User login
Treatment Trends and Outcomes in Healthcare-Associated Pneumonia
Bacterial pneumonia remains an important cause of morbidity and mortality in the United States, and is the 8th leading cause of death with 55,227 deaths among adults annually.1 In 2005, the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) collaborated to update guidelines for hospital-acquired pneumonia (HAP), ventilator-associated pneumonia, and healthcare-associated pneumonia (HCAP).2 This broad document outlines an evidence-based approach to diagnostic testing and antibiotic management based on the epidemiology and risk factors for these conditions. The guideline specifies the following criteria for HCAP: hospitalization in the past 90 days, residence in a skilled nursing facility (SNF), home infusion therapy, hemodialysis, home wound care, family members with multidrug resistant organisms (MDRO), and immunosuppressive diseases or medications, with the presumption that these patients are more likely to be harboring MDRO and should thus be treated empirically with broad-spectrum antibiotic therapy. Prior studies have shown that patients with HCAP have a more severe illness, are more likely to have MDRO, are more likely to be inadequately treated, and are at a higher risk for mortality than patients with community-acquired pneumonia (CAP).3,4
These guidelines are controversial, especially in regard to the recommendations to empirically treat broadly with 2 antibiotics targeting Pseudomonas species, whether patients with HCAP merit broader spectrum coverage than patients with CAP, and whether the criteria for defining HCAP are adequate to predict which patients are harboring MDRO. It has subsequently been proposed that HCAP is more related to CAP than to HAP, and a recent update to the guideline removed recommendations for treatment of HCAP and will be placing HCAP into the guidelines for CAP instead.5 We sought to investigate the degree of uptake of the ATS and IDSA guideline recommendations by physicians over time, and whether this led to a change in outcomes among patients who met the criteria for HCAP.
METHODS
Setting and Patients
We identified patients discharged between July 1, 2007, and November 30, 2011, from 488 US hospitals that participated in the Premier database (Premier Inc., Charlotte, North Carolina), an inpatient database developed for measuring quality and healthcare utilization. The database is frequently used for healthcare research and has been described previously.6 Member hospitals are in all regions of the US and are generally reflective of US hospitals. This database contains multiple data elements, including sociodemographic information, International Classification of Diseases, 9th Revision-Clinical Modification (ICD-9-CM) diagnosis and procedure codes, hospital and physician information, source of admission, and discharge status. It also includes a date-stamped log of all billed items and services, including diagnostic tests, medications, and other treatments. Because the data do not contain identifiable information, the institutional review board at our medical center determined that this study did not constitute human subjects research.
We included all patients aged ≥18 years with a principal diagnosis of pneumonia or with a secondary diagnosis of pneumonia paired with a principal diagnosis of respiratory failure, acute respiratory distress syndrome, respiratory arrest, sepsis, or influenza. Patients were excluded if they were transferred to or from another acute care institution, had a length of stay of 1 day or less, had cystic fibrosis, did not have a chest radiograph, or did not receive antibiotics within 48 hours of admission.
For each patient, we extracted age, gender, principal diagnosis, comorbidities, and the specialty of the attending physician. Comorbidities were identified from ICD-9-CM secondary diagnosis codes and Diagnosis Related Groups by using Healthcare Cost and Utilization Project Comorbidity Software, version 3.1, based on the work of Elixhauser (Agency for Healthcare Research and Quality, Rockville, Maryland).7 In order to ensure that patients had HCAP, we required the presence of ≥1 HCAP criteria, including hospitalization in the past 90 days, hemodialysis, admission from an SNF, or immune suppression (which was derived from either a secondary diagnosis for neutropenia, hematological malignancy, organ transplant, acquired immunodeficiency virus, or receiving immunosuppressant drugs or corticosteroids [equivalent to ≥20 mg/day of prednisone]).
Definitions of Guideline-Concordant and Discordant Antibiotic Therapy
The ATS and IDSA guidelines recommended the following antibiotic combinations for HCAP: an antipseudomonal cephalosporin or carbapenem or a beta-lactam/lactamase inhibitor, plus an antipseudomonal quinolone or aminoglycoside, plus an antibiotic with activity versus methicillin resistant Staphylococcus aureus (MRSA), such as vancomycin or linezolid. Based on these guidelines, we defined the receipt of fully guideline-concordant antibiotics as 2 recommended antibiotics for Pseudomonas species plus 1 for MRSA administered by the second day of admission. Partially guideline-concordant antibiotics were defined as 1 recommended antibiotic for Pseudomonas species plus 1 for MRSA by the second day of hospitalization. Guideline-discordant antibiotics were defined as all other combinations.
Statistical Analysis
Descriptive statistics on patient characteristics are presented as frequency, proportions for categorical factors, and median with interquartile range (IQR) for continuous variables for the full cohort and by treatment group, defined as fully or partially guideline-concordant antibiotic therapy or discordant therapy. Hospital rates of fully guideline-concordant treatment are presented overall and by hospital characteristics. The association of hospital characteristics with rates of fully guideline-concordant therapy were assessed by using 1-way analysis of variance tests.
To assess trends across hospitals for the association between the use of guideline-concordant therapy and mortality, progression to respiratory failure as measured by the late initiation of invasive mechanical ventilation (day 3 or later), and the length of stay among survivors, we divided the 4.5-year study period into 9 intervals of 6 months each; 292 hospitals that submitted data for all 9 time points were examined in this analysis. Based on the distribution of length of stay in the first time period, we created an indicator variable for extended length of stay with length of stay at or above the 75th percentile, defined as extended. For each hospital at each 6-month interval, we then computed risk-standardized guideline-concordant treatment (RS-treatment) rates and risk-standardized in-hospital outcome rates similar to methods used by the Centers for Medicare and Medicaid Services for public reporting.8 For each hospital at each time interval, we estimated a predicted rate of guideline-concordant treatment as the sum of predicted probabilities of guideline-concordant treatment from patient factors and the random intercept for the hospital in which they were admitted. We then calculated the expected rate of guideline-concordant treatment as the sum of expected probabilities of treatment received from patient factors only. RS-treatment was then calculated as the ratio of predicted to expected rates multiplied by the overall unadjusted mean treatment rate from all patients.9 We repeated the same modeling strategy to calculate risk-standardized outcome (RS-outcome) rates for each hospital across all time points. All models were adjusted for patient demographics and comorbidities. Similar models using administrative data have moderate discrimination for mortality.10
We then fit mixed-effects linear models with random hospital intercept and slope across time for the RS-treatment and outcome rates, respectively. From these models, we estimated the mean slope for RS-treatment and for RS-outcome over time. In addition, we estimated a slope or trend over time for each hospital for treatment and for outcome and evaluated the correlation between the treatment and outcome trends.
All analyses were performed using the Statistical Analysis System version 9.4 (SAS Institute Inc., Cary, NC) and STATA release 13 (StataCorp, LLC, College Station, Texas).
RESULTS
DISCUSSION
In this large, retrospective cohort study, we found that there was a substantial gap between the empiric antibiotics recommended by the ATS and IDSA guidelines and the empiric antibiotics that patients actually received. Over the study period, we saw an increased adherence to guidelines, in spite of growing evidence that HCAP risk factors do not adequately predict which patients are at risk for infection with an MDRO.11 We used this change in antibiotic prescribing behavior over time to determine if there was a clinical impact on patient outcomes and found that at the hospital level, there were no improvements in mortality, excess length of stay, or progression to respiratory failure despite a doubling in guideline-concordant antibiotic use.
At least 2 other large studies have assessed the association between guideline-concordant therapy and outcomes in HCAP.12,13 Both found that guideline-concordant therapy was associated with increased mortality, despite propensity matching. Both were conducted at the individual patient level by using administrative data, and results were likely affected by unmeasured clinical confounders, with sicker patients being more likely to receive guideline-concordant therapy. Our focus on the outcomes at the hospital level avoids this selection bias because the overall severity of illness of patients at any given hospital would not be expected to change over the study period, while physician uptake of antibiotic prescribing guidelines would be expected to increase over time. Determining the correlation between increases in guideline adherence and changes in patient outcome may offer a better assessment of the impact of guideline adherence. In this regard, our results are similar to those achieved by 1 quality improvement collaborative that was aimed at increasing guideline concordant therapy in ICUs. Despite an increase in guideline concordance from 33% to 47% of patients, they found no change in overall mortality.14
There were several limitations to our study. We did not have access to microbiologic data, so we were unable to determine which patients had MDRO infection or determine antibiotic-pathogen matching. However, the treating physicians in our study population presumably did not have access to this data at the time of treatment either because the time period we examined was within the first 48 hours of hospitalization, the interval during which cultures are incubating and the patients are being treated empirically. In addition, there may have been HCAP patients that we failed to identify, such as patients who were admitted in the past 90 days to a hospital that does not submit data to Premier. However, it is unlikely that prescribing for such patients should differ systematically from what we observed. While the database draws from 488 hospitals nationwide, it is possible that practices may be different at facilities that are not contained within the Premier database, such as Veterans Administration Hospitals. Similarly, we did not have readings for chest x-rays; hence, there could be some patients in the dataset who did not have pneumonia. However, we tried to overcome this by including only those patients with a principal diagnosis of pneumonia or sepsis with a secondary pneumonia diagnosis, a chest x-ray, and antibiotics administered within the first 48 hours of admission.
There are likely several reasons why so few HCAP patients in our study received guideline-concordant antibiotics. A lack of knowledge about the ATS and IDSA guidelines may have impacted the physicians in our study population. El-Solh et al.15 surveyed physicians about the ATS-IDSA guidelines 4 years after publication and found that only 45% were familiar with the document. We found that the rate of prescribing at least partially guideline-concordant antibiotics rose steadily over time, supporting the idea that the newness of the guidelines was 1 barrier. Additionally, prior studies have shown that many physicians may not agree with or choose to follow guidelines, with only 20% of physicians indicating that guidelines have a major impact on their clinical decision making,16 and the majority do not choose HCAP guideline-concordant antibiotics when tested.17 Alternatively, clinicians may not follow the guidelines because of a belief that the HCAP criteria do not adequately indicate patients who are at risk for MDRO. Previous studies have demonstrated the relative inability of HCAP risk factors to predict patients who harbor MDRO18 and suggest that better tools such as clinical scoring systems, which include not only the traditional HCAP risk factors but also prior exposure to antibiotics, prior culture data, and a cumulative assessment of both intrinsic and extrinsic factors, could more accurately predict MDRO and lead to a more judicious use of broad-spectrum antimicrobial agents.19-25 Indeed, these collective findings have led the authors of the recently updated guidelines to remove HCAP as a clinical entity from the hospital-acquired or ventilator-associated pneumonia guidelines and place them instead in the upcoming updated guidelines on the management of CAP.5 Of these 3 explanations, the lack of familiarity fits best with our observation that guideline-concordant therapy increased steadily over time with no evidence of reaching a plateau. Ironically, as consensus was building that HCAP is a poor marker for MDROs, routine empiric treatment with vancomycin and piperacillin-tazobactam (“vanco and zosyn”) have become routine in many hospitals. Additional studies are needed to know if this trend has stabilized or reversed.
CONCLUSIONS
In conclusion, clinicians in our large, nationally representative sample treated the majority of HCAP patients as though they had CAP. Although there was an increase in the administration of guideline-concordant therapy over time, this increase was not associated with improved outcomes. This study supports the growing consensus that HCAP criteria do not accurately predict which patients benefit from broad-spectrum antibiotics for pneumonia, and most patients fare well with antibiotics targeting common community-acquired organisms.
Disclosure
This work was supported by grant # R01HS018723 from the Agency for Healthcare Research and Quality. Dr. Lagu is also supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. Dr. Lindenauer is supported by grant K24HL132008 from the National Heart, Lung, and Blood Institute. The funding agency had no role in the data acquisition, analysis, or manuscript preparation for this study. Drs. Haessler and Rothberg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Haessler, Lagu, Lindenauer, Skiest, Zilberberg, Higgins, and Rothberg conceived of the study and analyzed and interpreted the data. Dr. Lindenauer acquired the data. Dr. Pekow and Ms. Priya carried out the statistical analyses. Dr. Haessler drafted the manuscript. All authors critically reviewed the manuscript for accuracy and integrity. All authors certify no potential conflicts of interest. Preliminary results from this study were presented in oral and poster format at IDWeek in 2012 and 2013.
1. Kochanek KD, Murphy SL, Xu JQ, Tejada-Vera B. Deaths: Final data for 2014. National vital statistics reports; vol 65 no 4. Hyattsville, MD: National Center for Health Statistics. 2016. PubMed
2. American Thoracic Society, Infectious Diseases Society of America. Guidelines for the Management of Adults with Hospital-acquired, Ventilator-associated, and Healthcare-associated Pneumonia. Am J Respir Crit Care Med. 2005;171(4):388-416. PubMed
3. Zilberberg MD, Shorr A. Healthcare-associated pneumonia: the state of the evidence to date. Curr Opin Pulm Med. 2011;17(3):142-147. PubMed
4. Kollef MK, Shorr A, Tabak YP, Gupta V, Liu LZ, Johannes RS. Epidemiology and Outcomes of Health-care-associated pneumonia. Chest. 2005;128(6):3854-3862. PubMed
5. Kalil AC, Metersky ML, Klompas M, et al. Management of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis. 2016;63(5):575-582. PubMed
6. Lindenauer PK, Pekow PS, Lahti MC, Lee Y, Benjamin EM, Rothberg MB. Association of corticosteroid dose and route of administration with risk of treatment failure in acute exacerbation of chronic obstructive pulmonary disease. JAMA. 2010;303(23):2359-2367. PubMed
7. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27. PubMed
8. Centers for Medicare & Medicaid Services. Frequently asked questions (FAQs): Implementation and maintenance of CMS mortality measures for AMI & HF. 2007. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQualityInits/downloads/HospitalMortalityAboutAMI_HF.pdf. Accessed November 1, 2016.
9. Normand SL, Shahian DM. Statistical and Clinical Aspects of Hospital Outcomes Profiling. Stat Sci. 2007;22(2):206-226.
10. Rothberg MB, Pekow PS, Priya A, et al. Using highly detailed administrative data to predict pneumonia mortality. PLoS One. 2014;9(1):e87382. PubMed
11. Jones BE, Jones MM, Huttner B, et al. Trends in antibiotic use and nosocomial pathogens in hospitalized veterans with pneumonia at 128 medical centers, 2006-2010. Clin Infect Dis. 2015;61(9):1403-1410. PubMed
12. Attridge RT, Frei CR, Restrepo MI, et al. Guideline-concordant therapy and outcomes in healthcare-associated pneumonia. Eur Respir J. 2011;38(4):878-887. PubMed
13. Rothberg MB, Zilberberg MD, Pekow PS, et al. Association of Guideline-based Antimicrobial Therapy and Outcomes in Healthcare-Associated Pneumonia. J Antimicrob Chemother. 2015;70(5):1573-1579. PubMed
14. Kett DH, Cano E, Quartin AA, et al. Improving Medicine through Pathway Assessment of Critical Therapy of Hospital-Acquired Pneumonia (IMPACT-HAP) Investigators. Implementation of guidelines for management of possible multidrug-resistant pneumonia in intensive care: an observational, multicentre cohort study. Lancet Infect Dis. 2011;11(3):181-189. PubMed
15. El-Solh AA, Alhajhusain A, Saliba RG, Drinka P. Physicians’ Attitudes Toward Guidelines for the Treatment of Hospitalized Nursing-Home -Acquired Pneumonia. J Am Med Dir Assoc. 2011;12(4):270-276. PubMed
16. Tunis S, Hayward R, Wilson M, et al. Internists’ Attitudes about Clinical Practice Guidelines. Ann Intern Med. 1994;120(11):956-963. PubMed
17. Seymann GB, Di Francesco L, Sharpe B, et al. The HCAP Gap: Differences between Self-Reported Practice Patterns and Published Guidelines for Health Care-Associated Pneumonia. Clin Infect Dis. 2009;49(12):1868-1874. PubMed
18. Chalmers JD, Rother C, Salih W, Ewig S. Healthcare associated pneumonia does not accurately identify potentially resistant pathogens: a systematic review and meta-analysis. Clin Infect Dis. 2014;58(3):330-339. PubMed
19. Shorr A, Zilberberg MD, Reichley R, et al. Validation of a Clinical Score for Assessing the Risk of Resistant Pathogens in Patients with Pneumonia Presenting to the Emergency Department. Clin Infect Dis. 2012;54(2):193-198. PubMed
20. Aliberti S, Pasquale MD, Zanaboni AM, et al. Stratifying Risk Factors for Multidrug-Resistant Pathogens in Hospitalized Patients Coming from the Community with Pneumonia. Clin Infect Dis. 2012;54(4):470-478. PubMed
21. Schreiber MP, Chan CM, Shorr AF. Resistant Pathogens in Nonnosocomial Pneumonia and Respiratory Failure: Is it Time to Refine the Definition of Health-care-Associated Pneumonia? Chest. 2010;137(6):1283-1288. PubMed
22. Madaras-Kelly KJ, Remington RE, Fan VS, Sloan KL. Predicting antibiotic resistance to community-acquired pneumonia antibiotics in culture-positive patients with healthcare-associated pneumonia. J Hosp Med. 2012;7(3):195-202. PubMed
23. Shindo Y, Ito R, Kobayashi D, et al. Risk factors for drug-resistant pathogens in community-acquired and healthcare-associated pneumonia. Am J Respir Crit Care Med. 2013;188(8):985-995. PubMed
24. Metersky ML, Frei CR, Mortensen EM. Predictors of Pseudomonas and methicillin-resistant Staphylococcus aureus in hospitalized patients with healthcare-associated pneumonia. Respirology. 2016;21(1):157-163. PubMed
25. Webb BJ, Dascomb K, Stenehjem E, Dean N. Predicting risk of drug-resistant organisms in pneumonia: moving beyond the HCAP model. Respir Med. 2015;109(1):1-10. PubMed
Bacterial pneumonia remains an important cause of morbidity and mortality in the United States, and is the 8th leading cause of death with 55,227 deaths among adults annually.1 In 2005, the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) collaborated to update guidelines for hospital-acquired pneumonia (HAP), ventilator-associated pneumonia, and healthcare-associated pneumonia (HCAP).2 This broad document outlines an evidence-based approach to diagnostic testing and antibiotic management based on the epidemiology and risk factors for these conditions. The guideline specifies the following criteria for HCAP: hospitalization in the past 90 days, residence in a skilled nursing facility (SNF), home infusion therapy, hemodialysis, home wound care, family members with multidrug resistant organisms (MDRO), and immunosuppressive diseases or medications, with the presumption that these patients are more likely to be harboring MDRO and should thus be treated empirically with broad-spectrum antibiotic therapy. Prior studies have shown that patients with HCAP have a more severe illness, are more likely to have MDRO, are more likely to be inadequately treated, and are at a higher risk for mortality than patients with community-acquired pneumonia (CAP).3,4
These guidelines are controversial, especially in regard to the recommendations to empirically treat broadly with 2 antibiotics targeting Pseudomonas species, whether patients with HCAP merit broader spectrum coverage than patients with CAP, and whether the criteria for defining HCAP are adequate to predict which patients are harboring MDRO. It has subsequently been proposed that HCAP is more related to CAP than to HAP, and a recent update to the guideline removed recommendations for treatment of HCAP and will be placing HCAP into the guidelines for CAP instead.5 We sought to investigate the degree of uptake of the ATS and IDSA guideline recommendations by physicians over time, and whether this led to a change in outcomes among patients who met the criteria for HCAP.
METHODS
Setting and Patients
We identified patients discharged between July 1, 2007, and November 30, 2011, from 488 US hospitals that participated in the Premier database (Premier Inc., Charlotte, North Carolina), an inpatient database developed for measuring quality and healthcare utilization. The database is frequently used for healthcare research and has been described previously.6 Member hospitals are in all regions of the US and are generally reflective of US hospitals. This database contains multiple data elements, including sociodemographic information, International Classification of Diseases, 9th Revision-Clinical Modification (ICD-9-CM) diagnosis and procedure codes, hospital and physician information, source of admission, and discharge status. It also includes a date-stamped log of all billed items and services, including diagnostic tests, medications, and other treatments. Because the data do not contain identifiable information, the institutional review board at our medical center determined that this study did not constitute human subjects research.
We included all patients aged ≥18 years with a principal diagnosis of pneumonia or with a secondary diagnosis of pneumonia paired with a principal diagnosis of respiratory failure, acute respiratory distress syndrome, respiratory arrest, sepsis, or influenza. Patients were excluded if they were transferred to or from another acute care institution, had a length of stay of 1 day or less, had cystic fibrosis, did not have a chest radiograph, or did not receive antibiotics within 48 hours of admission.
For each patient, we extracted age, gender, principal diagnosis, comorbidities, and the specialty of the attending physician. Comorbidities were identified from ICD-9-CM secondary diagnosis codes and Diagnosis Related Groups by using Healthcare Cost and Utilization Project Comorbidity Software, version 3.1, based on the work of Elixhauser (Agency for Healthcare Research and Quality, Rockville, Maryland).7 In order to ensure that patients had HCAP, we required the presence of ≥1 HCAP criteria, including hospitalization in the past 90 days, hemodialysis, admission from an SNF, or immune suppression (which was derived from either a secondary diagnosis for neutropenia, hematological malignancy, organ transplant, acquired immunodeficiency virus, or receiving immunosuppressant drugs or corticosteroids [equivalent to ≥20 mg/day of prednisone]).
Definitions of Guideline-Concordant and Discordant Antibiotic Therapy
The ATS and IDSA guidelines recommended the following antibiotic combinations for HCAP: an antipseudomonal cephalosporin or carbapenem or a beta-lactam/lactamase inhibitor, plus an antipseudomonal quinolone or aminoglycoside, plus an antibiotic with activity versus methicillin resistant Staphylococcus aureus (MRSA), such as vancomycin or linezolid. Based on these guidelines, we defined the receipt of fully guideline-concordant antibiotics as 2 recommended antibiotics for Pseudomonas species plus 1 for MRSA administered by the second day of admission. Partially guideline-concordant antibiotics were defined as 1 recommended antibiotic for Pseudomonas species plus 1 for MRSA by the second day of hospitalization. Guideline-discordant antibiotics were defined as all other combinations.
Statistical Analysis
Descriptive statistics on patient characteristics are presented as frequency, proportions for categorical factors, and median with interquartile range (IQR) for continuous variables for the full cohort and by treatment group, defined as fully or partially guideline-concordant antibiotic therapy or discordant therapy. Hospital rates of fully guideline-concordant treatment are presented overall and by hospital characteristics. The association of hospital characteristics with rates of fully guideline-concordant therapy were assessed by using 1-way analysis of variance tests.
To assess trends across hospitals for the association between the use of guideline-concordant therapy and mortality, progression to respiratory failure as measured by the late initiation of invasive mechanical ventilation (day 3 or later), and the length of stay among survivors, we divided the 4.5-year study period into 9 intervals of 6 months each; 292 hospitals that submitted data for all 9 time points were examined in this analysis. Based on the distribution of length of stay in the first time period, we created an indicator variable for extended length of stay with length of stay at or above the 75th percentile, defined as extended. For each hospital at each 6-month interval, we then computed risk-standardized guideline-concordant treatment (RS-treatment) rates and risk-standardized in-hospital outcome rates similar to methods used by the Centers for Medicare and Medicaid Services for public reporting.8 For each hospital at each time interval, we estimated a predicted rate of guideline-concordant treatment as the sum of predicted probabilities of guideline-concordant treatment from patient factors and the random intercept for the hospital in which they were admitted. We then calculated the expected rate of guideline-concordant treatment as the sum of expected probabilities of treatment received from patient factors only. RS-treatment was then calculated as the ratio of predicted to expected rates multiplied by the overall unadjusted mean treatment rate from all patients.9 We repeated the same modeling strategy to calculate risk-standardized outcome (RS-outcome) rates for each hospital across all time points. All models were adjusted for patient demographics and comorbidities. Similar models using administrative data have moderate discrimination for mortality.10
We then fit mixed-effects linear models with random hospital intercept and slope across time for the RS-treatment and outcome rates, respectively. From these models, we estimated the mean slope for RS-treatment and for RS-outcome over time. In addition, we estimated a slope or trend over time for each hospital for treatment and for outcome and evaluated the correlation between the treatment and outcome trends.
All analyses were performed using the Statistical Analysis System version 9.4 (SAS Institute Inc., Cary, NC) and STATA release 13 (StataCorp, LLC, College Station, Texas).
RESULTS
DISCUSSION
In this large, retrospective cohort study, we found that there was a substantial gap between the empiric antibiotics recommended by the ATS and IDSA guidelines and the empiric antibiotics that patients actually received. Over the study period, we saw an increased adherence to guidelines, in spite of growing evidence that HCAP risk factors do not adequately predict which patients are at risk for infection with an MDRO.11 We used this change in antibiotic prescribing behavior over time to determine if there was a clinical impact on patient outcomes and found that at the hospital level, there were no improvements in mortality, excess length of stay, or progression to respiratory failure despite a doubling in guideline-concordant antibiotic use.
At least 2 other large studies have assessed the association between guideline-concordant therapy and outcomes in HCAP.12,13 Both found that guideline-concordant therapy was associated with increased mortality, despite propensity matching. Both were conducted at the individual patient level by using administrative data, and results were likely affected by unmeasured clinical confounders, with sicker patients being more likely to receive guideline-concordant therapy. Our focus on the outcomes at the hospital level avoids this selection bias because the overall severity of illness of patients at any given hospital would not be expected to change over the study period, while physician uptake of antibiotic prescribing guidelines would be expected to increase over time. Determining the correlation between increases in guideline adherence and changes in patient outcome may offer a better assessment of the impact of guideline adherence. In this regard, our results are similar to those achieved by 1 quality improvement collaborative that was aimed at increasing guideline concordant therapy in ICUs. Despite an increase in guideline concordance from 33% to 47% of patients, they found no change in overall mortality.14
There were several limitations to our study. We did not have access to microbiologic data, so we were unable to determine which patients had MDRO infection or determine antibiotic-pathogen matching. However, the treating physicians in our study population presumably did not have access to this data at the time of treatment either because the time period we examined was within the first 48 hours of hospitalization, the interval during which cultures are incubating and the patients are being treated empirically. In addition, there may have been HCAP patients that we failed to identify, such as patients who were admitted in the past 90 days to a hospital that does not submit data to Premier. However, it is unlikely that prescribing for such patients should differ systematically from what we observed. While the database draws from 488 hospitals nationwide, it is possible that practices may be different at facilities that are not contained within the Premier database, such as Veterans Administration Hospitals. Similarly, we did not have readings for chest x-rays; hence, there could be some patients in the dataset who did not have pneumonia. However, we tried to overcome this by including only those patients with a principal diagnosis of pneumonia or sepsis with a secondary pneumonia diagnosis, a chest x-ray, and antibiotics administered within the first 48 hours of admission.
There are likely several reasons why so few HCAP patients in our study received guideline-concordant antibiotics. A lack of knowledge about the ATS and IDSA guidelines may have impacted the physicians in our study population. El-Solh et al.15 surveyed physicians about the ATS-IDSA guidelines 4 years after publication and found that only 45% were familiar with the document. We found that the rate of prescribing at least partially guideline-concordant antibiotics rose steadily over time, supporting the idea that the newness of the guidelines was 1 barrier. Additionally, prior studies have shown that many physicians may not agree with or choose to follow guidelines, with only 20% of physicians indicating that guidelines have a major impact on their clinical decision making,16 and the majority do not choose HCAP guideline-concordant antibiotics when tested.17 Alternatively, clinicians may not follow the guidelines because of a belief that the HCAP criteria do not adequately indicate patients who are at risk for MDRO. Previous studies have demonstrated the relative inability of HCAP risk factors to predict patients who harbor MDRO18 and suggest that better tools such as clinical scoring systems, which include not only the traditional HCAP risk factors but also prior exposure to antibiotics, prior culture data, and a cumulative assessment of both intrinsic and extrinsic factors, could more accurately predict MDRO and lead to a more judicious use of broad-spectrum antimicrobial agents.19-25 Indeed, these collective findings have led the authors of the recently updated guidelines to remove HCAP as a clinical entity from the hospital-acquired or ventilator-associated pneumonia guidelines and place them instead in the upcoming updated guidelines on the management of CAP.5 Of these 3 explanations, the lack of familiarity fits best with our observation that guideline-concordant therapy increased steadily over time with no evidence of reaching a plateau. Ironically, as consensus was building that HCAP is a poor marker for MDROs, routine empiric treatment with vancomycin and piperacillin-tazobactam (“vanco and zosyn”) have become routine in many hospitals. Additional studies are needed to know if this trend has stabilized or reversed.
CONCLUSIONS
In conclusion, clinicians in our large, nationally representative sample treated the majority of HCAP patients as though they had CAP. Although there was an increase in the administration of guideline-concordant therapy over time, this increase was not associated with improved outcomes. This study supports the growing consensus that HCAP criteria do not accurately predict which patients benefit from broad-spectrum antibiotics for pneumonia, and most patients fare well with antibiotics targeting common community-acquired organisms.
Disclosure
This work was supported by grant # R01HS018723 from the Agency for Healthcare Research and Quality. Dr. Lagu is also supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. Dr. Lindenauer is supported by grant K24HL132008 from the National Heart, Lung, and Blood Institute. The funding agency had no role in the data acquisition, analysis, or manuscript preparation for this study. Drs. Haessler and Rothberg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Haessler, Lagu, Lindenauer, Skiest, Zilberberg, Higgins, and Rothberg conceived of the study and analyzed and interpreted the data. Dr. Lindenauer acquired the data. Dr. Pekow and Ms. Priya carried out the statistical analyses. Dr. Haessler drafted the manuscript. All authors critically reviewed the manuscript for accuracy and integrity. All authors certify no potential conflicts of interest. Preliminary results from this study were presented in oral and poster format at IDWeek in 2012 and 2013.
Bacterial pneumonia remains an important cause of morbidity and mortality in the United States, and is the 8th leading cause of death with 55,227 deaths among adults annually.1 In 2005, the American Thoracic Society (ATS) and the Infectious Diseases Society of America (IDSA) collaborated to update guidelines for hospital-acquired pneumonia (HAP), ventilator-associated pneumonia, and healthcare-associated pneumonia (HCAP).2 This broad document outlines an evidence-based approach to diagnostic testing and antibiotic management based on the epidemiology and risk factors for these conditions. The guideline specifies the following criteria for HCAP: hospitalization in the past 90 days, residence in a skilled nursing facility (SNF), home infusion therapy, hemodialysis, home wound care, family members with multidrug resistant organisms (MDRO), and immunosuppressive diseases or medications, with the presumption that these patients are more likely to be harboring MDRO and should thus be treated empirically with broad-spectrum antibiotic therapy. Prior studies have shown that patients with HCAP have a more severe illness, are more likely to have MDRO, are more likely to be inadequately treated, and are at a higher risk for mortality than patients with community-acquired pneumonia (CAP).3,4
These guidelines are controversial, especially in regard to the recommendations to empirically treat broadly with 2 antibiotics targeting Pseudomonas species, whether patients with HCAP merit broader spectrum coverage than patients with CAP, and whether the criteria for defining HCAP are adequate to predict which patients are harboring MDRO. It has subsequently been proposed that HCAP is more related to CAP than to HAP, and a recent update to the guideline removed recommendations for treatment of HCAP and will be placing HCAP into the guidelines for CAP instead.5 We sought to investigate the degree of uptake of the ATS and IDSA guideline recommendations by physicians over time, and whether this led to a change in outcomes among patients who met the criteria for HCAP.
METHODS
Setting and Patients
We identified patients discharged between July 1, 2007, and November 30, 2011, from 488 US hospitals that participated in the Premier database (Premier Inc., Charlotte, North Carolina), an inpatient database developed for measuring quality and healthcare utilization. The database is frequently used for healthcare research and has been described previously.6 Member hospitals are in all regions of the US and are generally reflective of US hospitals. This database contains multiple data elements, including sociodemographic information, International Classification of Diseases, 9th Revision-Clinical Modification (ICD-9-CM) diagnosis and procedure codes, hospital and physician information, source of admission, and discharge status. It also includes a date-stamped log of all billed items and services, including diagnostic tests, medications, and other treatments. Because the data do not contain identifiable information, the institutional review board at our medical center determined that this study did not constitute human subjects research.
We included all patients aged ≥18 years with a principal diagnosis of pneumonia or with a secondary diagnosis of pneumonia paired with a principal diagnosis of respiratory failure, acute respiratory distress syndrome, respiratory arrest, sepsis, or influenza. Patients were excluded if they were transferred to or from another acute care institution, had a length of stay of 1 day or less, had cystic fibrosis, did not have a chest radiograph, or did not receive antibiotics within 48 hours of admission.
For each patient, we extracted age, gender, principal diagnosis, comorbidities, and the specialty of the attending physician. Comorbidities were identified from ICD-9-CM secondary diagnosis codes and Diagnosis Related Groups by using Healthcare Cost and Utilization Project Comorbidity Software, version 3.1, based on the work of Elixhauser (Agency for Healthcare Research and Quality, Rockville, Maryland).7 In order to ensure that patients had HCAP, we required the presence of ≥1 HCAP criteria, including hospitalization in the past 90 days, hemodialysis, admission from an SNF, or immune suppression (which was derived from either a secondary diagnosis for neutropenia, hematological malignancy, organ transplant, acquired immunodeficiency virus, or receiving immunosuppressant drugs or corticosteroids [equivalent to ≥20 mg/day of prednisone]).
Definitions of Guideline-Concordant and Discordant Antibiotic Therapy
The ATS and IDSA guidelines recommended the following antibiotic combinations for HCAP: an antipseudomonal cephalosporin or carbapenem or a beta-lactam/lactamase inhibitor, plus an antipseudomonal quinolone or aminoglycoside, plus an antibiotic with activity versus methicillin resistant Staphylococcus aureus (MRSA), such as vancomycin or linezolid. Based on these guidelines, we defined the receipt of fully guideline-concordant antibiotics as 2 recommended antibiotics for Pseudomonas species plus 1 for MRSA administered by the second day of admission. Partially guideline-concordant antibiotics were defined as 1 recommended antibiotic for Pseudomonas species plus 1 for MRSA by the second day of hospitalization. Guideline-discordant antibiotics were defined as all other combinations.
Statistical Analysis
Descriptive statistics on patient characteristics are presented as frequency, proportions for categorical factors, and median with interquartile range (IQR) for continuous variables for the full cohort and by treatment group, defined as fully or partially guideline-concordant antibiotic therapy or discordant therapy. Hospital rates of fully guideline-concordant treatment are presented overall and by hospital characteristics. The association of hospital characteristics with rates of fully guideline-concordant therapy were assessed by using 1-way analysis of variance tests.
To assess trends across hospitals for the association between the use of guideline-concordant therapy and mortality, progression to respiratory failure as measured by the late initiation of invasive mechanical ventilation (day 3 or later), and the length of stay among survivors, we divided the 4.5-year study period into 9 intervals of 6 months each; 292 hospitals that submitted data for all 9 time points were examined in this analysis. Based on the distribution of length of stay in the first time period, we created an indicator variable for extended length of stay with length of stay at or above the 75th percentile, defined as extended. For each hospital at each 6-month interval, we then computed risk-standardized guideline-concordant treatment (RS-treatment) rates and risk-standardized in-hospital outcome rates similar to methods used by the Centers for Medicare and Medicaid Services for public reporting.8 For each hospital at each time interval, we estimated a predicted rate of guideline-concordant treatment as the sum of predicted probabilities of guideline-concordant treatment from patient factors and the random intercept for the hospital in which they were admitted. We then calculated the expected rate of guideline-concordant treatment as the sum of expected probabilities of treatment received from patient factors only. RS-treatment was then calculated as the ratio of predicted to expected rates multiplied by the overall unadjusted mean treatment rate from all patients.9 We repeated the same modeling strategy to calculate risk-standardized outcome (RS-outcome) rates for each hospital across all time points. All models were adjusted for patient demographics and comorbidities. Similar models using administrative data have moderate discrimination for mortality.10
We then fit mixed-effects linear models with random hospital intercept and slope across time for the RS-treatment and outcome rates, respectively. From these models, we estimated the mean slope for RS-treatment and for RS-outcome over time. In addition, we estimated a slope or trend over time for each hospital for treatment and for outcome and evaluated the correlation between the treatment and outcome trends.
All analyses were performed using the Statistical Analysis System version 9.4 (SAS Institute Inc., Cary, NC) and STATA release 13 (StataCorp, LLC, College Station, Texas).
RESULTS
DISCUSSION
In this large, retrospective cohort study, we found that there was a substantial gap between the empiric antibiotics recommended by the ATS and IDSA guidelines and the empiric antibiotics that patients actually received. Over the study period, we saw an increased adherence to guidelines, in spite of growing evidence that HCAP risk factors do not adequately predict which patients are at risk for infection with an MDRO.11 We used this change in antibiotic prescribing behavior over time to determine if there was a clinical impact on patient outcomes and found that at the hospital level, there were no improvements in mortality, excess length of stay, or progression to respiratory failure despite a doubling in guideline-concordant antibiotic use.
At least 2 other large studies have assessed the association between guideline-concordant therapy and outcomes in HCAP.12,13 Both found that guideline-concordant therapy was associated with increased mortality, despite propensity matching. Both were conducted at the individual patient level by using administrative data, and results were likely affected by unmeasured clinical confounders, with sicker patients being more likely to receive guideline-concordant therapy. Our focus on the outcomes at the hospital level avoids this selection bias because the overall severity of illness of patients at any given hospital would not be expected to change over the study period, while physician uptake of antibiotic prescribing guidelines would be expected to increase over time. Determining the correlation between increases in guideline adherence and changes in patient outcome may offer a better assessment of the impact of guideline adherence. In this regard, our results are similar to those achieved by 1 quality improvement collaborative that was aimed at increasing guideline concordant therapy in ICUs. Despite an increase in guideline concordance from 33% to 47% of patients, they found no change in overall mortality.14
There were several limitations to our study. We did not have access to microbiologic data, so we were unable to determine which patients had MDRO infection or determine antibiotic-pathogen matching. However, the treating physicians in our study population presumably did not have access to this data at the time of treatment either because the time period we examined was within the first 48 hours of hospitalization, the interval during which cultures are incubating and the patients are being treated empirically. In addition, there may have been HCAP patients that we failed to identify, such as patients who were admitted in the past 90 days to a hospital that does not submit data to Premier. However, it is unlikely that prescribing for such patients should differ systematically from what we observed. While the database draws from 488 hospitals nationwide, it is possible that practices may be different at facilities that are not contained within the Premier database, such as Veterans Administration Hospitals. Similarly, we did not have readings for chest x-rays; hence, there could be some patients in the dataset who did not have pneumonia. However, we tried to overcome this by including only those patients with a principal diagnosis of pneumonia or sepsis with a secondary pneumonia diagnosis, a chest x-ray, and antibiotics administered within the first 48 hours of admission.
There are likely several reasons why so few HCAP patients in our study received guideline-concordant antibiotics. A lack of knowledge about the ATS and IDSA guidelines may have impacted the physicians in our study population. El-Solh et al.15 surveyed physicians about the ATS-IDSA guidelines 4 years after publication and found that only 45% were familiar with the document. We found that the rate of prescribing at least partially guideline-concordant antibiotics rose steadily over time, supporting the idea that the newness of the guidelines was 1 barrier. Additionally, prior studies have shown that many physicians may not agree with or choose to follow guidelines, with only 20% of physicians indicating that guidelines have a major impact on their clinical decision making,16 and the majority do not choose HCAP guideline-concordant antibiotics when tested.17 Alternatively, clinicians may not follow the guidelines because of a belief that the HCAP criteria do not adequately indicate patients who are at risk for MDRO. Previous studies have demonstrated the relative inability of HCAP risk factors to predict patients who harbor MDRO18 and suggest that better tools such as clinical scoring systems, which include not only the traditional HCAP risk factors but also prior exposure to antibiotics, prior culture data, and a cumulative assessment of both intrinsic and extrinsic factors, could more accurately predict MDRO and lead to a more judicious use of broad-spectrum antimicrobial agents.19-25 Indeed, these collective findings have led the authors of the recently updated guidelines to remove HCAP as a clinical entity from the hospital-acquired or ventilator-associated pneumonia guidelines and place them instead in the upcoming updated guidelines on the management of CAP.5 Of these 3 explanations, the lack of familiarity fits best with our observation that guideline-concordant therapy increased steadily over time with no evidence of reaching a plateau. Ironically, as consensus was building that HCAP is a poor marker for MDROs, routine empiric treatment with vancomycin and piperacillin-tazobactam (“vanco and zosyn”) have become routine in many hospitals. Additional studies are needed to know if this trend has stabilized or reversed.
CONCLUSIONS
In conclusion, clinicians in our large, nationally representative sample treated the majority of HCAP patients as though they had CAP. Although there was an increase in the administration of guideline-concordant therapy over time, this increase was not associated with improved outcomes. This study supports the growing consensus that HCAP criteria do not accurately predict which patients benefit from broad-spectrum antibiotics for pneumonia, and most patients fare well with antibiotics targeting common community-acquired organisms.
Disclosure
This work was supported by grant # R01HS018723 from the Agency for Healthcare Research and Quality. Dr. Lagu is also supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. Dr. Lindenauer is supported by grant K24HL132008 from the National Heart, Lung, and Blood Institute. The funding agency had no role in the data acquisition, analysis, or manuscript preparation for this study. Drs. Haessler and Rothberg had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs. Haessler, Lagu, Lindenauer, Skiest, Zilberberg, Higgins, and Rothberg conceived of the study and analyzed and interpreted the data. Dr. Lindenauer acquired the data. Dr. Pekow and Ms. Priya carried out the statistical analyses. Dr. Haessler drafted the manuscript. All authors critically reviewed the manuscript for accuracy and integrity. All authors certify no potential conflicts of interest. Preliminary results from this study were presented in oral and poster format at IDWeek in 2012 and 2013.
1. Kochanek KD, Murphy SL, Xu JQ, Tejada-Vera B. Deaths: Final data for 2014. National vital statistics reports; vol 65 no 4. Hyattsville, MD: National Center for Health Statistics. 2016. PubMed
2. American Thoracic Society, Infectious Diseases Society of America. Guidelines for the Management of Adults with Hospital-acquired, Ventilator-associated, and Healthcare-associated Pneumonia. Am J Respir Crit Care Med. 2005;171(4):388-416. PubMed
3. Zilberberg MD, Shorr A. Healthcare-associated pneumonia: the state of the evidence to date. Curr Opin Pulm Med. 2011;17(3):142-147. PubMed
4. Kollef MK, Shorr A, Tabak YP, Gupta V, Liu LZ, Johannes RS. Epidemiology and Outcomes of Health-care-associated pneumonia. Chest. 2005;128(6):3854-3862. PubMed
5. Kalil AC, Metersky ML, Klompas M, et al. Management of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis. 2016;63(5):575-582. PubMed
6. Lindenauer PK, Pekow PS, Lahti MC, Lee Y, Benjamin EM, Rothberg MB. Association of corticosteroid dose and route of administration with risk of treatment failure in acute exacerbation of chronic obstructive pulmonary disease. JAMA. 2010;303(23):2359-2367. PubMed
7. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27. PubMed
8. Centers for Medicare & Medicaid Services. Frequently asked questions (FAQs): Implementation and maintenance of CMS mortality measures for AMI & HF. 2007. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQualityInits/downloads/HospitalMortalityAboutAMI_HF.pdf. Accessed November 1, 2016.
9. Normand SL, Shahian DM. Statistical and Clinical Aspects of Hospital Outcomes Profiling. Stat Sci. 2007;22(2):206-226.
10. Rothberg MB, Pekow PS, Priya A, et al. Using highly detailed administrative data to predict pneumonia mortality. PLoS One. 2014;9(1):e87382. PubMed
11. Jones BE, Jones MM, Huttner B, et al. Trends in antibiotic use and nosocomial pathogens in hospitalized veterans with pneumonia at 128 medical centers, 2006-2010. Clin Infect Dis. 2015;61(9):1403-1410. PubMed
12. Attridge RT, Frei CR, Restrepo MI, et al. Guideline-concordant therapy and outcomes in healthcare-associated pneumonia. Eur Respir J. 2011;38(4):878-887. PubMed
13. Rothberg MB, Zilberberg MD, Pekow PS, et al. Association of Guideline-based Antimicrobial Therapy and Outcomes in Healthcare-Associated Pneumonia. J Antimicrob Chemother. 2015;70(5):1573-1579. PubMed
14. Kett DH, Cano E, Quartin AA, et al. Improving Medicine through Pathway Assessment of Critical Therapy of Hospital-Acquired Pneumonia (IMPACT-HAP) Investigators. Implementation of guidelines for management of possible multidrug-resistant pneumonia in intensive care: an observational, multicentre cohort study. Lancet Infect Dis. 2011;11(3):181-189. PubMed
15. El-Solh AA, Alhajhusain A, Saliba RG, Drinka P. Physicians’ Attitudes Toward Guidelines for the Treatment of Hospitalized Nursing-Home -Acquired Pneumonia. J Am Med Dir Assoc. 2011;12(4):270-276. PubMed
16. Tunis S, Hayward R, Wilson M, et al. Internists’ Attitudes about Clinical Practice Guidelines. Ann Intern Med. 1994;120(11):956-963. PubMed
17. Seymann GB, Di Francesco L, Sharpe B, et al. The HCAP Gap: Differences between Self-Reported Practice Patterns and Published Guidelines for Health Care-Associated Pneumonia. Clin Infect Dis. 2009;49(12):1868-1874. PubMed
18. Chalmers JD, Rother C, Salih W, Ewig S. Healthcare associated pneumonia does not accurately identify potentially resistant pathogens: a systematic review and meta-analysis. Clin Infect Dis. 2014;58(3):330-339. PubMed
19. Shorr A, Zilberberg MD, Reichley R, et al. Validation of a Clinical Score for Assessing the Risk of Resistant Pathogens in Patients with Pneumonia Presenting to the Emergency Department. Clin Infect Dis. 2012;54(2):193-198. PubMed
20. Aliberti S, Pasquale MD, Zanaboni AM, et al. Stratifying Risk Factors for Multidrug-Resistant Pathogens in Hospitalized Patients Coming from the Community with Pneumonia. Clin Infect Dis. 2012;54(4):470-478. PubMed
21. Schreiber MP, Chan CM, Shorr AF. Resistant Pathogens in Nonnosocomial Pneumonia and Respiratory Failure: Is it Time to Refine the Definition of Health-care-Associated Pneumonia? Chest. 2010;137(6):1283-1288. PubMed
22. Madaras-Kelly KJ, Remington RE, Fan VS, Sloan KL. Predicting antibiotic resistance to community-acquired pneumonia antibiotics in culture-positive patients with healthcare-associated pneumonia. J Hosp Med. 2012;7(3):195-202. PubMed
23. Shindo Y, Ito R, Kobayashi D, et al. Risk factors for drug-resistant pathogens in community-acquired and healthcare-associated pneumonia. Am J Respir Crit Care Med. 2013;188(8):985-995. PubMed
24. Metersky ML, Frei CR, Mortensen EM. Predictors of Pseudomonas and methicillin-resistant Staphylococcus aureus in hospitalized patients with healthcare-associated pneumonia. Respirology. 2016;21(1):157-163. PubMed
25. Webb BJ, Dascomb K, Stenehjem E, Dean N. Predicting risk of drug-resistant organisms in pneumonia: moving beyond the HCAP model. Respir Med. 2015;109(1):1-10. PubMed
1. Kochanek KD, Murphy SL, Xu JQ, Tejada-Vera B. Deaths: Final data for 2014. National vital statistics reports; vol 65 no 4. Hyattsville, MD: National Center for Health Statistics. 2016. PubMed
2. American Thoracic Society, Infectious Diseases Society of America. Guidelines for the Management of Adults with Hospital-acquired, Ventilator-associated, and Healthcare-associated Pneumonia. Am J Respir Crit Care Med. 2005;171(4):388-416. PubMed
3. Zilberberg MD, Shorr A. Healthcare-associated pneumonia: the state of the evidence to date. Curr Opin Pulm Med. 2011;17(3):142-147. PubMed
4. Kollef MK, Shorr A, Tabak YP, Gupta V, Liu LZ, Johannes RS. Epidemiology and Outcomes of Health-care-associated pneumonia. Chest. 2005;128(6):3854-3862. PubMed
5. Kalil AC, Metersky ML, Klompas M, et al. Management of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society. Clin Infect Dis. 2016;63(5):575-582. PubMed
6. Lindenauer PK, Pekow PS, Lahti MC, Lee Y, Benjamin EM, Rothberg MB. Association of corticosteroid dose and route of administration with risk of treatment failure in acute exacerbation of chronic obstructive pulmonary disease. JAMA. 2010;303(23):2359-2367. PubMed
7. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27. PubMed
8. Centers for Medicare & Medicaid Services. Frequently asked questions (FAQs): Implementation and maintenance of CMS mortality measures for AMI & HF. 2007. https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/HospitalQualityInits/downloads/HospitalMortalityAboutAMI_HF.pdf. Accessed November 1, 2016.
9. Normand SL, Shahian DM. Statistical and Clinical Aspects of Hospital Outcomes Profiling. Stat Sci. 2007;22(2):206-226.
10. Rothberg MB, Pekow PS, Priya A, et al. Using highly detailed administrative data to predict pneumonia mortality. PLoS One. 2014;9(1):e87382. PubMed
11. Jones BE, Jones MM, Huttner B, et al. Trends in antibiotic use and nosocomial pathogens in hospitalized veterans with pneumonia at 128 medical centers, 2006-2010. Clin Infect Dis. 2015;61(9):1403-1410. PubMed
12. Attridge RT, Frei CR, Restrepo MI, et al. Guideline-concordant therapy and outcomes in healthcare-associated pneumonia. Eur Respir J. 2011;38(4):878-887. PubMed
13. Rothberg MB, Zilberberg MD, Pekow PS, et al. Association of Guideline-based Antimicrobial Therapy and Outcomes in Healthcare-Associated Pneumonia. J Antimicrob Chemother. 2015;70(5):1573-1579. PubMed
14. Kett DH, Cano E, Quartin AA, et al. Improving Medicine through Pathway Assessment of Critical Therapy of Hospital-Acquired Pneumonia (IMPACT-HAP) Investigators. Implementation of guidelines for management of possible multidrug-resistant pneumonia in intensive care: an observational, multicentre cohort study. Lancet Infect Dis. 2011;11(3):181-189. PubMed
15. El-Solh AA, Alhajhusain A, Saliba RG, Drinka P. Physicians’ Attitudes Toward Guidelines for the Treatment of Hospitalized Nursing-Home -Acquired Pneumonia. J Am Med Dir Assoc. 2011;12(4):270-276. PubMed
16. Tunis S, Hayward R, Wilson M, et al. Internists’ Attitudes about Clinical Practice Guidelines. Ann Intern Med. 1994;120(11):956-963. PubMed
17. Seymann GB, Di Francesco L, Sharpe B, et al. The HCAP Gap: Differences between Self-Reported Practice Patterns and Published Guidelines for Health Care-Associated Pneumonia. Clin Infect Dis. 2009;49(12):1868-1874. PubMed
18. Chalmers JD, Rother C, Salih W, Ewig S. Healthcare associated pneumonia does not accurately identify potentially resistant pathogens: a systematic review and meta-analysis. Clin Infect Dis. 2014;58(3):330-339. PubMed
19. Shorr A, Zilberberg MD, Reichley R, et al. Validation of a Clinical Score for Assessing the Risk of Resistant Pathogens in Patients with Pneumonia Presenting to the Emergency Department. Clin Infect Dis. 2012;54(2):193-198. PubMed
20. Aliberti S, Pasquale MD, Zanaboni AM, et al. Stratifying Risk Factors for Multidrug-Resistant Pathogens in Hospitalized Patients Coming from the Community with Pneumonia. Clin Infect Dis. 2012;54(4):470-478. PubMed
21. Schreiber MP, Chan CM, Shorr AF. Resistant Pathogens in Nonnosocomial Pneumonia and Respiratory Failure: Is it Time to Refine the Definition of Health-care-Associated Pneumonia? Chest. 2010;137(6):1283-1288. PubMed
22. Madaras-Kelly KJ, Remington RE, Fan VS, Sloan KL. Predicting antibiotic resistance to community-acquired pneumonia antibiotics in culture-positive patients with healthcare-associated pneumonia. J Hosp Med. 2012;7(3):195-202. PubMed
23. Shindo Y, Ito R, Kobayashi D, et al. Risk factors for drug-resistant pathogens in community-acquired and healthcare-associated pneumonia. Am J Respir Crit Care Med. 2013;188(8):985-995. PubMed
24. Metersky ML, Frei CR, Mortensen EM. Predictors of Pseudomonas and methicillin-resistant Staphylococcus aureus in hospitalized patients with healthcare-associated pneumonia. Respirology. 2016;21(1):157-163. PubMed
25. Webb BJ, Dascomb K, Stenehjem E, Dean N. Predicting risk of drug-resistant organisms in pneumonia: moving beyond the HCAP model. Respir Med. 2015;109(1):1-10. PubMed
© 2017 Society of Hospital Medicine
Agreement on Dyspnea Severity
Breathlessness, or dyspnea, is defined as a subjective experience of breathing discomfort that is comprised of qualitatively distinct sensations that vary in intensity.[1] Dyspnea is a leading reason for patients presenting for emergency care,[2] and it is an important predictor for hospitalization and mortality in patients with cardiopulmonary disease.[3, 4, 5]
Several professional societies' guidelines recommend that patients should be asked to quantify the intensity of their breathlessness using a standardized scale, and that these ratings should be documented in medical records to guide dyspnea awareness and management.[1, 6, 7] During the evaluation and treatment of patients with acute cardiopulmonary conditions, the clinician estimates the severity of the illness and response to therapy based on multiple objective measures as well as the patient's perception of dyspnea. A patient‐centered care approach depends upon the physicians having a shared understanding of what the patient is experiencing. Without this appreciation, the healthcare provider cannot make appropriate treatment decisions to ensure alleviation of presenting symptoms. Understanding the severity of patients' dyspnea is critical to avoid under‐ or overtreatment of patients with acute cardiopulmonary conditions, but only a few studies have compared patient and provider perceptions of dyspnea intensity.[8, 9] Discordance between physician's impression of severity of dyspnea and patient's perception may result in suboptimal management and patient dissatisfaction with care. Furthermore, several studies have shown that, when physicians and patients agree with the assessment of well‐being, treatment adherence and outcomes improve.[10, 11]
Therefore, we evaluated the extent and directionality of agreement between patients' perception and healthcare providers' impression of dyspnea and explored which factors contribute to discordance. Additionally, we examined how healthcare providers document dyspnea severity.
METHODS
Study Setting and Population
The study was conducted between June 2012 and August 2012 at Baystate Medical Center (BMC), a 740‐bed tertiary care hospital in western Massachusetts. In 2012, the BMC hospitalist group had 48 attending physicians, of whom 47% were female, 48% had 0 to 3 years of attending experience, and 16% had 10 years of experience.
We enrolled consecutive admissions of English‐speaking adult patients, with a working diagnosis of heart failure (HF), chronic obstructive pulmonary disease (COPD), asthma, pneumonia, or a generic diagnosis of shortness of breath. Because we surveyed only hospitalists, we did not include patients admitted to an intensive care unit.
All participants gave informed consent to be part of the study. The research protocol was approved by the Baystate Health Institutional Review Board, Springfield, Massachusetts.
Dyspnea Assessment
Dyspnea intensity was assessed on an 11‐point (010) numerical rating scale (NRS).[12, 13] A trained research assistant interviewed patients on day 1 and 2 after admission, and on the day of discharge between 8 am and 12 pm on weekdays. The patient was asked: On a scale from 0 to 10, how bad is your shortness of breath at rest now, with 0 being no shortness of breath and 10 the worst shortness of breath you could ever imagine? The hospitalist or the senior resident and day‐shift nurse taking care of the patients were asked by the research assistant to rate the patient's dyspnea using the same scoring instrument shortly after they saw the patient. The physicians and nurses based their determination of dyspnea on their usual interview/examination of the patient. The patient, physician and the nurse were not aware of each other's rating. The research assistant scheduled the interviews to minimize the time intervals between the patient assessment and provider's rating. For this reason, the number of assessments per patient varied. Nurses were more readily available than physicians, which resulted in a larger number of patient‐nurse response pairs than patient‐physician pairs. All assessments were done in the morning, between 9 am and 12 pm, with a range of 3 hours between provider's assessment of the patient and the interview.
Dyspnea Agreement
Agreement was defined as a score within 1 between patient and healthcare provider; differences of 2 points were considered over‐ or underestimations. The decision to use this cutoff was based on prior studies, which found that a difference in the range of 1.6 to 2.2 cm was meaningful for the patient when assessment was done on the visual analog scale.[8, 14, 15] We also evaluated the direction of discordance. If the patient's rating of dyspnea severity was higher than the provider's rating, we defined this as underestimation by the provider; in the instance where a provider's score of dyspnea severity was higher than the patient's score, we defined this as overestimation. In a sensitivity analysis, agreement was defined as a score within 2 between patient and healthcare provider, and any difference 3 was considered disagreement.
Other Variables
We obtained information from the medical records about patient demographics, body mass index (BMI), smoking status, and vital signs. We calculated the oxygen saturation index as the ratio between the oxygen saturation and the fractional inspired oxygen (SpO2/FIO2). Comorbidities were assessed based on the International Classification od Diseases, Ninth Revision, Clinical Modification codes from the hospital financial decision support system. We calculated an overall combined comorbidity score based on the method described by Gagne, which is based on elements from the Charlson Comorbidity Index and from the Elixhauser comorbidities.[16]
Charts of the patients included in the study were retrospectively reviewed for physicians' and nurses' documentation of dyspnea at admission and at discharge. We recorded if dyspnea was mentioned and how it was assessed: whether it was described as present/absent; graded as mild, moderate, or severe; used a quantitative scale (010); used descriptors (eg, dyspnea when climbing stairs); and whether it was defined as improved or worsened without other qualifiers.
Statistical Analysis
Descriptive statistics of dyspnea scores, patient characteristics, comorbidities and vital signs were calculated and presented as medians with interquartile range (IQR) for continuous variables, and counts with percentages for categorical factors. Every patient‐provider concurrent scoring was included in the analysis as 1 dyad, which resulted in patients being included multiple times in the analysis. Patient‐physician and patient‐nurse dyads of dyspnea assessment were examined separately. Analyses included all dyads that were within same assessment period (same day and same time window).
The relationship between patient self‐perceived dyspnea severity and provider rating of was assessed in several ways. First, a weighted kappa coefficient was used as a measure of agreement between patient and nurse or physician scores. A weighted kappa analysis was chosen because it penalizes disagreements that are further apart from each other.
Second, we defined an indicator of discordance and constructed multivariable generalized estimating equation models that account for clustering of multiple dyads per patient, to assess the relationship of patient characteristics with discordance. Finally, we developed additional models to predict underestimation when compared to agreement or overestimation of dyspnea by the healthcare provider relative to the patient. Using the same definitions for agreement, we also compared the dyspnea assessment estimation between physicians and nurses.
We present the differences in dyspnea assessment between patient and healthcare provider and between nurses and physicians by Bland‐Altman plots.
All analyses were performed using SAS (version 9.3; SAS institute, Inc., Cary, NC), Stata (Stata statistical software release 13; StataCorp, College Station, TX), and RStudio version 0.99.892 (Bland‐Altman plots, R package version 0.3.1; The R Foundation for Statistical Computing, Vienna, Austria).[8, 9, 17]
RESULTS
Patient Characteristics
Among the 219 patients who met the screening criteria, 81 were not enrolled (Figure 1). Data from 138 patients, with both patient information and provider data on dyspnea assessment, were included. The median age of the patients was 72 years (IQR, 5880 years), 56.5% were women, 75.4% were white, and 28.3% were current smokers. Approximately 30% had a diagnosis of HF, 30% of COPD, and 13.0% of pneumonia. The median comorbidity score was 4 (IQR, 26), and 37.0% of the patients had a BMI 30. At admission, the median oxygen saturation index was 346 (IQR, 287.5460) indicating mild to moderate levels of hypoxia. (Table 1).
| Value | |
|---|---|
| |
| Age, median (IQR), y | 72 (5880) |
| Gender | |
| Female | 78 (56.5) |
| Male | 60 (43.5) |
| Race | |
| White | 104 (75.4) |
| Black | 16 (11.6) |
| Hispanic | 17 (12.3) |
| Other | 1 (0.7) |
| Body mass index, median (IQR) | 28 (23.334.6) |
| Obese (BMI 30) | 51 (37.0) |
| Smoker, current | 39 (28.3) |
| Admitting diagnosis | |
| Heart failure | 46 (33.3) |
| COPD/asthma | 41 (29.7) |
| Pneumonia | 18 (13.0) |
| Other | 33 (23.9) |
| Depression | 32 (23.2) |
| Comorbidity score, median (IQR) | 4 (26) |
| Respiratory rate at admission, median (IQR) | 20 (1924) |
| Oxygen saturation index at admission, median (IQR) | 346.4 (287.5460) |
| Patient NRS, median (IQR) | |
| At admission | 9 (710) |
| At discharge | 2 (14) |
| Discharged on home oxygen | 45 (32.6) |
| Respiratory rate at discharge, median (IQR) | 20 (1820) |
| Oxygen saturation index at discharge, median (IQR) | 475 (350485) |
Agreement Between Patients' Self‐Assessment and Providers' Assessment of Dyspnea Severity
Not all patients had complete data points, and more nurses were interviewed than physicians. Overall, 96 patient‐physician and 138 patient‐nurse pairs participated in the study. A total of 336 patient‐nurse rating dyad assessments of dyspnea and 124 patient‐physician rating dyads assessments were collected (Figure 1). The mean difference between patient and physicians and patient and nurses assessments of dyspnea was 1.23 (IQR, 3 to 0) and 0.21 (IQR, 2 to 2) respectively (a negative score means underestimation by the provider, a positive score means overestimation).
The unadjusted agreement on the severity of dyspnea was 36.3% for the patient‐physician dyads and 44.1% for the patient‐nurse dyads. Physicians underestimated their patients' dyspnea 37.9% of the time and overestimated it 25.8% of the time; nurses underestimated it 43.5% of the time and overestimated it in 12.4% of the study patients (Table 2). In 28.2% of the time, physicians were discordant more than 4 points of the patient assessment. Bland‐Altman plots show that there is greater variation in differences of dyspnea assessments with increase in shortness of breath scores (Figure 2). Nurses underestimated more when the dyspnea score was on the lower end. Physicians also tended to estimate either lower or higher when compared to patients when the dyspnea scores were <2 (Figure 2A,B).
| Underestimation | Concordance | Overestimation | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 3 | 2 | %* | 0 | 1 | % | 2 | 3 | % | |
| |||||||||
| Patient‐nurse dyads | 110 | 48 | 43.5 | 82 | 78 | 44.1 | 17 | 28 | 12.4 |
| Patient‐physician dyads | 33 | 14 | 37.9 | 21 | 24 | 36.3 | 12 | 20 | 25.8 |
The weighted kappa coefficient for agreement was 0.11 (95% confidence interval [CI]: 0.01 to 0.21) for patient‐physician assessment, 0.18 (95% CI: 0.12 to 0.24) for patient‐nurse, and 0.09 (0.02 to 0.20) for physician‐nurse indicating poor agreement. In a sensitivity analysis in which we used a higher threshold for defining discordance (difference of more than 2 points), the kappa coefficient increased to 0.21 (95% CI: 0.06 to 0.36) for patient‐physician assessments, to 0.24 (95% CI: 0.15 to 0.33) for patient‐nurse, and to 0.24 (95% CI: 0.09 to 0.39) for nurse‐physician assessments.
Predictors of Discordance and Underestimation of Dyspnea Severity Assessment
Principal diagnosis was the only factor associated with the physicians' discordant assessment of patients' dyspnea. Patients with admission diagnoses other than HF, COPD, or pneumonia (eg, pulmonary embolism) were more likely to have an accurate assessment of their dyspnea by providers (Table 3). Similar results were obtained in the sensitivity analysis by using a higher cutoff for defining discordance and when assessing predictors for underestimation (results not shown).
| Modeling Probability of Discordance | ||
|---|---|---|
| Physician‐Patient Dyads, OR (95% CI), N = 124 | Nurse‐Patient Dyads, OR (95% CI), N = 363 | |
| ||
| Univariate Analysis | ||
| Body mass index | 1.00 (0.991.01) | 1.00 (0.991.00) |
| Comorbidity score | 1.01 (0.981.05) | 0.99 (0.961.01) |
| Respiratory rate at admission | 1.00 (0.991.02) | 0.99 (0.981.00) |
| Oxygen saturation at admission | 1.00 (1.001.00) | 1.00 (1.001.00) |
| Age (binary) | ||
| 65 years | Referent | Referent |
| >65 years | 1.21 (0.572.55) | 0.96 (0.571.64) |
| Gender | ||
| Female | Referent | Referent |
| Male | 1.10 (0.522.32) | 0.81 (0.481.37) |
| Race | ||
| White | Referent | Referent |
| Nonwhite | 1.02 (0.442.37) | 1.06 (0.581.95) |
| Obese (BMI >30) | 1.43 (0.663.11) | 0.76 (0.441.30) |
| Smoker | 1.36 (0.613.05) | 1.04 (0.591.85) |
| Admitting diagnosis | ||
| Heart failure | Referent | Referent |
| COPD/asthma | 0.68 (0.251.83) | 1.91 (0.983.73)* |
| Pneumonia | 0.38 (0.101.40) | 1.07 (0.462.45) |
| Other | 0.30 (0.110.82)* | 1.54 (0.763.11) |
| Depression | 1.21 (0.572.55) | 1.01 (0.541.86) |
| Multivariable analysis | ||
| Admitting diagnosis | ||
| Congestive heart failure | Referent | Referent |
| COPD | 0.68 (0.251.83) | 1.91 (0.983.73)* |
| Pneumonia | 0.38 (0.101.40) | 1.07 (0.462.45) |
| Other | 0.30 (0.110.82)* | 1.54 (0.763.11) |
In the multivariable analysis that assessed patient‐nurse dyads, the diagnosis of COPD was associated with a marginally significant likelihood of discordance (OR: 1.91; 95% CI: 0.98 to 3.73) (Table 3). Similarly, multivariable analysis identified principal diagnosis to be the only predictor of underestimation, and COPD diagnosis was associated with increased odds of dyspnea underestimation by nurses. When we used a higher cutoff to define discordance, the principal diagnosis of COPD (OR: 3.43, 95% CI: 1.76 to 6.69) was associated with an increased risk of discordance, and smoking (OR: 0.54, 95% CI: 0.29 to 0.99) was associated with a decreased risk of discordance. Overall, 45 patients (32.6%) were discharged on oxygen. The odds of discrepancy (under‐ or overestimation) in dyspnea scores between patient and nurse were 1.7 times higher compared to patients who were not discharged on oxygen, but this association did not reach statistical significance; the odds of discrepancy between patient and physician were 3.88 (95% CI: 1.07 to 14.13).
Documentation of Dyspnea
We found that dyspnea was mentioned in the admission notes in 96% of the charts reviewed; physicians used a qualitative rating (mild, moderate, or severe) to indicate the severity of dyspnea in only 16% of cases, and in 53% a descriptor was added (eg, dyspnea with climbing stairs, gradually increased in the prior week). Nurses were more likely than physicians to use qualitative ratings of dyspnea (26% of cases), and they used a more uniform description of the patient's dyspnea (eg, at rest, at rest and on exertion, or on exertion) than physicians. At discharge, 83% of physicians noted in their discharge summary that dyspnea improved compared with admission but did not refer to the patient's baseline level of dyspnea.
DISCUSSION
In this prospective study of 138 patients hospitalized with cardiopulmonary disease, we found that the agreement between patient's experience of dyspnea and providers' assessment was poor, and the discordance was higher for physicians than for nurses. In more than half of the cases, differences between patient and healthcare providers' assessment of dyspnea were present. One‐third of the time, both physicians and nurses underestimated patients' reported levels of dyspnea. Admitting diagnosis was the only patient factor predicting lack of agreement, and patients with COPD were more likely to have their dyspnea underestimated by nurses. Healthcare providers predominantly documented the presence or absence of dyspnea and rarely used a more nuanced scale.
Discrepancies between patient and provider assessments for pain, depression, and overall health have been reported.[8, 18, 19, 20, 21] One explanation is that patients and healthcare providers measure different factors despite using the same terminology. Furthermore, patient's assessment may be confounded by other symptoms such as anxiety, fatigue, or pain. Physicians and nurses may underevaluate and underestimate the level of breathlessness; however, from the physician perspective, dyspnea is only 1 data point, and providers rely on other measures, such as oxygen saturation, heart rate, respiratory rate, evidence of increasing breathing effort, and arterial blood gas to drive decision making. In a recent study that evaluated the attitudes and beliefs of hospitalists regarding the assessment and management of dyspnea, we found that most hospitalists indicated that awareness of dyspnea severity influences their decision for treatment, diagnostic testing, and timing of the discharge. Moreover, whereas less than half of the respondents reported experience with standardized assessment of dyspnea severity, most stated that such data would be very useful in their practice.[22]
What is the clinical significance of having discordance between patients self‐assessment and providers impression of the patient's severity of dyspnea? First, inaccurate assessment of dyspnea by providers can lead to inadequate treatment and workup. For example, a physician who underestimates the severity of dyspnea may fail to recognize when a complication of the underlying disease develops or may underutilize symptomatic methods for relief of dyspnea. In contrast, a physician who overestimates dyspnea may continue with aggressive treatment when this is not necessary. Second, lack of awareness of dyspnea severity experienced by the patient may result in premature discharge and patient's frustration with the provider, as was shown in several studies evaluating physician‐patient agreement for pain perception.[21, 23] We found that discrepancy between patients and healthcare providers was more pronounced for patients with COPD. In another study using the same patient cohort, we reported that compared to patients with congestive heart failure, those with COPD had more residual dyspnea at discharge; 1 in 4 patients was discharged with a dyspnea score of 5 of greater, and almost half reported symptoms above their baseline.[24] The results from the current study may explain in part why patients with COPD are discharged with higher levels of dyspnea and should alert healthcare providers on the importance of patient‐reported breathlessness. Third, the high level of discordance between healthcare providers and patients may explain the undertreatment of dyspnea in patients with advanced disease. This is supported by our findings that the discordance between patients and physicians was higher if the patient was discharged on oxygen.
One key role of the provider during a clinical encounter is to elicit the patient's symptoms and achieve a shared understanding of what the patient is experiencing. From the patient's perspective, their self‐assessment of dyspnea is more important that the physician's assessment. Fortunately, there is a growing recognition and emphasis on using outcomes that matter to patients, such as dyspnea, to inform judgment about patient care and for clinical research. Numerical measures for assessment of dyspnea exist, are easy to use, and are sensitive to change in patients' dyspnea.[6, 25, 26, 27] Still, it is not standard practice for healthcare providers to ask patients to provide a rating of their dyspnea. When we examined the documentation of dyspnea in the medical record, we found that the description was vague, and providers did not use a standardized validated assessment. Although the dyspnea score decreased during hospitalization, the respiratory rate did not significantly change, indicating that this objective measure may not be reliable in patient assessment. The providers' knowledge of the intensity of the symptom expressed by patients will enable them to track improvement in symptoms over time or in response to therapy. In addition, in this era of multiple handoffs within a hospitalization or from primary care to the hospital, a more uniform assessment could allow providers to follow the severity and time course of dyspnea. The low level of agreement we found between patients and the providers lends support to recommendations regarding a structured dyspnea assessment into routine hospital practice.
Study Strengths and Limitations
This study has several strengths. This is 1 of the very few studies to report on the level of agreement between patients' and providers' assessments of dyspnea. We used a validated, simple dyspnea scale that provides consistency in rating.[28, 29] We enrolled patients with a broad set of diagnoses and complaints, which increases the generalizability of our results, and we surveyed both physicians and nurses. Last, our findings were robust across different cutoff points utilized to characterize discordance and across 3 frequent diagnoses.
The study has several limitations. First, we included only English‐speaking patients, and the results cannot be generalized to patients from other cultures who do not speak English. Second, this is a single‐center study, and practices may be different in other centers; for example, some hospitals may have already implemented a dyspnea assessment tool. Third, we did not collect information on the physician and nurse characteristics such as years in practice. However, a recent study that describes the agreement of breathlessness assessment between nurses, physicians, and mechanically ventilated patients found that underestimation of breathlessness by providers was not associated with professional competencies, previous patient care, or years of working in an intensive care unit.[9] In addition, a systematic review found that length of professional experience is often unrelated to performance measures and outcomes.[30] Finally, although we asked for physicians and nurses assessment close to their visit to the patient, assessment was done from memory, not at the bedside observing the patient.
CONCLUSION
We found that the extent of agreement between a structured patient self‐assessment of dyspnea and healthcare providers' assessment was low. Future studies should prospectively test whether routine assessment of dyspnea results in better acute and long‐term patient outcomes.
Acknowledgements
The authors acknowledge Ms. Anu Joshi for her help with formatting the manuscript and assisting with table preparations. The authors also acknowledge Ms. Katherine Dempsey, Jahnavi Sagi, Sashi Ariyaratne, and Mr. Pradeep Kumbaham for their help with collecting the data.
Disclosures: M.S.S. is the guarantor for this article and had full access to all of the data in the study, and takes responsibility for the integrity and accuracy of the data analysis. M.S.S., P.K.L., E.N., and M.B.R. conceived of the study. M.S.S. and B.M. acquired the data. M.S.S., A.P., P.S.P., R.J.G., and P.K.L. analyzed and interpreted the data. M.S.S. drafted the manuscript. All authors critically reviewed the manuscript for intellectual content. M.S.S. is supported by grant 1K01HL114631‐01A1 from the National Heart, Lung, and Blood Institute of the National Institutes of Health and by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through grant UL1RR025752. The authors report no conflicts of interest.
- , , , et al. An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea. Am J Respir Crit Care Med. 2012;185(4):435–452.
- , , . National Hospital Ambulatory Medical Care Survey: 2007 emergency department summary. Natl Health Stat Report. 2010;(26):1–31.
- , , , et al. The body‐mass index, airflow obstruction, dyspnea, and exercise capacity index in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(10):1005–1012.
- , , , . Dyspnea is a better predictor of 5‐year survival than airway obstruction in patients with COPD. Chest. 2002;121(5):1434–1440.
- , , . A multidimensional grading system (BODE index) as predictor of hospitalization for COPD. Chest. 2005;128(6):3810–3816.
- , , , et al. American College of Chest Physicians consensus statement on the management of dyspnea in patients with advanced lung or heart disease. Chest. 2010;137(3):674–691.
- , , , et al. Managing dyspnea in patients with advanced chronic obstructive pulmonary disease: a Canadian Thoracic Society clinical practice guideline. Can Respir J. 2011;18(2):69–78.
- , , . Physician vs patient assessment of dyspnea during acute decompensated heart failure. Congest Heart Fail. 2010;16(2):60–64.
- , , , , , . Underestimation of Patient Breathlessness by Nurses and Physicians during a Spontaneous Breathing Trial. Am J Respir Crit Care Med. 2015;192(12):1440–1448.
- , , , , , . The influence of patient‐practitioner agreement on outcome of care. Am J Public Health. 1981;71(2):127–131.
- , , , , , . Listening to parents: The role of symptom perception in pediatric palliative home care. Palliat Support Care. 2016;14(1):13–19.
- , . Validity of the numeric rating scale as a measure of dyspnea. Am J Crit Care. 1998;7(3):200–204.
- , , , , , . Dyspnea scales in the assessment of illiterate patients with chronic obstructive pulmonary disease. Am J Med Sci. 2000;320(4):240–243.
- , , , , . Measuring the dyspnea of decompensated heart failure with a visual analog scale: how much improvement is meaningful? Congest Heart Fail. 2004;10(4):188–191.
- , , , , , . Clinically meaningful changes in quantitative measures of asthma severity. Acad Emerg Med. 2000;7(4):327–334.
- , , , , . A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–759.
- . BlandAltmanLeh: plots (slightly extended) Bland‐Altman plots. Available at: https://cran.r‐project.org/web/packages/BlandAltmanLeh/index.html. Published December 23, 2015. Accessed March 10, 2016.
- , , , , . Correlation of patient and caregiver ratings of cancer pain. J Pain Symptom Manage. 1991;6(2):53–57.
- , , , et al. Depression symptomatology and diagnosis: discordance between patients and physicians in primary care settings. BMC Fam Pract. 2008;9:1.
- , , , , , . Patient‐physician discordance in assessments of global disease severity in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010;62(6):857–864.
- , , , et al. The influence of discordance in pain assessment on the functional status of patients with chronic nonmalignant pain. Am J Med Sci. 2006;332(1):18–23.
- , , , et al. Hospitalist attitudes toward the assessment and management of dyspnea in patients with acute cardiopulmonary diseases. J Hosp Med. 2015;10(11):724–730.
- , , , , . Brief report: patient‐physician agreement as a predictor of outcomes in patients with back pain. J Gen Intern Med. 2005;20(10):935–937.
- , , , , , . The trajectory of dyspnea in hospitalized patients [published online November 24, 2015]. J Pain Symptom Manage. doi: 10.1016/j.jpainsymman.2015.11.005.
- , , , , . Measurement of breathlessness in advanced disease: a systematic review. Respir Med. 2007;101(3):399–410.
- . Review of dyspnoea quantification in the emergency department: is a rating scale for breathlessness suitable for use as an admission prediction tool? Emerg Med Australas. 2007;19(5):394–404.
- , , , . Validation of a verbal dyspnoea rating scale in the emergency department. Emerg Med Australas. 2008;20(6):475–481.
- , , . Measurement of dyspnea: word labeled visual analog scale vs. verbal ordinal scale. Respir Physiol Neurobiol. 2003;134(2):77–83.
- , , , , , . Verbal numerical scales are as reliable and sensitive as visual analog scales for rating dyspnea in young and older subjects. Respir Physiol Neurobiol. 2007;157(2–3):360–365.
- , , . Systematic review: the relationship between clinical experience and quality of health care. Ann Intern Med. 2005;142(4):260–273.
Breathlessness, or dyspnea, is defined as a subjective experience of breathing discomfort that is comprised of qualitatively distinct sensations that vary in intensity.[1] Dyspnea is a leading reason for patients presenting for emergency care,[2] and it is an important predictor for hospitalization and mortality in patients with cardiopulmonary disease.[3, 4, 5]
Several professional societies' guidelines recommend that patients should be asked to quantify the intensity of their breathlessness using a standardized scale, and that these ratings should be documented in medical records to guide dyspnea awareness and management.[1, 6, 7] During the evaluation and treatment of patients with acute cardiopulmonary conditions, the clinician estimates the severity of the illness and response to therapy based on multiple objective measures as well as the patient's perception of dyspnea. A patient‐centered care approach depends upon the physicians having a shared understanding of what the patient is experiencing. Without this appreciation, the healthcare provider cannot make appropriate treatment decisions to ensure alleviation of presenting symptoms. Understanding the severity of patients' dyspnea is critical to avoid under‐ or overtreatment of patients with acute cardiopulmonary conditions, but only a few studies have compared patient and provider perceptions of dyspnea intensity.[8, 9] Discordance between physician's impression of severity of dyspnea and patient's perception may result in suboptimal management and patient dissatisfaction with care. Furthermore, several studies have shown that, when physicians and patients agree with the assessment of well‐being, treatment adherence and outcomes improve.[10, 11]
Therefore, we evaluated the extent and directionality of agreement between patients' perception and healthcare providers' impression of dyspnea and explored which factors contribute to discordance. Additionally, we examined how healthcare providers document dyspnea severity.
METHODS
Study Setting and Population
The study was conducted between June 2012 and August 2012 at Baystate Medical Center (BMC), a 740‐bed tertiary care hospital in western Massachusetts. In 2012, the BMC hospitalist group had 48 attending physicians, of whom 47% were female, 48% had 0 to 3 years of attending experience, and 16% had 10 years of experience.
We enrolled consecutive admissions of English‐speaking adult patients, with a working diagnosis of heart failure (HF), chronic obstructive pulmonary disease (COPD), asthma, pneumonia, or a generic diagnosis of shortness of breath. Because we surveyed only hospitalists, we did not include patients admitted to an intensive care unit.
All participants gave informed consent to be part of the study. The research protocol was approved by the Baystate Health Institutional Review Board, Springfield, Massachusetts.
Dyspnea Assessment
Dyspnea intensity was assessed on an 11‐point (010) numerical rating scale (NRS).[12, 13] A trained research assistant interviewed patients on day 1 and 2 after admission, and on the day of discharge between 8 am and 12 pm on weekdays. The patient was asked: On a scale from 0 to 10, how bad is your shortness of breath at rest now, with 0 being no shortness of breath and 10 the worst shortness of breath you could ever imagine? The hospitalist or the senior resident and day‐shift nurse taking care of the patients were asked by the research assistant to rate the patient's dyspnea using the same scoring instrument shortly after they saw the patient. The physicians and nurses based their determination of dyspnea on their usual interview/examination of the patient. The patient, physician and the nurse were not aware of each other's rating. The research assistant scheduled the interviews to minimize the time intervals between the patient assessment and provider's rating. For this reason, the number of assessments per patient varied. Nurses were more readily available than physicians, which resulted in a larger number of patient‐nurse response pairs than patient‐physician pairs. All assessments were done in the morning, between 9 am and 12 pm, with a range of 3 hours between provider's assessment of the patient and the interview.
Dyspnea Agreement
Agreement was defined as a score within 1 between patient and healthcare provider; differences of 2 points were considered over‐ or underestimations. The decision to use this cutoff was based on prior studies, which found that a difference in the range of 1.6 to 2.2 cm was meaningful for the patient when assessment was done on the visual analog scale.[8, 14, 15] We also evaluated the direction of discordance. If the patient's rating of dyspnea severity was higher than the provider's rating, we defined this as underestimation by the provider; in the instance where a provider's score of dyspnea severity was higher than the patient's score, we defined this as overestimation. In a sensitivity analysis, agreement was defined as a score within 2 between patient and healthcare provider, and any difference 3 was considered disagreement.
Other Variables
We obtained information from the medical records about patient demographics, body mass index (BMI), smoking status, and vital signs. We calculated the oxygen saturation index as the ratio between the oxygen saturation and the fractional inspired oxygen (SpO2/FIO2). Comorbidities were assessed based on the International Classification od Diseases, Ninth Revision, Clinical Modification codes from the hospital financial decision support system. We calculated an overall combined comorbidity score based on the method described by Gagne, which is based on elements from the Charlson Comorbidity Index and from the Elixhauser comorbidities.[16]
Charts of the patients included in the study were retrospectively reviewed for physicians' and nurses' documentation of dyspnea at admission and at discharge. We recorded if dyspnea was mentioned and how it was assessed: whether it was described as present/absent; graded as mild, moderate, or severe; used a quantitative scale (010); used descriptors (eg, dyspnea when climbing stairs); and whether it was defined as improved or worsened without other qualifiers.
Statistical Analysis
Descriptive statistics of dyspnea scores, patient characteristics, comorbidities and vital signs were calculated and presented as medians with interquartile range (IQR) for continuous variables, and counts with percentages for categorical factors. Every patient‐provider concurrent scoring was included in the analysis as 1 dyad, which resulted in patients being included multiple times in the analysis. Patient‐physician and patient‐nurse dyads of dyspnea assessment were examined separately. Analyses included all dyads that were within same assessment period (same day and same time window).
The relationship between patient self‐perceived dyspnea severity and provider rating of was assessed in several ways. First, a weighted kappa coefficient was used as a measure of agreement between patient and nurse or physician scores. A weighted kappa analysis was chosen because it penalizes disagreements that are further apart from each other.
Second, we defined an indicator of discordance and constructed multivariable generalized estimating equation models that account for clustering of multiple dyads per patient, to assess the relationship of patient characteristics with discordance. Finally, we developed additional models to predict underestimation when compared to agreement or overestimation of dyspnea by the healthcare provider relative to the patient. Using the same definitions for agreement, we also compared the dyspnea assessment estimation between physicians and nurses.
We present the differences in dyspnea assessment between patient and healthcare provider and between nurses and physicians by Bland‐Altman plots.
All analyses were performed using SAS (version 9.3; SAS institute, Inc., Cary, NC), Stata (Stata statistical software release 13; StataCorp, College Station, TX), and RStudio version 0.99.892 (Bland‐Altman plots, R package version 0.3.1; The R Foundation for Statistical Computing, Vienna, Austria).[8, 9, 17]
RESULTS
Patient Characteristics
Among the 219 patients who met the screening criteria, 81 were not enrolled (Figure 1). Data from 138 patients, with both patient information and provider data on dyspnea assessment, were included. The median age of the patients was 72 years (IQR, 5880 years), 56.5% were women, 75.4% were white, and 28.3% were current smokers. Approximately 30% had a diagnosis of HF, 30% of COPD, and 13.0% of pneumonia. The median comorbidity score was 4 (IQR, 26), and 37.0% of the patients had a BMI 30. At admission, the median oxygen saturation index was 346 (IQR, 287.5460) indicating mild to moderate levels of hypoxia. (Table 1).
| Value | |
|---|---|
| |
| Age, median (IQR), y | 72 (5880) |
| Gender | |
| Female | 78 (56.5) |
| Male | 60 (43.5) |
| Race | |
| White | 104 (75.4) |
| Black | 16 (11.6) |
| Hispanic | 17 (12.3) |
| Other | 1 (0.7) |
| Body mass index, median (IQR) | 28 (23.334.6) |
| Obese (BMI 30) | 51 (37.0) |
| Smoker, current | 39 (28.3) |
| Admitting diagnosis | |
| Heart failure | 46 (33.3) |
| COPD/asthma | 41 (29.7) |
| Pneumonia | 18 (13.0) |
| Other | 33 (23.9) |
| Depression | 32 (23.2) |
| Comorbidity score, median (IQR) | 4 (26) |
| Respiratory rate at admission, median (IQR) | 20 (1924) |
| Oxygen saturation index at admission, median (IQR) | 346.4 (287.5460) |
| Patient NRS, median (IQR) | |
| At admission | 9 (710) |
| At discharge | 2 (14) |
| Discharged on home oxygen | 45 (32.6) |
| Respiratory rate at discharge, median (IQR) | 20 (1820) |
| Oxygen saturation index at discharge, median (IQR) | 475 (350485) |
Agreement Between Patients' Self‐Assessment and Providers' Assessment of Dyspnea Severity
Not all patients had complete data points, and more nurses were interviewed than physicians. Overall, 96 patient‐physician and 138 patient‐nurse pairs participated in the study. A total of 336 patient‐nurse rating dyad assessments of dyspnea and 124 patient‐physician rating dyads assessments were collected (Figure 1). The mean difference between patient and physicians and patient and nurses assessments of dyspnea was 1.23 (IQR, 3 to 0) and 0.21 (IQR, 2 to 2) respectively (a negative score means underestimation by the provider, a positive score means overestimation).
The unadjusted agreement on the severity of dyspnea was 36.3% for the patient‐physician dyads and 44.1% for the patient‐nurse dyads. Physicians underestimated their patients' dyspnea 37.9% of the time and overestimated it 25.8% of the time; nurses underestimated it 43.5% of the time and overestimated it in 12.4% of the study patients (Table 2). In 28.2% of the time, physicians were discordant more than 4 points of the patient assessment. Bland‐Altman plots show that there is greater variation in differences of dyspnea assessments with increase in shortness of breath scores (Figure 2). Nurses underestimated more when the dyspnea score was on the lower end. Physicians also tended to estimate either lower or higher when compared to patients when the dyspnea scores were <2 (Figure 2A,B).
| Underestimation | Concordance | Overestimation | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 3 | 2 | %* | 0 | 1 | % | 2 | 3 | % | |
| |||||||||
| Patient‐nurse dyads | 110 | 48 | 43.5 | 82 | 78 | 44.1 | 17 | 28 | 12.4 |
| Patient‐physician dyads | 33 | 14 | 37.9 | 21 | 24 | 36.3 | 12 | 20 | 25.8 |
The weighted kappa coefficient for agreement was 0.11 (95% confidence interval [CI]: 0.01 to 0.21) for patient‐physician assessment, 0.18 (95% CI: 0.12 to 0.24) for patient‐nurse, and 0.09 (0.02 to 0.20) for physician‐nurse indicating poor agreement. In a sensitivity analysis in which we used a higher threshold for defining discordance (difference of more than 2 points), the kappa coefficient increased to 0.21 (95% CI: 0.06 to 0.36) for patient‐physician assessments, to 0.24 (95% CI: 0.15 to 0.33) for patient‐nurse, and to 0.24 (95% CI: 0.09 to 0.39) for nurse‐physician assessments.
Predictors of Discordance and Underestimation of Dyspnea Severity Assessment
Principal diagnosis was the only factor associated with the physicians' discordant assessment of patients' dyspnea. Patients with admission diagnoses other than HF, COPD, or pneumonia (eg, pulmonary embolism) were more likely to have an accurate assessment of their dyspnea by providers (Table 3). Similar results were obtained in the sensitivity analysis by using a higher cutoff for defining discordance and when assessing predictors for underestimation (results not shown).
| Modeling Probability of Discordance | ||
|---|---|---|
| Physician‐Patient Dyads, OR (95% CI), N = 124 | Nurse‐Patient Dyads, OR (95% CI), N = 363 | |
| ||
| Univariate Analysis | ||
| Body mass index | 1.00 (0.991.01) | 1.00 (0.991.00) |
| Comorbidity score | 1.01 (0.981.05) | 0.99 (0.961.01) |
| Respiratory rate at admission | 1.00 (0.991.02) | 0.99 (0.981.00) |
| Oxygen saturation at admission | 1.00 (1.001.00) | 1.00 (1.001.00) |
| Age (binary) | ||
| 65 years | Referent | Referent |
| >65 years | 1.21 (0.572.55) | 0.96 (0.571.64) |
| Gender | ||
| Female | Referent | Referent |
| Male | 1.10 (0.522.32) | 0.81 (0.481.37) |
| Race | ||
| White | Referent | Referent |
| Nonwhite | 1.02 (0.442.37) | 1.06 (0.581.95) |
| Obese (BMI >30) | 1.43 (0.663.11) | 0.76 (0.441.30) |
| Smoker | 1.36 (0.613.05) | 1.04 (0.591.85) |
| Admitting diagnosis | ||
| Heart failure | Referent | Referent |
| COPD/asthma | 0.68 (0.251.83) | 1.91 (0.983.73)* |
| Pneumonia | 0.38 (0.101.40) | 1.07 (0.462.45) |
| Other | 0.30 (0.110.82)* | 1.54 (0.763.11) |
| Depression | 1.21 (0.572.55) | 1.01 (0.541.86) |
| Multivariable analysis | ||
| Admitting diagnosis | ||
| Congestive heart failure | Referent | Referent |
| COPD | 0.68 (0.251.83) | 1.91 (0.983.73)* |
| Pneumonia | 0.38 (0.101.40) | 1.07 (0.462.45) |
| Other | 0.30 (0.110.82)* | 1.54 (0.763.11) |
In the multivariable analysis that assessed patient‐nurse dyads, the diagnosis of COPD was associated with a marginally significant likelihood of discordance (OR: 1.91; 95% CI: 0.98 to 3.73) (Table 3). Similarly, multivariable analysis identified principal diagnosis to be the only predictor of underestimation, and COPD diagnosis was associated with increased odds of dyspnea underestimation by nurses. When we used a higher cutoff to define discordance, the principal diagnosis of COPD (OR: 3.43, 95% CI: 1.76 to 6.69) was associated with an increased risk of discordance, and smoking (OR: 0.54, 95% CI: 0.29 to 0.99) was associated with a decreased risk of discordance. Overall, 45 patients (32.6%) were discharged on oxygen. The odds of discrepancy (under‐ or overestimation) in dyspnea scores between patient and nurse were 1.7 times higher compared to patients who were not discharged on oxygen, but this association did not reach statistical significance; the odds of discrepancy between patient and physician were 3.88 (95% CI: 1.07 to 14.13).
Documentation of Dyspnea
We found that dyspnea was mentioned in the admission notes in 96% of the charts reviewed; physicians used a qualitative rating (mild, moderate, or severe) to indicate the severity of dyspnea in only 16% of cases, and in 53% a descriptor was added (eg, dyspnea with climbing stairs, gradually increased in the prior week). Nurses were more likely than physicians to use qualitative ratings of dyspnea (26% of cases), and they used a more uniform description of the patient's dyspnea (eg, at rest, at rest and on exertion, or on exertion) than physicians. At discharge, 83% of physicians noted in their discharge summary that dyspnea improved compared with admission but did not refer to the patient's baseline level of dyspnea.
DISCUSSION
In this prospective study of 138 patients hospitalized with cardiopulmonary disease, we found that the agreement between patient's experience of dyspnea and providers' assessment was poor, and the discordance was higher for physicians than for nurses. In more than half of the cases, differences between patient and healthcare providers' assessment of dyspnea were present. One‐third of the time, both physicians and nurses underestimated patients' reported levels of dyspnea. Admitting diagnosis was the only patient factor predicting lack of agreement, and patients with COPD were more likely to have their dyspnea underestimated by nurses. Healthcare providers predominantly documented the presence or absence of dyspnea and rarely used a more nuanced scale.
Discrepancies between patient and provider assessments for pain, depression, and overall health have been reported.[8, 18, 19, 20, 21] One explanation is that patients and healthcare providers measure different factors despite using the same terminology. Furthermore, patient's assessment may be confounded by other symptoms such as anxiety, fatigue, or pain. Physicians and nurses may underevaluate and underestimate the level of breathlessness; however, from the physician perspective, dyspnea is only 1 data point, and providers rely on other measures, such as oxygen saturation, heart rate, respiratory rate, evidence of increasing breathing effort, and arterial blood gas to drive decision making. In a recent study that evaluated the attitudes and beliefs of hospitalists regarding the assessment and management of dyspnea, we found that most hospitalists indicated that awareness of dyspnea severity influences their decision for treatment, diagnostic testing, and timing of the discharge. Moreover, whereas less than half of the respondents reported experience with standardized assessment of dyspnea severity, most stated that such data would be very useful in their practice.[22]
What is the clinical significance of having discordance between patients self‐assessment and providers impression of the patient's severity of dyspnea? First, inaccurate assessment of dyspnea by providers can lead to inadequate treatment and workup. For example, a physician who underestimates the severity of dyspnea may fail to recognize when a complication of the underlying disease develops or may underutilize symptomatic methods for relief of dyspnea. In contrast, a physician who overestimates dyspnea may continue with aggressive treatment when this is not necessary. Second, lack of awareness of dyspnea severity experienced by the patient may result in premature discharge and patient's frustration with the provider, as was shown in several studies evaluating physician‐patient agreement for pain perception.[21, 23] We found that discrepancy between patients and healthcare providers was more pronounced for patients with COPD. In another study using the same patient cohort, we reported that compared to patients with congestive heart failure, those with COPD had more residual dyspnea at discharge; 1 in 4 patients was discharged with a dyspnea score of 5 of greater, and almost half reported symptoms above their baseline.[24] The results from the current study may explain in part why patients with COPD are discharged with higher levels of dyspnea and should alert healthcare providers on the importance of patient‐reported breathlessness. Third, the high level of discordance between healthcare providers and patients may explain the undertreatment of dyspnea in patients with advanced disease. This is supported by our findings that the discordance between patients and physicians was higher if the patient was discharged on oxygen.
One key role of the provider during a clinical encounter is to elicit the patient's symptoms and achieve a shared understanding of what the patient is experiencing. From the patient's perspective, their self‐assessment of dyspnea is more important that the physician's assessment. Fortunately, there is a growing recognition and emphasis on using outcomes that matter to patients, such as dyspnea, to inform judgment about patient care and for clinical research. Numerical measures for assessment of dyspnea exist, are easy to use, and are sensitive to change in patients' dyspnea.[6, 25, 26, 27] Still, it is not standard practice for healthcare providers to ask patients to provide a rating of their dyspnea. When we examined the documentation of dyspnea in the medical record, we found that the description was vague, and providers did not use a standardized validated assessment. Although the dyspnea score decreased during hospitalization, the respiratory rate did not significantly change, indicating that this objective measure may not be reliable in patient assessment. The providers' knowledge of the intensity of the symptom expressed by patients will enable them to track improvement in symptoms over time or in response to therapy. In addition, in this era of multiple handoffs within a hospitalization or from primary care to the hospital, a more uniform assessment could allow providers to follow the severity and time course of dyspnea. The low level of agreement we found between patients and the providers lends support to recommendations regarding a structured dyspnea assessment into routine hospital practice.
Study Strengths and Limitations
This study has several strengths. This is 1 of the very few studies to report on the level of agreement between patients' and providers' assessments of dyspnea. We used a validated, simple dyspnea scale that provides consistency in rating.[28, 29] We enrolled patients with a broad set of diagnoses and complaints, which increases the generalizability of our results, and we surveyed both physicians and nurses. Last, our findings were robust across different cutoff points utilized to characterize discordance and across 3 frequent diagnoses.
The study has several limitations. First, we included only English‐speaking patients, and the results cannot be generalized to patients from other cultures who do not speak English. Second, this is a single‐center study, and practices may be different in other centers; for example, some hospitals may have already implemented a dyspnea assessment tool. Third, we did not collect information on the physician and nurse characteristics such as years in practice. However, a recent study that describes the agreement of breathlessness assessment between nurses, physicians, and mechanically ventilated patients found that underestimation of breathlessness by providers was not associated with professional competencies, previous patient care, or years of working in an intensive care unit.[9] In addition, a systematic review found that length of professional experience is often unrelated to performance measures and outcomes.[30] Finally, although we asked for physicians and nurses assessment close to their visit to the patient, assessment was done from memory, not at the bedside observing the patient.
CONCLUSION
We found that the extent of agreement between a structured patient self‐assessment of dyspnea and healthcare providers' assessment was low. Future studies should prospectively test whether routine assessment of dyspnea results in better acute and long‐term patient outcomes.
Acknowledgements
The authors acknowledge Ms. Anu Joshi for her help with formatting the manuscript and assisting with table preparations. The authors also acknowledge Ms. Katherine Dempsey, Jahnavi Sagi, Sashi Ariyaratne, and Mr. Pradeep Kumbaham for their help with collecting the data.
Disclosures: M.S.S. is the guarantor for this article and had full access to all of the data in the study, and takes responsibility for the integrity and accuracy of the data analysis. M.S.S., P.K.L., E.N., and M.B.R. conceived of the study. M.S.S. and B.M. acquired the data. M.S.S., A.P., P.S.P., R.J.G., and P.K.L. analyzed and interpreted the data. M.S.S. drafted the manuscript. All authors critically reviewed the manuscript for intellectual content. M.S.S. is supported by grant 1K01HL114631‐01A1 from the National Heart, Lung, and Blood Institute of the National Institutes of Health and by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through grant UL1RR025752. The authors report no conflicts of interest.
Breathlessness, or dyspnea, is defined as a subjective experience of breathing discomfort that is comprised of qualitatively distinct sensations that vary in intensity.[1] Dyspnea is a leading reason for patients presenting for emergency care,[2] and it is an important predictor for hospitalization and mortality in patients with cardiopulmonary disease.[3, 4, 5]
Several professional societies' guidelines recommend that patients should be asked to quantify the intensity of their breathlessness using a standardized scale, and that these ratings should be documented in medical records to guide dyspnea awareness and management.[1, 6, 7] During the evaluation and treatment of patients with acute cardiopulmonary conditions, the clinician estimates the severity of the illness and response to therapy based on multiple objective measures as well as the patient's perception of dyspnea. A patient‐centered care approach depends upon the physicians having a shared understanding of what the patient is experiencing. Without this appreciation, the healthcare provider cannot make appropriate treatment decisions to ensure alleviation of presenting symptoms. Understanding the severity of patients' dyspnea is critical to avoid under‐ or overtreatment of patients with acute cardiopulmonary conditions, but only a few studies have compared patient and provider perceptions of dyspnea intensity.[8, 9] Discordance between physician's impression of severity of dyspnea and patient's perception may result in suboptimal management and patient dissatisfaction with care. Furthermore, several studies have shown that, when physicians and patients agree with the assessment of well‐being, treatment adherence and outcomes improve.[10, 11]
Therefore, we evaluated the extent and directionality of agreement between patients' perception and healthcare providers' impression of dyspnea and explored which factors contribute to discordance. Additionally, we examined how healthcare providers document dyspnea severity.
METHODS
Study Setting and Population
The study was conducted between June 2012 and August 2012 at Baystate Medical Center (BMC), a 740‐bed tertiary care hospital in western Massachusetts. In 2012, the BMC hospitalist group had 48 attending physicians, of whom 47% were female, 48% had 0 to 3 years of attending experience, and 16% had 10 years of experience.
We enrolled consecutive admissions of English‐speaking adult patients, with a working diagnosis of heart failure (HF), chronic obstructive pulmonary disease (COPD), asthma, pneumonia, or a generic diagnosis of shortness of breath. Because we surveyed only hospitalists, we did not include patients admitted to an intensive care unit.
All participants gave informed consent to be part of the study. The research protocol was approved by the Baystate Health Institutional Review Board, Springfield, Massachusetts.
Dyspnea Assessment
Dyspnea intensity was assessed on an 11‐point (010) numerical rating scale (NRS).[12, 13] A trained research assistant interviewed patients on day 1 and 2 after admission, and on the day of discharge between 8 am and 12 pm on weekdays. The patient was asked: On a scale from 0 to 10, how bad is your shortness of breath at rest now, with 0 being no shortness of breath and 10 the worst shortness of breath you could ever imagine? The hospitalist or the senior resident and day‐shift nurse taking care of the patients were asked by the research assistant to rate the patient's dyspnea using the same scoring instrument shortly after they saw the patient. The physicians and nurses based their determination of dyspnea on their usual interview/examination of the patient. The patient, physician and the nurse were not aware of each other's rating. The research assistant scheduled the interviews to minimize the time intervals between the patient assessment and provider's rating. For this reason, the number of assessments per patient varied. Nurses were more readily available than physicians, which resulted in a larger number of patient‐nurse response pairs than patient‐physician pairs. All assessments were done in the morning, between 9 am and 12 pm, with a range of 3 hours between provider's assessment of the patient and the interview.
Dyspnea Agreement
Agreement was defined as a score within 1 between patient and healthcare provider; differences of 2 points were considered over‐ or underestimations. The decision to use this cutoff was based on prior studies, which found that a difference in the range of 1.6 to 2.2 cm was meaningful for the patient when assessment was done on the visual analog scale.[8, 14, 15] We also evaluated the direction of discordance. If the patient's rating of dyspnea severity was higher than the provider's rating, we defined this as underestimation by the provider; in the instance where a provider's score of dyspnea severity was higher than the patient's score, we defined this as overestimation. In a sensitivity analysis, agreement was defined as a score within 2 between patient and healthcare provider, and any difference 3 was considered disagreement.
Other Variables
We obtained information from the medical records about patient demographics, body mass index (BMI), smoking status, and vital signs. We calculated the oxygen saturation index as the ratio between the oxygen saturation and the fractional inspired oxygen (SpO2/FIO2). Comorbidities were assessed based on the International Classification od Diseases, Ninth Revision, Clinical Modification codes from the hospital financial decision support system. We calculated an overall combined comorbidity score based on the method described by Gagne, which is based on elements from the Charlson Comorbidity Index and from the Elixhauser comorbidities.[16]
Charts of the patients included in the study were retrospectively reviewed for physicians' and nurses' documentation of dyspnea at admission and at discharge. We recorded if dyspnea was mentioned and how it was assessed: whether it was described as present/absent; graded as mild, moderate, or severe; used a quantitative scale (010); used descriptors (eg, dyspnea when climbing stairs); and whether it was defined as improved or worsened without other qualifiers.
Statistical Analysis
Descriptive statistics of dyspnea scores, patient characteristics, comorbidities and vital signs were calculated and presented as medians with interquartile range (IQR) for continuous variables, and counts with percentages for categorical factors. Every patient‐provider concurrent scoring was included in the analysis as 1 dyad, which resulted in patients being included multiple times in the analysis. Patient‐physician and patient‐nurse dyads of dyspnea assessment were examined separately. Analyses included all dyads that were within same assessment period (same day and same time window).
The relationship between patient self‐perceived dyspnea severity and provider rating of was assessed in several ways. First, a weighted kappa coefficient was used as a measure of agreement between patient and nurse or physician scores. A weighted kappa analysis was chosen because it penalizes disagreements that are further apart from each other.
Second, we defined an indicator of discordance and constructed multivariable generalized estimating equation models that account for clustering of multiple dyads per patient, to assess the relationship of patient characteristics with discordance. Finally, we developed additional models to predict underestimation when compared to agreement or overestimation of dyspnea by the healthcare provider relative to the patient. Using the same definitions for agreement, we also compared the dyspnea assessment estimation between physicians and nurses.
We present the differences in dyspnea assessment between patient and healthcare provider and between nurses and physicians by Bland‐Altman plots.
All analyses were performed using SAS (version 9.3; SAS institute, Inc., Cary, NC), Stata (Stata statistical software release 13; StataCorp, College Station, TX), and RStudio version 0.99.892 (Bland‐Altman plots, R package version 0.3.1; The R Foundation for Statistical Computing, Vienna, Austria).[8, 9, 17]
RESULTS
Patient Characteristics
Among the 219 patients who met the screening criteria, 81 were not enrolled (Figure 1). Data from 138 patients, with both patient information and provider data on dyspnea assessment, were included. The median age of the patients was 72 years (IQR, 5880 years), 56.5% were women, 75.4% were white, and 28.3% were current smokers. Approximately 30% had a diagnosis of HF, 30% of COPD, and 13.0% of pneumonia. The median comorbidity score was 4 (IQR, 26), and 37.0% of the patients had a BMI 30. At admission, the median oxygen saturation index was 346 (IQR, 287.5460) indicating mild to moderate levels of hypoxia. (Table 1).
| Value | |
|---|---|
| |
| Age, median (IQR), y | 72 (5880) |
| Gender | |
| Female | 78 (56.5) |
| Male | 60 (43.5) |
| Race | |
| White | 104 (75.4) |
| Black | 16 (11.6) |
| Hispanic | 17 (12.3) |
| Other | 1 (0.7) |
| Body mass index, median (IQR) | 28 (23.334.6) |
| Obese (BMI 30) | 51 (37.0) |
| Smoker, current | 39 (28.3) |
| Admitting diagnosis | |
| Heart failure | 46 (33.3) |
| COPD/asthma | 41 (29.7) |
| Pneumonia | 18 (13.0) |
| Other | 33 (23.9) |
| Depression | 32 (23.2) |
| Comorbidity score, median (IQR) | 4 (26) |
| Respiratory rate at admission, median (IQR) | 20 (1924) |
| Oxygen saturation index at admission, median (IQR) | 346.4 (287.5460) |
| Patient NRS, median (IQR) | |
| At admission | 9 (710) |
| At discharge | 2 (14) |
| Discharged on home oxygen | 45 (32.6) |
| Respiratory rate at discharge, median (IQR) | 20 (1820) |
| Oxygen saturation index at discharge, median (IQR) | 475 (350485) |
Agreement Between Patients' Self‐Assessment and Providers' Assessment of Dyspnea Severity
Not all patients had complete data points, and more nurses were interviewed than physicians. Overall, 96 patient‐physician and 138 patient‐nurse pairs participated in the study. A total of 336 patient‐nurse rating dyad assessments of dyspnea and 124 patient‐physician rating dyads assessments were collected (Figure 1). The mean difference between patient and physicians and patient and nurses assessments of dyspnea was 1.23 (IQR, 3 to 0) and 0.21 (IQR, 2 to 2) respectively (a negative score means underestimation by the provider, a positive score means overestimation).
The unadjusted agreement on the severity of dyspnea was 36.3% for the patient‐physician dyads and 44.1% for the patient‐nurse dyads. Physicians underestimated their patients' dyspnea 37.9% of the time and overestimated it 25.8% of the time; nurses underestimated it 43.5% of the time and overestimated it in 12.4% of the study patients (Table 2). In 28.2% of the time, physicians were discordant more than 4 points of the patient assessment. Bland‐Altman plots show that there is greater variation in differences of dyspnea assessments with increase in shortness of breath scores (Figure 2). Nurses underestimated more when the dyspnea score was on the lower end. Physicians also tended to estimate either lower or higher when compared to patients when the dyspnea scores were <2 (Figure 2A,B).
| Underestimation | Concordance | Overestimation | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 3 | 2 | %* | 0 | 1 | % | 2 | 3 | % | |
| |||||||||
| Patient‐nurse dyads | 110 | 48 | 43.5 | 82 | 78 | 44.1 | 17 | 28 | 12.4 |
| Patient‐physician dyads | 33 | 14 | 37.9 | 21 | 24 | 36.3 | 12 | 20 | 25.8 |
The weighted kappa coefficient for agreement was 0.11 (95% confidence interval [CI]: 0.01 to 0.21) for patient‐physician assessment, 0.18 (95% CI: 0.12 to 0.24) for patient‐nurse, and 0.09 (0.02 to 0.20) for physician‐nurse indicating poor agreement. In a sensitivity analysis in which we used a higher threshold for defining discordance (difference of more than 2 points), the kappa coefficient increased to 0.21 (95% CI: 0.06 to 0.36) for patient‐physician assessments, to 0.24 (95% CI: 0.15 to 0.33) for patient‐nurse, and to 0.24 (95% CI: 0.09 to 0.39) for nurse‐physician assessments.
Predictors of Discordance and Underestimation of Dyspnea Severity Assessment
Principal diagnosis was the only factor associated with the physicians' discordant assessment of patients' dyspnea. Patients with admission diagnoses other than HF, COPD, or pneumonia (eg, pulmonary embolism) were more likely to have an accurate assessment of their dyspnea by providers (Table 3). Similar results were obtained in the sensitivity analysis by using a higher cutoff for defining discordance and when assessing predictors for underestimation (results not shown).
| Modeling Probability of Discordance | ||
|---|---|---|
| Physician‐Patient Dyads, OR (95% CI), N = 124 | Nurse‐Patient Dyads, OR (95% CI), N = 363 | |
| ||
| Univariate Analysis | ||
| Body mass index | 1.00 (0.991.01) | 1.00 (0.991.00) |
| Comorbidity score | 1.01 (0.981.05) | 0.99 (0.961.01) |
| Respiratory rate at admission | 1.00 (0.991.02) | 0.99 (0.981.00) |
| Oxygen saturation at admission | 1.00 (1.001.00) | 1.00 (1.001.00) |
| Age (binary) | ||
| 65 years | Referent | Referent |
| >65 years | 1.21 (0.572.55) | 0.96 (0.571.64) |
| Gender | ||
| Female | Referent | Referent |
| Male | 1.10 (0.522.32) | 0.81 (0.481.37) |
| Race | ||
| White | Referent | Referent |
| Nonwhite | 1.02 (0.442.37) | 1.06 (0.581.95) |
| Obese (BMI >30) | 1.43 (0.663.11) | 0.76 (0.441.30) |
| Smoker | 1.36 (0.613.05) | 1.04 (0.591.85) |
| Admitting diagnosis | ||
| Heart failure | Referent | Referent |
| COPD/asthma | 0.68 (0.251.83) | 1.91 (0.983.73)* |
| Pneumonia | 0.38 (0.101.40) | 1.07 (0.462.45) |
| Other | 0.30 (0.110.82)* | 1.54 (0.763.11) |
| Depression | 1.21 (0.572.55) | 1.01 (0.541.86) |
| Multivariable analysis | ||
| Admitting diagnosis | ||
| Congestive heart failure | Referent | Referent |
| COPD | 0.68 (0.251.83) | 1.91 (0.983.73)* |
| Pneumonia | 0.38 (0.101.40) | 1.07 (0.462.45) |
| Other | 0.30 (0.110.82)* | 1.54 (0.763.11) |
In the multivariable analysis that assessed patient‐nurse dyads, the diagnosis of COPD was associated with a marginally significant likelihood of discordance (OR: 1.91; 95% CI: 0.98 to 3.73) (Table 3). Similarly, multivariable analysis identified principal diagnosis to be the only predictor of underestimation, and COPD diagnosis was associated with increased odds of dyspnea underestimation by nurses. When we used a higher cutoff to define discordance, the principal diagnosis of COPD (OR: 3.43, 95% CI: 1.76 to 6.69) was associated with an increased risk of discordance, and smoking (OR: 0.54, 95% CI: 0.29 to 0.99) was associated with a decreased risk of discordance. Overall, 45 patients (32.6%) were discharged on oxygen. The odds of discrepancy (under‐ or overestimation) in dyspnea scores between patient and nurse were 1.7 times higher compared to patients who were not discharged on oxygen, but this association did not reach statistical significance; the odds of discrepancy between patient and physician were 3.88 (95% CI: 1.07 to 14.13).
Documentation of Dyspnea
We found that dyspnea was mentioned in the admission notes in 96% of the charts reviewed; physicians used a qualitative rating (mild, moderate, or severe) to indicate the severity of dyspnea in only 16% of cases, and in 53% a descriptor was added (eg, dyspnea with climbing stairs, gradually increased in the prior week). Nurses were more likely than physicians to use qualitative ratings of dyspnea (26% of cases), and they used a more uniform description of the patient's dyspnea (eg, at rest, at rest and on exertion, or on exertion) than physicians. At discharge, 83% of physicians noted in their discharge summary that dyspnea improved compared with admission but did not refer to the patient's baseline level of dyspnea.
DISCUSSION
In this prospective study of 138 patients hospitalized with cardiopulmonary disease, we found that the agreement between patient's experience of dyspnea and providers' assessment was poor, and the discordance was higher for physicians than for nurses. In more than half of the cases, differences between patient and healthcare providers' assessment of dyspnea were present. One‐third of the time, both physicians and nurses underestimated patients' reported levels of dyspnea. Admitting diagnosis was the only patient factor predicting lack of agreement, and patients with COPD were more likely to have their dyspnea underestimated by nurses. Healthcare providers predominantly documented the presence or absence of dyspnea and rarely used a more nuanced scale.
Discrepancies between patient and provider assessments for pain, depression, and overall health have been reported.[8, 18, 19, 20, 21] One explanation is that patients and healthcare providers measure different factors despite using the same terminology. Furthermore, patient's assessment may be confounded by other symptoms such as anxiety, fatigue, or pain. Physicians and nurses may underevaluate and underestimate the level of breathlessness; however, from the physician perspective, dyspnea is only 1 data point, and providers rely on other measures, such as oxygen saturation, heart rate, respiratory rate, evidence of increasing breathing effort, and arterial blood gas to drive decision making. In a recent study that evaluated the attitudes and beliefs of hospitalists regarding the assessment and management of dyspnea, we found that most hospitalists indicated that awareness of dyspnea severity influences their decision for treatment, diagnostic testing, and timing of the discharge. Moreover, whereas less than half of the respondents reported experience with standardized assessment of dyspnea severity, most stated that such data would be very useful in their practice.[22]
What is the clinical significance of having discordance between patients self‐assessment and providers impression of the patient's severity of dyspnea? First, inaccurate assessment of dyspnea by providers can lead to inadequate treatment and workup. For example, a physician who underestimates the severity of dyspnea may fail to recognize when a complication of the underlying disease develops or may underutilize symptomatic methods for relief of dyspnea. In contrast, a physician who overestimates dyspnea may continue with aggressive treatment when this is not necessary. Second, lack of awareness of dyspnea severity experienced by the patient may result in premature discharge and patient's frustration with the provider, as was shown in several studies evaluating physician‐patient agreement for pain perception.[21, 23] We found that discrepancy between patients and healthcare providers was more pronounced for patients with COPD. In another study using the same patient cohort, we reported that compared to patients with congestive heart failure, those with COPD had more residual dyspnea at discharge; 1 in 4 patients was discharged with a dyspnea score of 5 of greater, and almost half reported symptoms above their baseline.[24] The results from the current study may explain in part why patients with COPD are discharged with higher levels of dyspnea and should alert healthcare providers on the importance of patient‐reported breathlessness. Third, the high level of discordance between healthcare providers and patients may explain the undertreatment of dyspnea in patients with advanced disease. This is supported by our findings that the discordance between patients and physicians was higher if the patient was discharged on oxygen.
One key role of the provider during a clinical encounter is to elicit the patient's symptoms and achieve a shared understanding of what the patient is experiencing. From the patient's perspective, their self‐assessment of dyspnea is more important that the physician's assessment. Fortunately, there is a growing recognition and emphasis on using outcomes that matter to patients, such as dyspnea, to inform judgment about patient care and for clinical research. Numerical measures for assessment of dyspnea exist, are easy to use, and are sensitive to change in patients' dyspnea.[6, 25, 26, 27] Still, it is not standard practice for healthcare providers to ask patients to provide a rating of their dyspnea. When we examined the documentation of dyspnea in the medical record, we found that the description was vague, and providers did not use a standardized validated assessment. Although the dyspnea score decreased during hospitalization, the respiratory rate did not significantly change, indicating that this objective measure may not be reliable in patient assessment. The providers' knowledge of the intensity of the symptom expressed by patients will enable them to track improvement in symptoms over time or in response to therapy. In addition, in this era of multiple handoffs within a hospitalization or from primary care to the hospital, a more uniform assessment could allow providers to follow the severity and time course of dyspnea. The low level of agreement we found between patients and the providers lends support to recommendations regarding a structured dyspnea assessment into routine hospital practice.
Study Strengths and Limitations
This study has several strengths. This is 1 of the very few studies to report on the level of agreement between patients' and providers' assessments of dyspnea. We used a validated, simple dyspnea scale that provides consistency in rating.[28, 29] We enrolled patients with a broad set of diagnoses and complaints, which increases the generalizability of our results, and we surveyed both physicians and nurses. Last, our findings were robust across different cutoff points utilized to characterize discordance and across 3 frequent diagnoses.
The study has several limitations. First, we included only English‐speaking patients, and the results cannot be generalized to patients from other cultures who do not speak English. Second, this is a single‐center study, and practices may be different in other centers; for example, some hospitals may have already implemented a dyspnea assessment tool. Third, we did not collect information on the physician and nurse characteristics such as years in practice. However, a recent study that describes the agreement of breathlessness assessment between nurses, physicians, and mechanically ventilated patients found that underestimation of breathlessness by providers was not associated with professional competencies, previous patient care, or years of working in an intensive care unit.[9] In addition, a systematic review found that length of professional experience is often unrelated to performance measures and outcomes.[30] Finally, although we asked for physicians and nurses assessment close to their visit to the patient, assessment was done from memory, not at the bedside observing the patient.
CONCLUSION
We found that the extent of agreement between a structured patient self‐assessment of dyspnea and healthcare providers' assessment was low. Future studies should prospectively test whether routine assessment of dyspnea results in better acute and long‐term patient outcomes.
Acknowledgements
The authors acknowledge Ms. Anu Joshi for her help with formatting the manuscript and assisting with table preparations. The authors also acknowledge Ms. Katherine Dempsey, Jahnavi Sagi, Sashi Ariyaratne, and Mr. Pradeep Kumbaham for their help with collecting the data.
Disclosures: M.S.S. is the guarantor for this article and had full access to all of the data in the study, and takes responsibility for the integrity and accuracy of the data analysis. M.S.S., P.K.L., E.N., and M.B.R. conceived of the study. M.S.S. and B.M. acquired the data. M.S.S., A.P., P.S.P., R.J.G., and P.K.L. analyzed and interpreted the data. M.S.S. drafted the manuscript. All authors critically reviewed the manuscript for intellectual content. M.S.S. is supported by grant 1K01HL114631‐01A1 from the National Heart, Lung, and Blood Institute of the National Institutes of Health and by the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through grant UL1RR025752. The authors report no conflicts of interest.
- , , , et al. An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea. Am J Respir Crit Care Med. 2012;185(4):435–452.
- , , . National Hospital Ambulatory Medical Care Survey: 2007 emergency department summary. Natl Health Stat Report. 2010;(26):1–31.
- , , , et al. The body‐mass index, airflow obstruction, dyspnea, and exercise capacity index in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(10):1005–1012.
- , , , . Dyspnea is a better predictor of 5‐year survival than airway obstruction in patients with COPD. Chest. 2002;121(5):1434–1440.
- , , . A multidimensional grading system (BODE index) as predictor of hospitalization for COPD. Chest. 2005;128(6):3810–3816.
- , , , et al. American College of Chest Physicians consensus statement on the management of dyspnea in patients with advanced lung or heart disease. Chest. 2010;137(3):674–691.
- , , , et al. Managing dyspnea in patients with advanced chronic obstructive pulmonary disease: a Canadian Thoracic Society clinical practice guideline. Can Respir J. 2011;18(2):69–78.
- , , . Physician vs patient assessment of dyspnea during acute decompensated heart failure. Congest Heart Fail. 2010;16(2):60–64.
- , , , , , . Underestimation of Patient Breathlessness by Nurses and Physicians during a Spontaneous Breathing Trial. Am J Respir Crit Care Med. 2015;192(12):1440–1448.
- , , , , , . The influence of patient‐practitioner agreement on outcome of care. Am J Public Health. 1981;71(2):127–131.
- , , , , , . Listening to parents: The role of symptom perception in pediatric palliative home care. Palliat Support Care. 2016;14(1):13–19.
- , . Validity of the numeric rating scale as a measure of dyspnea. Am J Crit Care. 1998;7(3):200–204.
- , , , , , . Dyspnea scales in the assessment of illiterate patients with chronic obstructive pulmonary disease. Am J Med Sci. 2000;320(4):240–243.
- , , , , . Measuring the dyspnea of decompensated heart failure with a visual analog scale: how much improvement is meaningful? Congest Heart Fail. 2004;10(4):188–191.
- , , , , , . Clinically meaningful changes in quantitative measures of asthma severity. Acad Emerg Med. 2000;7(4):327–334.
- , , , , . A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–759.
- . BlandAltmanLeh: plots (slightly extended) Bland‐Altman plots. Available at: https://cran.r‐project.org/web/packages/BlandAltmanLeh/index.html. Published December 23, 2015. Accessed March 10, 2016.
- , , , , . Correlation of patient and caregiver ratings of cancer pain. J Pain Symptom Manage. 1991;6(2):53–57.
- , , , et al. Depression symptomatology and diagnosis: discordance between patients and physicians in primary care settings. BMC Fam Pract. 2008;9:1.
- , , , , , . Patient‐physician discordance in assessments of global disease severity in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010;62(6):857–864.
- , , , et al. The influence of discordance in pain assessment on the functional status of patients with chronic nonmalignant pain. Am J Med Sci. 2006;332(1):18–23.
- , , , et al. Hospitalist attitudes toward the assessment and management of dyspnea in patients with acute cardiopulmonary diseases. J Hosp Med. 2015;10(11):724–730.
- , , , , . Brief report: patient‐physician agreement as a predictor of outcomes in patients with back pain. J Gen Intern Med. 2005;20(10):935–937.
- , , , , , . The trajectory of dyspnea in hospitalized patients [published online November 24, 2015]. J Pain Symptom Manage. doi: 10.1016/j.jpainsymman.2015.11.005.
- , , , , . Measurement of breathlessness in advanced disease: a systematic review. Respir Med. 2007;101(3):399–410.
- . Review of dyspnoea quantification in the emergency department: is a rating scale for breathlessness suitable for use as an admission prediction tool? Emerg Med Australas. 2007;19(5):394–404.
- , , , . Validation of a verbal dyspnoea rating scale in the emergency department. Emerg Med Australas. 2008;20(6):475–481.
- , , . Measurement of dyspnea: word labeled visual analog scale vs. verbal ordinal scale. Respir Physiol Neurobiol. 2003;134(2):77–83.
- , , , , , . Verbal numerical scales are as reliable and sensitive as visual analog scales for rating dyspnea in young and older subjects. Respir Physiol Neurobiol. 2007;157(2–3):360–365.
- , , . Systematic review: the relationship between clinical experience and quality of health care. Ann Intern Med. 2005;142(4):260–273.
- , , , et al. An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea. Am J Respir Crit Care Med. 2012;185(4):435–452.
- , , . National Hospital Ambulatory Medical Care Survey: 2007 emergency department summary. Natl Health Stat Report. 2010;(26):1–31.
- , , , et al. The body‐mass index, airflow obstruction, dyspnea, and exercise capacity index in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(10):1005–1012.
- , , , . Dyspnea is a better predictor of 5‐year survival than airway obstruction in patients with COPD. Chest. 2002;121(5):1434–1440.
- , , . A multidimensional grading system (BODE index) as predictor of hospitalization for COPD. Chest. 2005;128(6):3810–3816.
- , , , et al. American College of Chest Physicians consensus statement on the management of dyspnea in patients with advanced lung or heart disease. Chest. 2010;137(3):674–691.
- , , , et al. Managing dyspnea in patients with advanced chronic obstructive pulmonary disease: a Canadian Thoracic Society clinical practice guideline. Can Respir J. 2011;18(2):69–78.
- , , . Physician vs patient assessment of dyspnea during acute decompensated heart failure. Congest Heart Fail. 2010;16(2):60–64.
- , , , , , . Underestimation of Patient Breathlessness by Nurses and Physicians during a Spontaneous Breathing Trial. Am J Respir Crit Care Med. 2015;192(12):1440–1448.
- , , , , , . The influence of patient‐practitioner agreement on outcome of care. Am J Public Health. 1981;71(2):127–131.
- , , , , , . Listening to parents: The role of symptom perception in pediatric palliative home care. Palliat Support Care. 2016;14(1):13–19.
- , . Validity of the numeric rating scale as a measure of dyspnea. Am J Crit Care. 1998;7(3):200–204.
- , , , , , . Dyspnea scales in the assessment of illiterate patients with chronic obstructive pulmonary disease. Am J Med Sci. 2000;320(4):240–243.
- , , , , . Measuring the dyspnea of decompensated heart failure with a visual analog scale: how much improvement is meaningful? Congest Heart Fail. 2004;10(4):188–191.
- , , , , , . Clinically meaningful changes in quantitative measures of asthma severity. Acad Emerg Med. 2000;7(4):327–334.
- , , , , . A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–759.
- . BlandAltmanLeh: plots (slightly extended) Bland‐Altman plots. Available at: https://cran.r‐project.org/web/packages/BlandAltmanLeh/index.html. Published December 23, 2015. Accessed March 10, 2016.
- , , , , . Correlation of patient and caregiver ratings of cancer pain. J Pain Symptom Manage. 1991;6(2):53–57.
- , , , et al. Depression symptomatology and diagnosis: discordance between patients and physicians in primary care settings. BMC Fam Pract. 2008;9:1.
- , , , , , . Patient‐physician discordance in assessments of global disease severity in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010;62(6):857–864.
- , , , et al. The influence of discordance in pain assessment on the functional status of patients with chronic nonmalignant pain. Am J Med Sci. 2006;332(1):18–23.
- , , , et al. Hospitalist attitudes toward the assessment and management of dyspnea in patients with acute cardiopulmonary diseases. J Hosp Med. 2015;10(11):724–730.
- , , , , . Brief report: patient‐physician agreement as a predictor of outcomes in patients with back pain. J Gen Intern Med. 2005;20(10):935–937.
- , , , , , . The trajectory of dyspnea in hospitalized patients [published online November 24, 2015]. J Pain Symptom Manage. doi: 10.1016/j.jpainsymman.2015.11.005.
- , , , , . Measurement of breathlessness in advanced disease: a systematic review. Respir Med. 2007;101(3):399–410.
- . Review of dyspnoea quantification in the emergency department: is a rating scale for breathlessness suitable for use as an admission prediction tool? Emerg Med Australas. 2007;19(5):394–404.
- , , , . Validation of a verbal dyspnoea rating scale in the emergency department. Emerg Med Australas. 2008;20(6):475–481.
- , , . Measurement of dyspnea: word labeled visual analog scale vs. verbal ordinal scale. Respir Physiol Neurobiol. 2003;134(2):77–83.
- , , , , , . Verbal numerical scales are as reliable and sensitive as visual analog scales for rating dyspnea in young and older subjects. Respir Physiol Neurobiol. 2007;157(2–3):360–365.
- , , . Systematic review: the relationship between clinical experience and quality of health care. Ann Intern Med. 2005;142(4):260–273.
PICC Use in Adults With Pneumonia
Pneumonia is the most common cause of unplanned hospitalization in the United States.[1] Despite its clinical toll, the management of this disease has evolved markedly. Expanding vaccination programs, efforts to improve timeliness of antibiotic therapy, and improved processes of care are but a few developments that have improved outcomes for patients afflicted with this illness.[2, 3]
Use of peripherally inserted central catheters (PICCs) is an example of a modern development in the management of patients with pneumonia.[4, 5, 6, 7] PICCs provide many of the benefits associated with central venous catheters (CVCs) including reliable venous access for delivery of antibiotics, phlebotomy, and invasive hemodynamic monitoring. However, as they are placed in veins of the upper extremity, PICCs bypass insertion risks (eg, injury to the carotid vessels or pneumothorax) associated with placement of traditional CVCs.[8] Because they offer durable venous access, PICCs also facilitate care transitions while continuing intravenous antimicrobial therapy in patients with pneumonia.
However, accumulating evidence also suggests that PICCs are associated with important complications, including central lineassociated bloodstream infectionand venous thromboembolism.[9, 10] Furthermore, knowledge gaps in clinicians regarding indications for appropriate use and management of complications associated with PICCs have been recognized.[10, 11] These elements are problematic because reports of unjustified and inappropriate PICC use are growing in the literature.[12, 13] Such concerns have prompted a number of policy calls to improve PICC use, including Choosing Wisely recommendations by various professional societies.[14, 15]
As little is known about the prevalence or patterns of PICC use in adults hospitalized with pneumonia, we conducted a retrospective cohort study using data from a large network of US hospitals.
METHODS
Setting and Participants
We included patients from hospitals that participated in Premier's inpatient dataset, a large, fee‐supported, multipayer administrative database that has been used extensively in health services research to measure quality of care and comparative effectiveness of interventions.[16] Participating hospitals represent all regions of the United States and include teaching and nonteaching facilities in rural and urban locations. In addition to variables found in the uniform billing form, the Premier inpatient database also includes a date‐stamped list of charges for procedures conducted during hospitalization such as PICC placement. As PICC‐specific data are not available in most nationally representative datasets, Premier offers unique insights into utilization, timing, and factors associated with use of PICCs in hospitalized settings.
We included adult patients aged 18 years who were (1) admitted with a principal diagnosis of pneumonia present on admission, or secondary diagnosis of pneumonia if paired with a principal diagnosis of sepsis, respiratory failure, or influenza; (2) received at least 1 day of antibiotics between July 1, 2007 and November 30, 2011, and (3) underwent chest x‐ray or computed tomography (CT) at the time of admission. International Classification of Disease, 9th Revision, Clinical Modification (ICD‐9‐CM) codes were used for patient selection. Patients who were not admitted (eg, observation cases), had cystic fibrosis, or marked as pneumonia not present on admission were excluded. For patients who had more than 1 hospitalization during the study period, a single admission was randomly selected for inclusion.
Patient, Physician, and Hospital Data
For all patients, age, gender, marital status, insurance, race, and ethnicity were captured. Using software provided by the Healthcare Costs and Utilization Project, we categorized information on 29 comorbid conditions and computed a combined comorbidity score as described by Gagne et al.[17] Patients were considered to have healthcare‐associated pneumonia (HCAP) if they were: (1) admitted from a skilled nursing or a long‐term care facility, (2) hospitalized in the previous 90 days, (3) on dialysis, or (4) receiving immunosuppressing medications (eg, chemotherapy or steroids equivalent to at least 20 mg of prednisone per day) at the time of admission. Information on specialty of the admitting physician and hospital characteristics (eg, size, location, teaching status) were sourced through Premier data.
Receipt of PICCs and Related Therapies
Among eligible adult patients hospitalized with pneumonia, we identified patients who received a PICC at any time during hospitalization via PICC‐specific billing codes. Non‐PICC devices (eg, midlines, Hickman catheters) were not included. For all insertions, we assessed day of PICC placement relative to admission date. Data on type of PICC (eg, power‐injection capable, antibiotic coating) or PICC characteristics (size, number of lumens) were not available. We used billing codes to assess use of invasive or noninvasive ventilation, vasopressors, and administration of pneumonia‐specific antibiotics (eg, ‐lactams, macrolides, fluoroquinolones). Early exposure was defined when a billing code appeared within 2 days of hospital admission.
Outcomes of Interest
The primary outcome of interest was receipt of a PICC. Additionally, we assessed factors associated with PICC placement and variation in risk‐standardized rates of PICC use between hospitals.
Statistical Analyses
Patient and hospital characteristics were summarized using frequencies for categorical variables and medians with interquartile ranges for continuous variables. We examined association of individual patient and hospital characteristics with use of PICCs using generalized estimating equations models with a logit link for categorical variables and identity link for continuous variables, accounting for patient clustering within hospitals.
| Characteristic | Total, No. (%) | No PICC, No. (%) | PICC, No. (%) | P Value* |
|---|---|---|---|---|
| ||||
| 545,250 (100) | 503,401 (92.3) | 41,849 (7.7) | ||
| Demographics | ||||
| Age, median (Q1Q3), y | 71 (5782) | 72 (5782) | 69 (5780) | <0.001 |
| Gender | <0.001 | |||
| Male | 256,448 (47.0) | 237,232 (47.1) | 19,216 (45.9) | |
| Female | 288,802 (53.0) | 266,169 (52.9) | 22,633 (54.1) | |
| Race/ethnicity | <0.001 | |||
| White | 377,255 (69.2) | 346,689 (68.9) | 30,566 (73.0) | |
| Black | 63,345 (11.6) | 58,407 (11.6) | 4,938 (11.8) | |
| Hispanic | 22,855 (4.2) | 21,716 (4.3) | 1,139 (2.7) | |
| Other | 81,795 (15.0) | 76,589 (15.2) | 5,206 (12.4) | |
| Admitting specialty | <0.001 | |||
| Internal medicine | 236,859 (43.4) | 218,689 (43.4) | 18,170 (43.4) | |
| Hospital medicine | 116,499 (21.4) | 107,671 (21.4) | 8,828 (21.1) | |
| Family practice | 80,388 (14.7) | 75,482 (15.0) | 4,906 (11.7) | |
| Critical care and pulmonary | 35,670 (6.5) | 30,529 (6.1) | 41,849 (12.3) | |
| Geriatrics | 4,812 (0.9) | 4,098 (0.8) | 714 (1.7) | |
| Other | 71,022 (13.0) | 66,932 (13.3) | 4,090 (9.8) | |
| Insurance | <0.001 | |||
| Medicare | 370,303 (67.9) | 341,379 (67.8) | 28,924 (69.1) | |
| Medicaid | 45,505 (8.3) | 41,100 (8.2) | 4,405 (10.5) | |
| Managed care | 69,984 (12.8) | 65,280 (13.0) | 4,704 (11.2) | |
| Commercialindemnity | 20,672 (3.8) | 19,251 (3.8) | 1,421 (3.4) | |
| Other | 38,786 (7.1) | 36,391 (7.2) | 2,395 (5.7) | |
| Comorbidities | ||||
| Gagne combined comorbidity score, median (Q1Q3) | 2 (15) | 2 (14) | 4 (26) | <0.001 |
| Hypertension | 332,347 (60.9) | 306,964 (61.0) | 25,383 (60.7) | 0.13 |
| Chronic pulmonary disease | 255,403 (46.8) | 234,619 (46.6) | 20,784 (49.7) | <0.001 |
| Diabetes | 171,247 (31.4) | 155,540 (30.9) | 15,707 (37.5) | <0.001 |
| Congestive heart failure | 146,492 (26.9) | 131,041 (26.0) | 15,451 (36.9) | <0.001 |
| Atrial fibrillation | 108,405 (19.9) | 97,124 (19.3) | 11,281 (27.0) | <0.001 |
| Renal failure | 104,404 (19.1) | 94,277 (18.7) | 10,127 (24.2) | <0.001 |
| Nicotine replacement therapy/tobacco use | 89,938 (16.5) | 83,247 (16.5) | 6,691 (16.0) | <0.001 |
| Obesity | 60,242 (11.0) | 53,268 (10.6) | 6,974 (16.7) | <0.001 |
| Coagulopathy | 41,717 (7.6) | 35,371 (7.0) | 6,346 (15.2) | <0.001 |
| Prior stroke (1 year) | 26,787 (4.9) | 24,046 (4.78) | 2,741 (6.55) | <0.001 |
| Metastatic cancer | 21,868 (4.0) | 20,244 (4.0) | 1,624 (3.9) | 0.16 |
| Solid tumor w/out metastasis | 21,083 (3.9) | 19,380 (3.8) | 1,703 (4.1) | 0.12 |
| Prior VTE (1 year) | 19,090 (3.5) | 16,906 (3.4) | 2,184 (5.2) | <0.001 |
| Chronic liver disease | 16,273 (3.0) | 14,207 (2.8) | 2,066 (4.9) | <0.001 |
| Prior bacteremia (1 year) | 4,106 (0.7) | 3,584 (0.7) | 522 (1.2) | <0.001 |
| Nephrotic syndrome | 671 (0.1) | 607 (0.1) | 64 (0.2) | 0.03 |
| Morbidity markers | ||||
| Type of pneumonia | <0.001 | |||
| CAP | 376,370 (69.1) | 352,900 (70.1) | 23,830 (56.9) | |
| HCAP | 168,520 (30.9) | 150,501 (29.9) | 18,019 (43.1) | |
| Sepsis present on admission | 114,578 (21.0) | 96,467 (19.2) | 18,111 (43.3) | <0.001 |
| Non‐invasive ventilation | 47,913(8.8) | 40,599 (8.1) | 7,314 (17.5) | <0.001 |
| Invasive mechanical ventilation | 56,179 (10.3) | 44,228 (8.8) | 11,951 (28.6) | <0.001 |
| ICU status | 97,703 (17.9) | 80,380 (16.0) | 17,323 (41.4) | <0.001 |
| Vasopressor use | 48,353 (8.9) | 38,030 (7.6) | 10,323 (24.7) | <0.001 |
| Antibiotic/medication use | ||||
| Anti‐MRSA agent (vancomycin) | 146,068 (26.8) | 123,327 (24.5) | 22,741 (54.3) | <0.001 |
| Third‐generation cephalosporin | 250,782 (46.0) | 235,556 (46.8) | 15,226 (36.4) | <0.001 |
| Anti‐Pseudomonal cephalosporin | 41,798 (7.7) | 36,982 (7.3) | 4,816 (11.5) | <0.001 |
| Anti‐Pseudomonal ‐lactam | 122,215 (22.4) | 105,741 (21.0) | 16,474 (39.4) | <0.001 |
| Fluroquinolone | 288,051 (52.8) | 267,131 (53.1) | 20,920 (50.0) | <0.001 |
| Macrolide | 223,737 (41.0) | 210,954 (41.9) | 12,783 (30.5) | <0.001 |
| Aminoglycoside | 15,415 (2.8) | 12,661 (2.5) | 2,754 (6.6) | <0.001 |
| Oral steroids | 44,486 (8.2) | 41,586 (8.3) | 2,900 (6.9) | <0.001 |
| Intravenous steroids | 146,308 (26.8) | 133,920 (26.6) | 12,388 (29.6) | <0.001 |
| VTE prophylaxis with LMWH | 190,735 (35.0) | 174,612 (34.7) | 16,123 (38.5) | 0.01 |
| Discharge disposition | ||||
| Home | 282,146 (51.7) | 272,604(54.1) | 9,542 (22.8) | <0.001 |
| Home with home health | 71,977 (13.2) | 65,289 (13.0) | 6,688 (16.0) | <0.001 |
| Skilled nursing facility | 111,541 (20.5) | 97,113 (19.3) | 14,428 (34.5) | <0.001 |
| Hospice | 20,428 (3.7) | 17,902 (3.6) | 2,526 (6.0) | <0.001 |
| Expired | 47,733 (8.7) | 40,768 (8.1) | 6,965 (16.6) | <0.001 |
| Other | 11,425 (2.1) | 9,725 (1.9) | 1,700 (4.1) | <0.001 |
We then developed a multivariable hierarchical generalized linear model (HGLM) for PICC placement with a random effect for hospital. In this model, we included patient demographics, comorbidities, sepsis on admission, type of pneumonia (eg, HCAP vs community‐associated pneumonia [CAP]), admitting physician specialty, and indicators for early receipt of specific treatments such as guideline‐recommended antibiotics, vasopressors, ventilation (invasive or noninvasive), and pneumatic compression devices for prophylaxis of deep vein thrombosis.
To understand and estimate between‐hospital variation in PICC use, we calculated risk‐standardized rates of PICC use (RSPICC) across hospitals using HGLM methods. These methods are also employed by the Centers for Medicare and Medicaid Services to calculate risk‐standardized measures for public reporting.[18] Because hospital rates of PICC use were highly skewed (21.2% [n = 105] of hospitals had no patients with PICCs), we restricted this model to the 343 hospitals that had at least 5 patients with a PICC to obtain stable estimates. For each hospital, we estimated a predicted rate of PICC use (pPICC) as the sum of predicted probabilities of PICC receipt from patient factors and the random intercept for hospital in which they were admitted. We then calculated an expected rate of PICC use (ePICC) per hospital as the sum of expected probabilities of PICC receipt from patient factors only. RSPICC for each hospital was then computed as the product of the overall unadjusted mean PICC rate (PICC) from all patients and the ratio of the predicted to expected PICC rate (uPICC*[pPICC/ePICC]).[19] Kruskal‐Wallis tests were used to evaluate the association between hospital characteristics with RSPICC rates. To evaluate the impact of the hospital in variation in PICC use, we assessed the change in likelihood ratio of a hierarchical model with hospital random effects compared to a logistic regression model with patient factors only. In addition, we estimated the intraclass correlation (ICC) to assess the proportion of variation in PICC use associated with the hospital, and the median odds ratio (MOR) from the hierarchical model. The MOR is the median of a set of odds ratios comparing 2 patients with the same set of characteristics treated at 2 randomly selected hospitals.[20, 21, 22] All analyses were performed using the Statistical Analysis System version 9.3 (SAS Institute, Inc., Cary, NC) and Stata 13 (StataCorp Inc., College Station, TX).
Ethical and Regulatory Oversight
Permission to conduct this study was obtained from the institutional review board at Baystate Medical Center, Springfield, Massachusetts. The study did not qualify as human subjects research and made use of fully deidentified data.
RESULTS
Between July 2007 and November 2011, 634,285 admissions representing 545,250 unique patients from 495 hospitals met eligibility criteria and were included in the study (Figure 1). Included patients had a median age of 71 years (interquartile range [IQR]: 5782), and 53.0% were female. Most patients were Caucasian (69.2%), unmarried (51.6%), and insured by Medicare (67.9%). Patients were admitted to the hospital by internal medicine providers (43.4%), hospitalists (21.4%), and family practice providers (14.7%); notably, critical care and pulmonary medicine providers admitted 6.5% of patients. The median Gagne comorbidity score was 2 (IQR: 15). Hypertension, chronic obstructive pulmonary disease, diabetes, and congestive heart failure were among the most common comorbidities observed (Table 1).
Among eligible patients, 41,849 (7.7%) received a PICC during hospitalization. Approximately a quarter of all patients who received PICCs did so by hospital day 2; 90% underwent insertion by hospital day 11 (mean = 5.4 days, median = 4 days). Patients who received PICCs were younger (median IQR: 69 years, 5780 years) but otherwise demographically similar to those that did not receive PICCs (median IQR: 72 years, 5782 years). Compared to other specialties, patients admitted by critical care/pulmonary providers were twice as likely to receive PICCs (12.3% vs 6.1%, P < .001). Patients who received PICCs had higher comorbidity scores than those who did not (median Gagne comorbidity score 4 vs 2, P < 0.001) and were more likely to be diagnosed with HCAP (43.1% vs 29.9%, P < 0.001) than CAP (56.9% vs 70.1%, P < 0.001).
PICC recipients were also more likely to receive intensive care unit (ICU) level of care (41.4% vs 16%, P < 0.001) and both noninvasive (17.5% vs 8.1%, P < 0.001) and invasive ventilation (28.6% vs 8.8%, P < 0.001) upon admission. Vasopressor use was also significantly more frequent in patients who received PICCs (24.7% vs 7.6%, P < 0.001) compared to those who did not receive these devices. Patients with PICCs were more often discharged to skilled nursing facilities (34.5% vs 19.3%) than those without PICCs.
Characteristics Associated With PICC Use Following Multivariable Modeling
Using HGLM with a random hospital effect, multiple patient characteristics were associated with PICC use (Table 2). Patients 65 years of age were less likely to receive a PICC compared to younger patients (odds ratio [OR]: 0.81, 95% confidence interval [CI]: 0.79‐0.84). Weight loss (OR: 2.03, 95% CI: 1.97‐2.10), sepsis on admission (OR: 1.80, 95% CI: 1.75‐1.85), and ICU status on hospital day 1 or 2 (OR: 1.70, 95% CI: 1.64‐1.75) represented 3 factors most strongly associated with PICC use.
| Patient Characteristic | Odds Ratio | 95% Confidence Intervals |
|---|---|---|
| ||
| Age group (>66 vs 65 years) | 0.82 | 0.790.84 |
| Race/ethnicity | ||
| Other | 1.02 | 0.971.06 |
| Black | 0.99 | 0.951.03 |
| Hispanic | 0.82 | 0.760.88 |
| White | Referent | |
| Marital status | ||
| Other/missing | 1.07 | 1.011.14 |
| Single | 1.02 | 1.001.05 |
| Married | Referent | |
| Insurance payor | ||
| Other | 0.85 | 0.800.89 |
| Medicaid | 1.13 | 1.081.18 |
| Managed care | 0.95 | 0.910.99 |
| Commercialindemnity | 0.93 | 0.871.00 |
| Medicare | Referent | |
| Admitting physician specialty | ||
| Pulmonary/critical care medicine | 1.18 | 1.131.24 |
| Family practice | 1.01 | 0.971.05 |
| Geriatric medicine (FP and IM) | 1.85 | 1.662.05 |
| Hospitalist | 0.94 | 0.910.98 |
| Other specialties | 1.02 | 0.971.06 |
| Internal medicine | Referent | |
| Comorbidities | ||
| Congestive heart failure | 1.27 | 1.241.31 |
| Valvular disease | 1.11 | 1.071.15 |
| Pulmonary circulation disorders | 1.37 | 1.321.42 |
| Peripheral vascular disease | 1.09 | 1.051.13 |
| Hypertension | 0.94 | 0.920.97 |
| Paralysis | 1.59 | 1.511.67 |
| Other neurological disorders | 1.20 | 1.161.23 |
| Chronic lung disease | 1.10 | 1.071.12 |
| Diabetes | 1.13 | 1.101.16 |
| Hypothyroidism | 1.03 | 1.001.06 |
| Liver disease | 1.16 | 1.101.23 |
| Ulcer | 1.86 | 1.153.02 |
| Lymphoma | 0.88 | 0.810.96 |
| Metastatic cancer | 0.75 | 0.710.80 |
| Solid tumor without metastasis | 0.93 | 0.880.98 |
| Arthritis | 1.22 | 1.161.28 |
| Obesity | 1.47 | 1.421.52 |
| Weight loss | 2.03 | 1.972.10 |
| Blood loss | 1.69 | 1.551.85 |
| Deficiency anemias | 1.40 | 1.371.44 |
| Alcohol abuse | 1.19 | 1.131.26 |
| Drug abuse | 1.31 | 1.231.39 |
| Psychoses | 1.16 | 1.111.21 |
| Depression | 1.10 | 1.061.13 |
| Renal failure | 0.96 | 0.930.98 |
| Type of pneumonia | ||
| HCAP | 1.03 | 1.011.06 |
| CAP | Referent | |
| Sepsis (POA) | 1.80 | 1.751.85 |
| Antibiotic exposure | ||
| Anti‐MRSA agent | 1.72 | 1.671.76 |
| Anti‐Pseudomonal carbapenem | 1.37 | 1.311.44 |
| Non‐Pseudomonal carbapenem | 1.48 | 1.331.66 |
| Third‐generation cephalosporin | 1.04 | 1.011.07 |
| Anti‐Pseudomonal cephalosporin | 1.25 | 1.201.30 |
| Anti‐Pseudomonal ‐lactam | 1.27 | 1.231.31 |
| Aztreonam | 1.31 | 1.231.40 |
| Non‐Pseudomonal ‐lactam | 1.36 | 1.231.50 |
| ‐lactam | 1.55 | 1.261.90 |
| Respiratory quinolone | 0.90 | 0.870.92 |
| Macrolide | 0.85 | 0.820.88 |
| Doxycycline | 0.94 | 0.871.01 |
| Aminoglycoside | 1.21 | 1.141.27 |
| Vasopressors | 1.06 | 1.031.10 |
| Noninvasive ventilation | 1.29 | 1.251.34 |
| Invasive ventilation | 1.66 | 1.611.72 |
| Intensive care unit on admission | 1.70 | 1.641.75 |
| Atrial fibrillation | 1.26 | 1.221.29 |
| Upper extremity chronic DVT | 1.61 | 1.132.28 |
| Nicotine replacement therapy/tobacco abuse | 0.91 | 0.880.94 |
| Aspirin | 0.94 | 0.920.97 |
| Warfarin | 0.90 | 0.860.94 |
| LMWH, prophylactic dose | 1.10 | 1.081.13 |
| LMWH, treatment dose | 1.22 | 1.161.29 |
| Intravenous steroids | 1.05 | 1.021.08 |
| Bacteremia (prior year) | 1.14 | 1.021.27 |
| VTE (prior year) | 1.11 | 1.061.18 |
| Pneumatic compression device | 1.25 | 1.081.45 |
| Invasive ventilation (prior year) | 1.17 | 1.111.24 |
| Irritable bowel disease | 1.19 | 1.051.36 |
Therapy with potent parenteral antimicrobials including antimethicillin‐resistant Staphylococcus aureus agents (OR: 1.72, 95% CI: 1.67‐1.76), antipseudomonal ‐lactamases (OR: 1.27, 95% CI: 1.23‐1.31), and carbapenems (OR: 1.37, 95% CI: 1.31‐1.44) were significantly associated with PICC use. Conversely, use of macrolides (OR: 0.85, 95% CI: 0.82‐0.88) or respiratory fluoroquinolones (OR: 0.90, 95% CI: 0.87‐0.92) were associated with lower likelihood of PICC use. After adjusting for antimicrobial therapy, HCAP was only slightly more likely to result in PICC use than CAP (OR: 1.03, 95% CI: 1.01‐1.06). Compared to internal medicine providers, admission by geriatricians and critical care/pulmonary specialists was associated with greater likelihood of PICC use (OR: 1.85, 95% CI: 1.66‐2.05 and OR: 1.18, 95% CI: =1.13‐1.24, respectively). Admission by hospitalists was associated with a modestly lower likelihood of PICC placement (OR: 0.94, 95% CI: 0.91‐0.98).
Hospital Level Variation in PICC Use
To ensure stable estimates of hospital PICC use, we excluded 152 facilities (31%): 10% had no patients with PICCs and 21% had <5 patients who received a PICC. Therefore, RSPICC was estimated for 343 of 495 facilities (69%) (Figure 2). In these facilities, RSPICC varied from 0.3% to 41.7%. Hospital RSPICC was significantly associated with hospital location (median 11.9% vs 7.8% for urban vs rural hospitals respectively, P = 0.05). RSPICCs were also greater among hospitals in Southern (11.3%), Western (12.7%), and Midwest (12.0%) regions of the nation compared to those in the Northeast (8.4%) (P = 0.02) (Table 3).
| Hospital Characteristic (No.) | Median (IQR), % | P Value |
|---|---|---|
| ||
| Bed size | 0.12 | |
| 200 beds (106) | 9.1 (4.816.3) | |
| 201 beds (237) | 11.6 (5.817.6) | |
| Rural/urban | 0.05 | |
| Urban (275) | 11.9 (5.517.4) | |
| Rural (68) | 7.8 (5.014.0) | |
| Region | 0.02 | |
| Northeast (50) | 8.4 (3.913.0) | |
| Midwest (69) | 12.0 (5.817.4) | |
| West (57) | 12.7 (7.617.0) | |
| South (167) | 11.3 (4.817.8) | |
| Teaching status | 0.77 | |
| Nonteaching (246) | 10.9 (5.017.4) | |
| Teaching (97) | 12.0 (5.816.9) | |
A likelihood ratio test comparing the hierarchical model to a logistic model with patient factors only was highly significant (P < 0.001), indicating that the hospital where the patient was treated had a major impact on receipt of PICC after accounting for patient factors. The MOR was 2.71, which is a larger effect than we found for any of the individual patient characteristics. The proportion of variance explained by hospitals was 25% (95% CI: 22%‐28%), as measured by the ICC.
DISCUSSION
In this study of 545,250 adults hospitalized with pneumonia, we found that approximately 8% of patients received a PICC. Patients who received PICCs had more comorbidities, were more frequently diagnosed with HCAP, and were more often admitted to the ICU, where they experienced greater rates of mechanical ventilation, noninvasive ventilation, and vasopressor use compared to those who did not receive a PICC. Additionally, risk‐adjusted rates of PICC use varied as much as 10‐fold across institutions. In fact, almost 70% of the total variation in rates of PICC use remained unexplained by hospital or patient characteristics. Although use of PICCs is often clinically nuanced in ways that are difficult to capture in large datasets (eg, difficult venous access or inability to tolerate oral medications), the substantial variation of PICC use observed suggests that physician and institutional practice styles are the major determinants of PICC placement during a hospitalization for pneumonia. Because PICCs are associated with serious complications, and evidence regarding discretionary use is accumulating, a research agenda examining reasons for such use and related outcomes appears necessary.
The placement of PICCs has grown substantially in hospitalized patients all over the world.[23, 24] Although originally developed for total parenteral nutrition in surgical patients,[25] contemporary reports of PICC use in critical illness,[26] diseases such as cystic fibrosis,[27] and even pregnancy[28] are now common. Although PICCs are clinically invaluable in many of these conditions, growing use of these devices has led to the realization that benefits may be offset by complications.[9, 10, 29, 30] Additionally, recent data suggest that not all PICCs may be used for appropriate reasons. For instance, in a decade‐long study at a tertiary care center, changes in patterns of PICC use including shortened dwell times, multiple insertions in a single patient, and unclear indications for use were reported.[11] In another study at an academic medical center, a substantial proportion of PICCs were found to be idle or unjustified.[12] It comes as little surprise, then, that a recent multicenter study found that 1 out of every 5 clinicians did not even know that their patient had a PICC.[29] Although calls to improve PICC use in the hospital setting have emerged, strategies to do so are limited by data that emanate from single‐center reports or retrospective designs. No other studies reporting use of PICCs across US hospitals for any clinical condition currently exist.[31]
We found that patients with weight loss, those with greater combined comorbidity scores, and those who were critically ill or diagnosed with sepsis were more likely to receive PICCs than others. These observations suggest that PICC use may reflect underlying severity of illness, as advanced care such as ventilator support was often associated with PICC use. Additionally, discharge to a skilled nursing facility was frequently associated with PICC placement, a finding consistent with a recent study evaluating the use of PICCs in these settings.[32] However, a substantial proportion of PICC use remained unexplained by available patient or hospital factors. Although our study was not specifically designed to examine this question, a possible reason may relate to unmeasured institutional factors that influence the propensity to use a PICC, recently termed as PICC culture.[33] For example, it is plausible that hospitals with nursing‐led PICC teams or interventional radiology (such as teaching hospitals) are more likely to use PICCs than those without such operators. This hypothesis may explain why urban, larger, and teaching hospitals exhibited higher rates of PICC use. Conversely, providers may have an affinity toward PICC use that is predicated not just by operator availability, but also local hospital norms. Understanding why some facilities use PICCs at higher rates than others and implications of such variation with respect to patient safety, cost, and outcomes is important. Study designs that use mixed‐methods approaches or seek to qualitatively understand reasons behind PICC use are likely to be valuable in this enquiry.
Our study has limitations. First, we used an administrative dataset and ICD‐9‐CM codes rather than clinical data from medical records to identify cases of pneumonia or comorbidities. Our estimates of PICC use across hospitals thus may not fully account for differences in severity of illness, and it is possible that patients needed a PICC for reasons that we could not observe. However, the substantial variation observed in rates of PICC use across hospitals is unlikely to be explained by differences in patient severity of illness, documentation, or coding practices. Second, as PICC removal codes were not available, we are unable to comment on how often hospitalized pneumonia patients were discharged with PICCs or received antimicrobial therapy beyond their inpatient stay. Third, although we observed that a number of patient and hospital factors were associated with PICC receipt, our study was not designed to determine the reasons underlying these patterns.
These limitations aside, our study has important strengths. To our knowledge, this is the first study to report utilization and outcomes associated with PICC use among those hospitalized with pneumonia across the United States. The inclusion of a large number of patients receiving care in diverse facilities lends a high degree of external validity to our findings. Second, we used advanced modeling to identify factors associated with PICC use in hospitalized patients with pneumonia, producing innovative and novel findings. Third, our study is the first to show the existence of substantial variation in rates of PICC use across US hospitals within the single disease state of pneumonia. Understanding the drivers of this variability is important as it may inform future studies, policies, and practices to improve PICC use in hospitalized patients.
In conclusion, we found that PICC use in patients hospitalized with pneumonia is common and highly variable. Future studies examining the contextual factors behind PICC use and their association with outcomes are needed to facilitate efforts to standardize PICC use across hospitals.
Disclosures
Dr. Chopra is supported by a career development award (1‐K08‐HS022835‐01) from the Agency of Healthcare Research and Quality. The authors report no conflicts of interest.
- , . Reasons for being admitted to the hospital through the emergency department, 2003. Healthcare Cost and Utilization Project Statistical Brief 2. Rockville, MD: Agency for Healthcare Research and Quality. Available at: www.hcup‐us.ahrq.gov/reports/statbriefs/sb2.pdf. Published February 2006. Accessed June 27, 2014.
- , , , et al. National patterns of risk‐standardized mortality and readmission after hospitalization for acute myocardial infarction, heart failure, and pneumonia: update on publicly reported outcomes measures based on the 2013 release. J Gen Intern Med. 2014;29(10):1333–1340.
- , , , et al. Quality of care for elderly patients hospitalized for pneumonia in the United States, 2006 to 2010. JAMA Intern Med. 2014;174(11):1806–1814.
- , . PICC lines: the latest home care challenge. RN. 1990;53(1):44–51.
- , , , . Peripherally inserted central catheters in an acute‐care hospital. Arch Intern Med. 1994;154(16):1833–1837.
- , . The peripherally inserted central catheter: a retrospective look at three years of insertions. J Intraven Nurs. 1993;16(2):92–103.
- , , , . Peripherally inserted central catheters in general medicine. Mayo Clin Proc. 1997;72(3):225–233.
- , , . Two‐year trends of peripherally inserted central catheter‐line complications at a tertiary‐care hospital: role of nursing expertise. Infect Control Hosp Epidemiol. 2001;22(6):377–379.
- , , , , , . PICC‐associated bloodstream infections: prevalence, patterns, and predictors. Am J Med. 2014;127(4):319–328.
- , , , , . The risk of bloodstream infection associated with peripherally inserted central catheters compared with central venous catheters in adults: a systematic review and meta‐analysis. Infect Control Hosp Epidemiol. 2013;34(9):908–918.
- , , , , , . Peripherally inserted central catheters: use at a tertiary care pediatric center. J Vasc Interv Radiol. 2013;24(9):1323–1331.
- , , , et al. Temporary central venous catheter utilization patterns in a large tertiary care center: tracking the “idle central venous catheter”. Infect Control Hosp Epidemiol. 2012;33(1):50–57.
- , , , , . Inappropriate intravascular device use: a prospective study. Journal Hosp Infect. 2011;78(2):128–132.
- , , , et al. Enhancing patient‐centered care: SGIM and choosing wisely. J Gen Intern Med. 2014;29(3):432–433.
- , , , et al. Critical and honest conversations: the evidence behind the “Choosing Wisely” campaign recommendations by the American Society of Nephrology. Clin J Am Soc Nephrol. 2012;7(10):1664–1672.
- , , , et al. Using highly detailed administrative data to predict pneumonia mortality. PLoS One. 2014;9(1):e87382.
- , , , , . A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–759.
- , , , , . Hospitals with the highest intensive care utilization provide lower quality pneumonia care to the elderly. Crit Care Med. 2015;43(6):1178–1186.
- , . Statistical and clinical aspects of hospital outcomes profiling. Stat Sci. 2007;22(2):206–226.
- , . Appropriate assessment of neighborhood effects on individual health: integrating random and fixed effects in multilevel logistic regression. Am J Epidemiol. 2005;161(1):81–88.
- , , , . Interpreting parameters in the logistic regression model with random effects. Biometrics. 2000;56(3):909–914.
- , , , . Hospital‐level associations with 30‐day patient mortality after cardiac surgery: a tutorial on the application and interpretation of marginal and multilevel logistic regression. BMC Med Res Methodol. 2012;12:28.
- , , , . Experiences of the first PICC team in the Czech Republic. Br J Nurs. 2015;24(suppl 2):S4–S10.
- , , , et al. Greece reports prototype intervention with first peripherally inserted central catheter: case report and literature review. J Vasc Nurs. 2012;30(3):88–93.
- Total intravenous nutrition with peripherally inserted silicone elastomer central venous catheters. Arch Surg. 1975;110(5):644–646.
- , . Focus on peripherally inserted central catheters in critically ill patients. World J Crit Care Med. 2014;3(4):80–94.
- , , , et al. Quality improvement initiative to reduce deep vein thrombosis associated with peripherally inserted central catheters in adults with cystic fibrosis. Ann Am Thorac Soc. 2014;11(9):1404–1410.
- , , , . Peripherally Inserted central catheter (PICC) complications during pregnancy. JPEN J Parenter Enteral Nutr. 2013;38(5):595–601.
- , , , et al. Do clinicians know which of their patients have central venous catheters?: a multicenter observational study. Ann Intern Med. 2014;161(8):562–567.
- , , , et al. Risk of venous thromboembolism associated with peripherally inserted central catheters: a systematic review and meta‐analysis. Lancet. 2013;382(9889):311–325.
- , , . The problem with peripherally inserted central catheters. JAMA. 2012;308(15):1527–1528.
- , , , et al. Peripherally inserted central catheter use in skilled nursing facilities: a pilot study. J Am Geriatr Soc. 2015;63(9):1894–1899.
- , , , , . Inpatient venous access practices: PICC culture and the kidney patient. J Vasc Access. 2015;16(3):206–210.
Pneumonia is the most common cause of unplanned hospitalization in the United States.[1] Despite its clinical toll, the management of this disease has evolved markedly. Expanding vaccination programs, efforts to improve timeliness of antibiotic therapy, and improved processes of care are but a few developments that have improved outcomes for patients afflicted with this illness.[2, 3]
Use of peripherally inserted central catheters (PICCs) is an example of a modern development in the management of patients with pneumonia.[4, 5, 6, 7] PICCs provide many of the benefits associated with central venous catheters (CVCs) including reliable venous access for delivery of antibiotics, phlebotomy, and invasive hemodynamic monitoring. However, as they are placed in veins of the upper extremity, PICCs bypass insertion risks (eg, injury to the carotid vessels or pneumothorax) associated with placement of traditional CVCs.[8] Because they offer durable venous access, PICCs also facilitate care transitions while continuing intravenous antimicrobial therapy in patients with pneumonia.
However, accumulating evidence also suggests that PICCs are associated with important complications, including central lineassociated bloodstream infectionand venous thromboembolism.[9, 10] Furthermore, knowledge gaps in clinicians regarding indications for appropriate use and management of complications associated with PICCs have been recognized.[10, 11] These elements are problematic because reports of unjustified and inappropriate PICC use are growing in the literature.[12, 13] Such concerns have prompted a number of policy calls to improve PICC use, including Choosing Wisely recommendations by various professional societies.[14, 15]
As little is known about the prevalence or patterns of PICC use in adults hospitalized with pneumonia, we conducted a retrospective cohort study using data from a large network of US hospitals.
METHODS
Setting and Participants
We included patients from hospitals that participated in Premier's inpatient dataset, a large, fee‐supported, multipayer administrative database that has been used extensively in health services research to measure quality of care and comparative effectiveness of interventions.[16] Participating hospitals represent all regions of the United States and include teaching and nonteaching facilities in rural and urban locations. In addition to variables found in the uniform billing form, the Premier inpatient database also includes a date‐stamped list of charges for procedures conducted during hospitalization such as PICC placement. As PICC‐specific data are not available in most nationally representative datasets, Premier offers unique insights into utilization, timing, and factors associated with use of PICCs in hospitalized settings.
We included adult patients aged 18 years who were (1) admitted with a principal diagnosis of pneumonia present on admission, or secondary diagnosis of pneumonia if paired with a principal diagnosis of sepsis, respiratory failure, or influenza; (2) received at least 1 day of antibiotics between July 1, 2007 and November 30, 2011, and (3) underwent chest x‐ray or computed tomography (CT) at the time of admission. International Classification of Disease, 9th Revision, Clinical Modification (ICD‐9‐CM) codes were used for patient selection. Patients who were not admitted (eg, observation cases), had cystic fibrosis, or marked as pneumonia not present on admission were excluded. For patients who had more than 1 hospitalization during the study period, a single admission was randomly selected for inclusion.
Patient, Physician, and Hospital Data
For all patients, age, gender, marital status, insurance, race, and ethnicity were captured. Using software provided by the Healthcare Costs and Utilization Project, we categorized information on 29 comorbid conditions and computed a combined comorbidity score as described by Gagne et al.[17] Patients were considered to have healthcare‐associated pneumonia (HCAP) if they were: (1) admitted from a skilled nursing or a long‐term care facility, (2) hospitalized in the previous 90 days, (3) on dialysis, or (4) receiving immunosuppressing medications (eg, chemotherapy or steroids equivalent to at least 20 mg of prednisone per day) at the time of admission. Information on specialty of the admitting physician and hospital characteristics (eg, size, location, teaching status) were sourced through Premier data.
Receipt of PICCs and Related Therapies
Among eligible adult patients hospitalized with pneumonia, we identified patients who received a PICC at any time during hospitalization via PICC‐specific billing codes. Non‐PICC devices (eg, midlines, Hickman catheters) were not included. For all insertions, we assessed day of PICC placement relative to admission date. Data on type of PICC (eg, power‐injection capable, antibiotic coating) or PICC characteristics (size, number of lumens) were not available. We used billing codes to assess use of invasive or noninvasive ventilation, vasopressors, and administration of pneumonia‐specific antibiotics (eg, ‐lactams, macrolides, fluoroquinolones). Early exposure was defined when a billing code appeared within 2 days of hospital admission.
Outcomes of Interest
The primary outcome of interest was receipt of a PICC. Additionally, we assessed factors associated with PICC placement and variation in risk‐standardized rates of PICC use between hospitals.
Statistical Analyses
Patient and hospital characteristics were summarized using frequencies for categorical variables and medians with interquartile ranges for continuous variables. We examined association of individual patient and hospital characteristics with use of PICCs using generalized estimating equations models with a logit link for categorical variables and identity link for continuous variables, accounting for patient clustering within hospitals.
| Characteristic | Total, No. (%) | No PICC, No. (%) | PICC, No. (%) | P Value* |
|---|---|---|---|---|
| ||||
| 545,250 (100) | 503,401 (92.3) | 41,849 (7.7) | ||
| Demographics | ||||
| Age, median (Q1Q3), y | 71 (5782) | 72 (5782) | 69 (5780) | <0.001 |
| Gender | <0.001 | |||
| Male | 256,448 (47.0) | 237,232 (47.1) | 19,216 (45.9) | |
| Female | 288,802 (53.0) | 266,169 (52.9) | 22,633 (54.1) | |
| Race/ethnicity | <0.001 | |||
| White | 377,255 (69.2) | 346,689 (68.9) | 30,566 (73.0) | |
| Black | 63,345 (11.6) | 58,407 (11.6) | 4,938 (11.8) | |
| Hispanic | 22,855 (4.2) | 21,716 (4.3) | 1,139 (2.7) | |
| Other | 81,795 (15.0) | 76,589 (15.2) | 5,206 (12.4) | |
| Admitting specialty | <0.001 | |||
| Internal medicine | 236,859 (43.4) | 218,689 (43.4) | 18,170 (43.4) | |
| Hospital medicine | 116,499 (21.4) | 107,671 (21.4) | 8,828 (21.1) | |
| Family practice | 80,388 (14.7) | 75,482 (15.0) | 4,906 (11.7) | |
| Critical care and pulmonary | 35,670 (6.5) | 30,529 (6.1) | 41,849 (12.3) | |
| Geriatrics | 4,812 (0.9) | 4,098 (0.8) | 714 (1.7) | |
| Other | 71,022 (13.0) | 66,932 (13.3) | 4,090 (9.8) | |
| Insurance | <0.001 | |||
| Medicare | 370,303 (67.9) | 341,379 (67.8) | 28,924 (69.1) | |
| Medicaid | 45,505 (8.3) | 41,100 (8.2) | 4,405 (10.5) | |
| Managed care | 69,984 (12.8) | 65,280 (13.0) | 4,704 (11.2) | |
| Commercialindemnity | 20,672 (3.8) | 19,251 (3.8) | 1,421 (3.4) | |
| Other | 38,786 (7.1) | 36,391 (7.2) | 2,395 (5.7) | |
| Comorbidities | ||||
| Gagne combined comorbidity score, median (Q1Q3) | 2 (15) | 2 (14) | 4 (26) | <0.001 |
| Hypertension | 332,347 (60.9) | 306,964 (61.0) | 25,383 (60.7) | 0.13 |
| Chronic pulmonary disease | 255,403 (46.8) | 234,619 (46.6) | 20,784 (49.7) | <0.001 |
| Diabetes | 171,247 (31.4) | 155,540 (30.9) | 15,707 (37.5) | <0.001 |
| Congestive heart failure | 146,492 (26.9) | 131,041 (26.0) | 15,451 (36.9) | <0.001 |
| Atrial fibrillation | 108,405 (19.9) | 97,124 (19.3) | 11,281 (27.0) | <0.001 |
| Renal failure | 104,404 (19.1) | 94,277 (18.7) | 10,127 (24.2) | <0.001 |
| Nicotine replacement therapy/tobacco use | 89,938 (16.5) | 83,247 (16.5) | 6,691 (16.0) | <0.001 |
| Obesity | 60,242 (11.0) | 53,268 (10.6) | 6,974 (16.7) | <0.001 |
| Coagulopathy | 41,717 (7.6) | 35,371 (7.0) | 6,346 (15.2) | <0.001 |
| Prior stroke (1 year) | 26,787 (4.9) | 24,046 (4.78) | 2,741 (6.55) | <0.001 |
| Metastatic cancer | 21,868 (4.0) | 20,244 (4.0) | 1,624 (3.9) | 0.16 |
| Solid tumor w/out metastasis | 21,083 (3.9) | 19,380 (3.8) | 1,703 (4.1) | 0.12 |
| Prior VTE (1 year) | 19,090 (3.5) | 16,906 (3.4) | 2,184 (5.2) | <0.001 |
| Chronic liver disease | 16,273 (3.0) | 14,207 (2.8) | 2,066 (4.9) | <0.001 |
| Prior bacteremia (1 year) | 4,106 (0.7) | 3,584 (0.7) | 522 (1.2) | <0.001 |
| Nephrotic syndrome | 671 (0.1) | 607 (0.1) | 64 (0.2) | 0.03 |
| Morbidity markers | ||||
| Type of pneumonia | <0.001 | |||
| CAP | 376,370 (69.1) | 352,900 (70.1) | 23,830 (56.9) | |
| HCAP | 168,520 (30.9) | 150,501 (29.9) | 18,019 (43.1) | |
| Sepsis present on admission | 114,578 (21.0) | 96,467 (19.2) | 18,111 (43.3) | <0.001 |
| Non‐invasive ventilation | 47,913(8.8) | 40,599 (8.1) | 7,314 (17.5) | <0.001 |
| Invasive mechanical ventilation | 56,179 (10.3) | 44,228 (8.8) | 11,951 (28.6) | <0.001 |
| ICU status | 97,703 (17.9) | 80,380 (16.0) | 17,323 (41.4) | <0.001 |
| Vasopressor use | 48,353 (8.9) | 38,030 (7.6) | 10,323 (24.7) | <0.001 |
| Antibiotic/medication use | ||||
| Anti‐MRSA agent (vancomycin) | 146,068 (26.8) | 123,327 (24.5) | 22,741 (54.3) | <0.001 |
| Third‐generation cephalosporin | 250,782 (46.0) | 235,556 (46.8) | 15,226 (36.4) | <0.001 |
| Anti‐Pseudomonal cephalosporin | 41,798 (7.7) | 36,982 (7.3) | 4,816 (11.5) | <0.001 |
| Anti‐Pseudomonal ‐lactam | 122,215 (22.4) | 105,741 (21.0) | 16,474 (39.4) | <0.001 |
| Fluroquinolone | 288,051 (52.8) | 267,131 (53.1) | 20,920 (50.0) | <0.001 |
| Macrolide | 223,737 (41.0) | 210,954 (41.9) | 12,783 (30.5) | <0.001 |
| Aminoglycoside | 15,415 (2.8) | 12,661 (2.5) | 2,754 (6.6) | <0.001 |
| Oral steroids | 44,486 (8.2) | 41,586 (8.3) | 2,900 (6.9) | <0.001 |
| Intravenous steroids | 146,308 (26.8) | 133,920 (26.6) | 12,388 (29.6) | <0.001 |
| VTE prophylaxis with LMWH | 190,735 (35.0) | 174,612 (34.7) | 16,123 (38.5) | 0.01 |
| Discharge disposition | ||||
| Home | 282,146 (51.7) | 272,604(54.1) | 9,542 (22.8) | <0.001 |
| Home with home health | 71,977 (13.2) | 65,289 (13.0) | 6,688 (16.0) | <0.001 |
| Skilled nursing facility | 111,541 (20.5) | 97,113 (19.3) | 14,428 (34.5) | <0.001 |
| Hospice | 20,428 (3.7) | 17,902 (3.6) | 2,526 (6.0) | <0.001 |
| Expired | 47,733 (8.7) | 40,768 (8.1) | 6,965 (16.6) | <0.001 |
| Other | 11,425 (2.1) | 9,725 (1.9) | 1,700 (4.1) | <0.001 |
We then developed a multivariable hierarchical generalized linear model (HGLM) for PICC placement with a random effect for hospital. In this model, we included patient demographics, comorbidities, sepsis on admission, type of pneumonia (eg, HCAP vs community‐associated pneumonia [CAP]), admitting physician specialty, and indicators for early receipt of specific treatments such as guideline‐recommended antibiotics, vasopressors, ventilation (invasive or noninvasive), and pneumatic compression devices for prophylaxis of deep vein thrombosis.
To understand and estimate between‐hospital variation in PICC use, we calculated risk‐standardized rates of PICC use (RSPICC) across hospitals using HGLM methods. These methods are also employed by the Centers for Medicare and Medicaid Services to calculate risk‐standardized measures for public reporting.[18] Because hospital rates of PICC use were highly skewed (21.2% [n = 105] of hospitals had no patients with PICCs), we restricted this model to the 343 hospitals that had at least 5 patients with a PICC to obtain stable estimates. For each hospital, we estimated a predicted rate of PICC use (pPICC) as the sum of predicted probabilities of PICC receipt from patient factors and the random intercept for hospital in which they were admitted. We then calculated an expected rate of PICC use (ePICC) per hospital as the sum of expected probabilities of PICC receipt from patient factors only. RSPICC for each hospital was then computed as the product of the overall unadjusted mean PICC rate (PICC) from all patients and the ratio of the predicted to expected PICC rate (uPICC*[pPICC/ePICC]).[19] Kruskal‐Wallis tests were used to evaluate the association between hospital characteristics with RSPICC rates. To evaluate the impact of the hospital in variation in PICC use, we assessed the change in likelihood ratio of a hierarchical model with hospital random effects compared to a logistic regression model with patient factors only. In addition, we estimated the intraclass correlation (ICC) to assess the proportion of variation in PICC use associated with the hospital, and the median odds ratio (MOR) from the hierarchical model. The MOR is the median of a set of odds ratios comparing 2 patients with the same set of characteristics treated at 2 randomly selected hospitals.[20, 21, 22] All analyses were performed using the Statistical Analysis System version 9.3 (SAS Institute, Inc., Cary, NC) and Stata 13 (StataCorp Inc., College Station, TX).
Ethical and Regulatory Oversight
Permission to conduct this study was obtained from the institutional review board at Baystate Medical Center, Springfield, Massachusetts. The study did not qualify as human subjects research and made use of fully deidentified data.
RESULTS
Between July 2007 and November 2011, 634,285 admissions representing 545,250 unique patients from 495 hospitals met eligibility criteria and were included in the study (Figure 1). Included patients had a median age of 71 years (interquartile range [IQR]: 5782), and 53.0% were female. Most patients were Caucasian (69.2%), unmarried (51.6%), and insured by Medicare (67.9%). Patients were admitted to the hospital by internal medicine providers (43.4%), hospitalists (21.4%), and family practice providers (14.7%); notably, critical care and pulmonary medicine providers admitted 6.5% of patients. The median Gagne comorbidity score was 2 (IQR: 15). Hypertension, chronic obstructive pulmonary disease, diabetes, and congestive heart failure were among the most common comorbidities observed (Table 1).
Among eligible patients, 41,849 (7.7%) received a PICC during hospitalization. Approximately a quarter of all patients who received PICCs did so by hospital day 2; 90% underwent insertion by hospital day 11 (mean = 5.4 days, median = 4 days). Patients who received PICCs were younger (median IQR: 69 years, 5780 years) but otherwise demographically similar to those that did not receive PICCs (median IQR: 72 years, 5782 years). Compared to other specialties, patients admitted by critical care/pulmonary providers were twice as likely to receive PICCs (12.3% vs 6.1%, P < .001). Patients who received PICCs had higher comorbidity scores than those who did not (median Gagne comorbidity score 4 vs 2, P < 0.001) and were more likely to be diagnosed with HCAP (43.1% vs 29.9%, P < 0.001) than CAP (56.9% vs 70.1%, P < 0.001).
PICC recipients were also more likely to receive intensive care unit (ICU) level of care (41.4% vs 16%, P < 0.001) and both noninvasive (17.5% vs 8.1%, P < 0.001) and invasive ventilation (28.6% vs 8.8%, P < 0.001) upon admission. Vasopressor use was also significantly more frequent in patients who received PICCs (24.7% vs 7.6%, P < 0.001) compared to those who did not receive these devices. Patients with PICCs were more often discharged to skilled nursing facilities (34.5% vs 19.3%) than those without PICCs.
Characteristics Associated With PICC Use Following Multivariable Modeling
Using HGLM with a random hospital effect, multiple patient characteristics were associated with PICC use (Table 2). Patients 65 years of age were less likely to receive a PICC compared to younger patients (odds ratio [OR]: 0.81, 95% confidence interval [CI]: 0.79‐0.84). Weight loss (OR: 2.03, 95% CI: 1.97‐2.10), sepsis on admission (OR: 1.80, 95% CI: 1.75‐1.85), and ICU status on hospital day 1 or 2 (OR: 1.70, 95% CI: 1.64‐1.75) represented 3 factors most strongly associated with PICC use.
| Patient Characteristic | Odds Ratio | 95% Confidence Intervals |
|---|---|---|
| ||
| Age group (>66 vs 65 years) | 0.82 | 0.790.84 |
| Race/ethnicity | ||
| Other | 1.02 | 0.971.06 |
| Black | 0.99 | 0.951.03 |
| Hispanic | 0.82 | 0.760.88 |
| White | Referent | |
| Marital status | ||
| Other/missing | 1.07 | 1.011.14 |
| Single | 1.02 | 1.001.05 |
| Married | Referent | |
| Insurance payor | ||
| Other | 0.85 | 0.800.89 |
| Medicaid | 1.13 | 1.081.18 |
| Managed care | 0.95 | 0.910.99 |
| Commercialindemnity | 0.93 | 0.871.00 |
| Medicare | Referent | |
| Admitting physician specialty | ||
| Pulmonary/critical care medicine | 1.18 | 1.131.24 |
| Family practice | 1.01 | 0.971.05 |
| Geriatric medicine (FP and IM) | 1.85 | 1.662.05 |
| Hospitalist | 0.94 | 0.910.98 |
| Other specialties | 1.02 | 0.971.06 |
| Internal medicine | Referent | |
| Comorbidities | ||
| Congestive heart failure | 1.27 | 1.241.31 |
| Valvular disease | 1.11 | 1.071.15 |
| Pulmonary circulation disorders | 1.37 | 1.321.42 |
| Peripheral vascular disease | 1.09 | 1.051.13 |
| Hypertension | 0.94 | 0.920.97 |
| Paralysis | 1.59 | 1.511.67 |
| Other neurological disorders | 1.20 | 1.161.23 |
| Chronic lung disease | 1.10 | 1.071.12 |
| Diabetes | 1.13 | 1.101.16 |
| Hypothyroidism | 1.03 | 1.001.06 |
| Liver disease | 1.16 | 1.101.23 |
| Ulcer | 1.86 | 1.153.02 |
| Lymphoma | 0.88 | 0.810.96 |
| Metastatic cancer | 0.75 | 0.710.80 |
| Solid tumor without metastasis | 0.93 | 0.880.98 |
| Arthritis | 1.22 | 1.161.28 |
| Obesity | 1.47 | 1.421.52 |
| Weight loss | 2.03 | 1.972.10 |
| Blood loss | 1.69 | 1.551.85 |
| Deficiency anemias | 1.40 | 1.371.44 |
| Alcohol abuse | 1.19 | 1.131.26 |
| Drug abuse | 1.31 | 1.231.39 |
| Psychoses | 1.16 | 1.111.21 |
| Depression | 1.10 | 1.061.13 |
| Renal failure | 0.96 | 0.930.98 |
| Type of pneumonia | ||
| HCAP | 1.03 | 1.011.06 |
| CAP | Referent | |
| Sepsis (POA) | 1.80 | 1.751.85 |
| Antibiotic exposure | ||
| Anti‐MRSA agent | 1.72 | 1.671.76 |
| Anti‐Pseudomonal carbapenem | 1.37 | 1.311.44 |
| Non‐Pseudomonal carbapenem | 1.48 | 1.331.66 |
| Third‐generation cephalosporin | 1.04 | 1.011.07 |
| Anti‐Pseudomonal cephalosporin | 1.25 | 1.201.30 |
| Anti‐Pseudomonal ‐lactam | 1.27 | 1.231.31 |
| Aztreonam | 1.31 | 1.231.40 |
| Non‐Pseudomonal ‐lactam | 1.36 | 1.231.50 |
| ‐lactam | 1.55 | 1.261.90 |
| Respiratory quinolone | 0.90 | 0.870.92 |
| Macrolide | 0.85 | 0.820.88 |
| Doxycycline | 0.94 | 0.871.01 |
| Aminoglycoside | 1.21 | 1.141.27 |
| Vasopressors | 1.06 | 1.031.10 |
| Noninvasive ventilation | 1.29 | 1.251.34 |
| Invasive ventilation | 1.66 | 1.611.72 |
| Intensive care unit on admission | 1.70 | 1.641.75 |
| Atrial fibrillation | 1.26 | 1.221.29 |
| Upper extremity chronic DVT | 1.61 | 1.132.28 |
| Nicotine replacement therapy/tobacco abuse | 0.91 | 0.880.94 |
| Aspirin | 0.94 | 0.920.97 |
| Warfarin | 0.90 | 0.860.94 |
| LMWH, prophylactic dose | 1.10 | 1.081.13 |
| LMWH, treatment dose | 1.22 | 1.161.29 |
| Intravenous steroids | 1.05 | 1.021.08 |
| Bacteremia (prior year) | 1.14 | 1.021.27 |
| VTE (prior year) | 1.11 | 1.061.18 |
| Pneumatic compression device | 1.25 | 1.081.45 |
| Invasive ventilation (prior year) | 1.17 | 1.111.24 |
| Irritable bowel disease | 1.19 | 1.051.36 |
Therapy with potent parenteral antimicrobials including antimethicillin‐resistant Staphylococcus aureus agents (OR: 1.72, 95% CI: 1.67‐1.76), antipseudomonal ‐lactamases (OR: 1.27, 95% CI: 1.23‐1.31), and carbapenems (OR: 1.37, 95% CI: 1.31‐1.44) were significantly associated with PICC use. Conversely, use of macrolides (OR: 0.85, 95% CI: 0.82‐0.88) or respiratory fluoroquinolones (OR: 0.90, 95% CI: 0.87‐0.92) were associated with lower likelihood of PICC use. After adjusting for antimicrobial therapy, HCAP was only slightly more likely to result in PICC use than CAP (OR: 1.03, 95% CI: 1.01‐1.06). Compared to internal medicine providers, admission by geriatricians and critical care/pulmonary specialists was associated with greater likelihood of PICC use (OR: 1.85, 95% CI: 1.66‐2.05 and OR: 1.18, 95% CI: =1.13‐1.24, respectively). Admission by hospitalists was associated with a modestly lower likelihood of PICC placement (OR: 0.94, 95% CI: 0.91‐0.98).
Hospital Level Variation in PICC Use
To ensure stable estimates of hospital PICC use, we excluded 152 facilities (31%): 10% had no patients with PICCs and 21% had <5 patients who received a PICC. Therefore, RSPICC was estimated for 343 of 495 facilities (69%) (Figure 2). In these facilities, RSPICC varied from 0.3% to 41.7%. Hospital RSPICC was significantly associated with hospital location (median 11.9% vs 7.8% for urban vs rural hospitals respectively, P = 0.05). RSPICCs were also greater among hospitals in Southern (11.3%), Western (12.7%), and Midwest (12.0%) regions of the nation compared to those in the Northeast (8.4%) (P = 0.02) (Table 3).
| Hospital Characteristic (No.) | Median (IQR), % | P Value |
|---|---|---|
| ||
| Bed size | 0.12 | |
| 200 beds (106) | 9.1 (4.816.3) | |
| 201 beds (237) | 11.6 (5.817.6) | |
| Rural/urban | 0.05 | |
| Urban (275) | 11.9 (5.517.4) | |
| Rural (68) | 7.8 (5.014.0) | |
| Region | 0.02 | |
| Northeast (50) | 8.4 (3.913.0) | |
| Midwest (69) | 12.0 (5.817.4) | |
| West (57) | 12.7 (7.617.0) | |
| South (167) | 11.3 (4.817.8) | |
| Teaching status | 0.77 | |
| Nonteaching (246) | 10.9 (5.017.4) | |
| Teaching (97) | 12.0 (5.816.9) | |
A likelihood ratio test comparing the hierarchical model to a logistic model with patient factors only was highly significant (P < 0.001), indicating that the hospital where the patient was treated had a major impact on receipt of PICC after accounting for patient factors. The MOR was 2.71, which is a larger effect than we found for any of the individual patient characteristics. The proportion of variance explained by hospitals was 25% (95% CI: 22%‐28%), as measured by the ICC.
DISCUSSION
In this study of 545,250 adults hospitalized with pneumonia, we found that approximately 8% of patients received a PICC. Patients who received PICCs had more comorbidities, were more frequently diagnosed with HCAP, and were more often admitted to the ICU, where they experienced greater rates of mechanical ventilation, noninvasive ventilation, and vasopressor use compared to those who did not receive a PICC. Additionally, risk‐adjusted rates of PICC use varied as much as 10‐fold across institutions. In fact, almost 70% of the total variation in rates of PICC use remained unexplained by hospital or patient characteristics. Although use of PICCs is often clinically nuanced in ways that are difficult to capture in large datasets (eg, difficult venous access or inability to tolerate oral medications), the substantial variation of PICC use observed suggests that physician and institutional practice styles are the major determinants of PICC placement during a hospitalization for pneumonia. Because PICCs are associated with serious complications, and evidence regarding discretionary use is accumulating, a research agenda examining reasons for such use and related outcomes appears necessary.
The placement of PICCs has grown substantially in hospitalized patients all over the world.[23, 24] Although originally developed for total parenteral nutrition in surgical patients,[25] contemporary reports of PICC use in critical illness,[26] diseases such as cystic fibrosis,[27] and even pregnancy[28] are now common. Although PICCs are clinically invaluable in many of these conditions, growing use of these devices has led to the realization that benefits may be offset by complications.[9, 10, 29, 30] Additionally, recent data suggest that not all PICCs may be used for appropriate reasons. For instance, in a decade‐long study at a tertiary care center, changes in patterns of PICC use including shortened dwell times, multiple insertions in a single patient, and unclear indications for use were reported.[11] In another study at an academic medical center, a substantial proportion of PICCs were found to be idle or unjustified.[12] It comes as little surprise, then, that a recent multicenter study found that 1 out of every 5 clinicians did not even know that their patient had a PICC.[29] Although calls to improve PICC use in the hospital setting have emerged, strategies to do so are limited by data that emanate from single‐center reports or retrospective designs. No other studies reporting use of PICCs across US hospitals for any clinical condition currently exist.[31]
We found that patients with weight loss, those with greater combined comorbidity scores, and those who were critically ill or diagnosed with sepsis were more likely to receive PICCs than others. These observations suggest that PICC use may reflect underlying severity of illness, as advanced care such as ventilator support was often associated with PICC use. Additionally, discharge to a skilled nursing facility was frequently associated with PICC placement, a finding consistent with a recent study evaluating the use of PICCs in these settings.[32] However, a substantial proportion of PICC use remained unexplained by available patient or hospital factors. Although our study was not specifically designed to examine this question, a possible reason may relate to unmeasured institutional factors that influence the propensity to use a PICC, recently termed as PICC culture.[33] For example, it is plausible that hospitals with nursing‐led PICC teams or interventional radiology (such as teaching hospitals) are more likely to use PICCs than those without such operators. This hypothesis may explain why urban, larger, and teaching hospitals exhibited higher rates of PICC use. Conversely, providers may have an affinity toward PICC use that is predicated not just by operator availability, but also local hospital norms. Understanding why some facilities use PICCs at higher rates than others and implications of such variation with respect to patient safety, cost, and outcomes is important. Study designs that use mixed‐methods approaches or seek to qualitatively understand reasons behind PICC use are likely to be valuable in this enquiry.
Our study has limitations. First, we used an administrative dataset and ICD‐9‐CM codes rather than clinical data from medical records to identify cases of pneumonia or comorbidities. Our estimates of PICC use across hospitals thus may not fully account for differences in severity of illness, and it is possible that patients needed a PICC for reasons that we could not observe. However, the substantial variation observed in rates of PICC use across hospitals is unlikely to be explained by differences in patient severity of illness, documentation, or coding practices. Second, as PICC removal codes were not available, we are unable to comment on how often hospitalized pneumonia patients were discharged with PICCs or received antimicrobial therapy beyond their inpatient stay. Third, although we observed that a number of patient and hospital factors were associated with PICC receipt, our study was not designed to determine the reasons underlying these patterns.
These limitations aside, our study has important strengths. To our knowledge, this is the first study to report utilization and outcomes associated with PICC use among those hospitalized with pneumonia across the United States. The inclusion of a large number of patients receiving care in diverse facilities lends a high degree of external validity to our findings. Second, we used advanced modeling to identify factors associated with PICC use in hospitalized patients with pneumonia, producing innovative and novel findings. Third, our study is the first to show the existence of substantial variation in rates of PICC use across US hospitals within the single disease state of pneumonia. Understanding the drivers of this variability is important as it may inform future studies, policies, and practices to improve PICC use in hospitalized patients.
In conclusion, we found that PICC use in patients hospitalized with pneumonia is common and highly variable. Future studies examining the contextual factors behind PICC use and their association with outcomes are needed to facilitate efforts to standardize PICC use across hospitals.
Disclosures
Dr. Chopra is supported by a career development award (1‐K08‐HS022835‐01) from the Agency of Healthcare Research and Quality. The authors report no conflicts of interest.
Pneumonia is the most common cause of unplanned hospitalization in the United States.[1] Despite its clinical toll, the management of this disease has evolved markedly. Expanding vaccination programs, efforts to improve timeliness of antibiotic therapy, and improved processes of care are but a few developments that have improved outcomes for patients afflicted with this illness.[2, 3]
Use of peripherally inserted central catheters (PICCs) is an example of a modern development in the management of patients with pneumonia.[4, 5, 6, 7] PICCs provide many of the benefits associated with central venous catheters (CVCs) including reliable venous access for delivery of antibiotics, phlebotomy, and invasive hemodynamic monitoring. However, as they are placed in veins of the upper extremity, PICCs bypass insertion risks (eg, injury to the carotid vessels or pneumothorax) associated with placement of traditional CVCs.[8] Because they offer durable venous access, PICCs also facilitate care transitions while continuing intravenous antimicrobial therapy in patients with pneumonia.
However, accumulating evidence also suggests that PICCs are associated with important complications, including central lineassociated bloodstream infectionand venous thromboembolism.[9, 10] Furthermore, knowledge gaps in clinicians regarding indications for appropriate use and management of complications associated with PICCs have been recognized.[10, 11] These elements are problematic because reports of unjustified and inappropriate PICC use are growing in the literature.[12, 13] Such concerns have prompted a number of policy calls to improve PICC use, including Choosing Wisely recommendations by various professional societies.[14, 15]
As little is known about the prevalence or patterns of PICC use in adults hospitalized with pneumonia, we conducted a retrospective cohort study using data from a large network of US hospitals.
METHODS
Setting and Participants
We included patients from hospitals that participated in Premier's inpatient dataset, a large, fee‐supported, multipayer administrative database that has been used extensively in health services research to measure quality of care and comparative effectiveness of interventions.[16] Participating hospitals represent all regions of the United States and include teaching and nonteaching facilities in rural and urban locations. In addition to variables found in the uniform billing form, the Premier inpatient database also includes a date‐stamped list of charges for procedures conducted during hospitalization such as PICC placement. As PICC‐specific data are not available in most nationally representative datasets, Premier offers unique insights into utilization, timing, and factors associated with use of PICCs in hospitalized settings.
We included adult patients aged 18 years who were (1) admitted with a principal diagnosis of pneumonia present on admission, or secondary diagnosis of pneumonia if paired with a principal diagnosis of sepsis, respiratory failure, or influenza; (2) received at least 1 day of antibiotics between July 1, 2007 and November 30, 2011, and (3) underwent chest x‐ray or computed tomography (CT) at the time of admission. International Classification of Disease, 9th Revision, Clinical Modification (ICD‐9‐CM) codes were used for patient selection. Patients who were not admitted (eg, observation cases), had cystic fibrosis, or marked as pneumonia not present on admission were excluded. For patients who had more than 1 hospitalization during the study period, a single admission was randomly selected for inclusion.
Patient, Physician, and Hospital Data
For all patients, age, gender, marital status, insurance, race, and ethnicity were captured. Using software provided by the Healthcare Costs and Utilization Project, we categorized information on 29 comorbid conditions and computed a combined comorbidity score as described by Gagne et al.[17] Patients were considered to have healthcare‐associated pneumonia (HCAP) if they were: (1) admitted from a skilled nursing or a long‐term care facility, (2) hospitalized in the previous 90 days, (3) on dialysis, or (4) receiving immunosuppressing medications (eg, chemotherapy or steroids equivalent to at least 20 mg of prednisone per day) at the time of admission. Information on specialty of the admitting physician and hospital characteristics (eg, size, location, teaching status) were sourced through Premier data.
Receipt of PICCs and Related Therapies
Among eligible adult patients hospitalized with pneumonia, we identified patients who received a PICC at any time during hospitalization via PICC‐specific billing codes. Non‐PICC devices (eg, midlines, Hickman catheters) were not included. For all insertions, we assessed day of PICC placement relative to admission date. Data on type of PICC (eg, power‐injection capable, antibiotic coating) or PICC characteristics (size, number of lumens) were not available. We used billing codes to assess use of invasive or noninvasive ventilation, vasopressors, and administration of pneumonia‐specific antibiotics (eg, ‐lactams, macrolides, fluoroquinolones). Early exposure was defined when a billing code appeared within 2 days of hospital admission.
Outcomes of Interest
The primary outcome of interest was receipt of a PICC. Additionally, we assessed factors associated with PICC placement and variation in risk‐standardized rates of PICC use between hospitals.
Statistical Analyses
Patient and hospital characteristics were summarized using frequencies for categorical variables and medians with interquartile ranges for continuous variables. We examined association of individual patient and hospital characteristics with use of PICCs using generalized estimating equations models with a logit link for categorical variables and identity link for continuous variables, accounting for patient clustering within hospitals.
| Characteristic | Total, No. (%) | No PICC, No. (%) | PICC, No. (%) | P Value* |
|---|---|---|---|---|
| ||||
| 545,250 (100) | 503,401 (92.3) | 41,849 (7.7) | ||
| Demographics | ||||
| Age, median (Q1Q3), y | 71 (5782) | 72 (5782) | 69 (5780) | <0.001 |
| Gender | <0.001 | |||
| Male | 256,448 (47.0) | 237,232 (47.1) | 19,216 (45.9) | |
| Female | 288,802 (53.0) | 266,169 (52.9) | 22,633 (54.1) | |
| Race/ethnicity | <0.001 | |||
| White | 377,255 (69.2) | 346,689 (68.9) | 30,566 (73.0) | |
| Black | 63,345 (11.6) | 58,407 (11.6) | 4,938 (11.8) | |
| Hispanic | 22,855 (4.2) | 21,716 (4.3) | 1,139 (2.7) | |
| Other | 81,795 (15.0) | 76,589 (15.2) | 5,206 (12.4) | |
| Admitting specialty | <0.001 | |||
| Internal medicine | 236,859 (43.4) | 218,689 (43.4) | 18,170 (43.4) | |
| Hospital medicine | 116,499 (21.4) | 107,671 (21.4) | 8,828 (21.1) | |
| Family practice | 80,388 (14.7) | 75,482 (15.0) | 4,906 (11.7) | |
| Critical care and pulmonary | 35,670 (6.5) | 30,529 (6.1) | 41,849 (12.3) | |
| Geriatrics | 4,812 (0.9) | 4,098 (0.8) | 714 (1.7) | |
| Other | 71,022 (13.0) | 66,932 (13.3) | 4,090 (9.8) | |
| Insurance | <0.001 | |||
| Medicare | 370,303 (67.9) | 341,379 (67.8) | 28,924 (69.1) | |
| Medicaid | 45,505 (8.3) | 41,100 (8.2) | 4,405 (10.5) | |
| Managed care | 69,984 (12.8) | 65,280 (13.0) | 4,704 (11.2) | |
| Commercialindemnity | 20,672 (3.8) | 19,251 (3.8) | 1,421 (3.4) | |
| Other | 38,786 (7.1) | 36,391 (7.2) | 2,395 (5.7) | |
| Comorbidities | ||||
| Gagne combined comorbidity score, median (Q1Q3) | 2 (15) | 2 (14) | 4 (26) | <0.001 |
| Hypertension | 332,347 (60.9) | 306,964 (61.0) | 25,383 (60.7) | 0.13 |
| Chronic pulmonary disease | 255,403 (46.8) | 234,619 (46.6) | 20,784 (49.7) | <0.001 |
| Diabetes | 171,247 (31.4) | 155,540 (30.9) | 15,707 (37.5) | <0.001 |
| Congestive heart failure | 146,492 (26.9) | 131,041 (26.0) | 15,451 (36.9) | <0.001 |
| Atrial fibrillation | 108,405 (19.9) | 97,124 (19.3) | 11,281 (27.0) | <0.001 |
| Renal failure | 104,404 (19.1) | 94,277 (18.7) | 10,127 (24.2) | <0.001 |
| Nicotine replacement therapy/tobacco use | 89,938 (16.5) | 83,247 (16.5) | 6,691 (16.0) | <0.001 |
| Obesity | 60,242 (11.0) | 53,268 (10.6) | 6,974 (16.7) | <0.001 |
| Coagulopathy | 41,717 (7.6) | 35,371 (7.0) | 6,346 (15.2) | <0.001 |
| Prior stroke (1 year) | 26,787 (4.9) | 24,046 (4.78) | 2,741 (6.55) | <0.001 |
| Metastatic cancer | 21,868 (4.0) | 20,244 (4.0) | 1,624 (3.9) | 0.16 |
| Solid tumor w/out metastasis | 21,083 (3.9) | 19,380 (3.8) | 1,703 (4.1) | 0.12 |
| Prior VTE (1 year) | 19,090 (3.5) | 16,906 (3.4) | 2,184 (5.2) | <0.001 |
| Chronic liver disease | 16,273 (3.0) | 14,207 (2.8) | 2,066 (4.9) | <0.001 |
| Prior bacteremia (1 year) | 4,106 (0.7) | 3,584 (0.7) | 522 (1.2) | <0.001 |
| Nephrotic syndrome | 671 (0.1) | 607 (0.1) | 64 (0.2) | 0.03 |
| Morbidity markers | ||||
| Type of pneumonia | <0.001 | |||
| CAP | 376,370 (69.1) | 352,900 (70.1) | 23,830 (56.9) | |
| HCAP | 168,520 (30.9) | 150,501 (29.9) | 18,019 (43.1) | |
| Sepsis present on admission | 114,578 (21.0) | 96,467 (19.2) | 18,111 (43.3) | <0.001 |
| Non‐invasive ventilation | 47,913(8.8) | 40,599 (8.1) | 7,314 (17.5) | <0.001 |
| Invasive mechanical ventilation | 56,179 (10.3) | 44,228 (8.8) | 11,951 (28.6) | <0.001 |
| ICU status | 97,703 (17.9) | 80,380 (16.0) | 17,323 (41.4) | <0.001 |
| Vasopressor use | 48,353 (8.9) | 38,030 (7.6) | 10,323 (24.7) | <0.001 |
| Antibiotic/medication use | ||||
| Anti‐MRSA agent (vancomycin) | 146,068 (26.8) | 123,327 (24.5) | 22,741 (54.3) | <0.001 |
| Third‐generation cephalosporin | 250,782 (46.0) | 235,556 (46.8) | 15,226 (36.4) | <0.001 |
| Anti‐Pseudomonal cephalosporin | 41,798 (7.7) | 36,982 (7.3) | 4,816 (11.5) | <0.001 |
| Anti‐Pseudomonal ‐lactam | 122,215 (22.4) | 105,741 (21.0) | 16,474 (39.4) | <0.001 |
| Fluroquinolone | 288,051 (52.8) | 267,131 (53.1) | 20,920 (50.0) | <0.001 |
| Macrolide | 223,737 (41.0) | 210,954 (41.9) | 12,783 (30.5) | <0.001 |
| Aminoglycoside | 15,415 (2.8) | 12,661 (2.5) | 2,754 (6.6) | <0.001 |
| Oral steroids | 44,486 (8.2) | 41,586 (8.3) | 2,900 (6.9) | <0.001 |
| Intravenous steroids | 146,308 (26.8) | 133,920 (26.6) | 12,388 (29.6) | <0.001 |
| VTE prophylaxis with LMWH | 190,735 (35.0) | 174,612 (34.7) | 16,123 (38.5) | 0.01 |
| Discharge disposition | ||||
| Home | 282,146 (51.7) | 272,604(54.1) | 9,542 (22.8) | <0.001 |
| Home with home health | 71,977 (13.2) | 65,289 (13.0) | 6,688 (16.0) | <0.001 |
| Skilled nursing facility | 111,541 (20.5) | 97,113 (19.3) | 14,428 (34.5) | <0.001 |
| Hospice | 20,428 (3.7) | 17,902 (3.6) | 2,526 (6.0) | <0.001 |
| Expired | 47,733 (8.7) | 40,768 (8.1) | 6,965 (16.6) | <0.001 |
| Other | 11,425 (2.1) | 9,725 (1.9) | 1,700 (4.1) | <0.001 |
We then developed a multivariable hierarchical generalized linear model (HGLM) for PICC placement with a random effect for hospital. In this model, we included patient demographics, comorbidities, sepsis on admission, type of pneumonia (eg, HCAP vs community‐associated pneumonia [CAP]), admitting physician specialty, and indicators for early receipt of specific treatments such as guideline‐recommended antibiotics, vasopressors, ventilation (invasive or noninvasive), and pneumatic compression devices for prophylaxis of deep vein thrombosis.
To understand and estimate between‐hospital variation in PICC use, we calculated risk‐standardized rates of PICC use (RSPICC) across hospitals using HGLM methods. These methods are also employed by the Centers for Medicare and Medicaid Services to calculate risk‐standardized measures for public reporting.[18] Because hospital rates of PICC use were highly skewed (21.2% [n = 105] of hospitals had no patients with PICCs), we restricted this model to the 343 hospitals that had at least 5 patients with a PICC to obtain stable estimates. For each hospital, we estimated a predicted rate of PICC use (pPICC) as the sum of predicted probabilities of PICC receipt from patient factors and the random intercept for hospital in which they were admitted. We then calculated an expected rate of PICC use (ePICC) per hospital as the sum of expected probabilities of PICC receipt from patient factors only. RSPICC for each hospital was then computed as the product of the overall unadjusted mean PICC rate (PICC) from all patients and the ratio of the predicted to expected PICC rate (uPICC*[pPICC/ePICC]).[19] Kruskal‐Wallis tests were used to evaluate the association between hospital characteristics with RSPICC rates. To evaluate the impact of the hospital in variation in PICC use, we assessed the change in likelihood ratio of a hierarchical model with hospital random effects compared to a logistic regression model with patient factors only. In addition, we estimated the intraclass correlation (ICC) to assess the proportion of variation in PICC use associated with the hospital, and the median odds ratio (MOR) from the hierarchical model. The MOR is the median of a set of odds ratios comparing 2 patients with the same set of characteristics treated at 2 randomly selected hospitals.[20, 21, 22] All analyses were performed using the Statistical Analysis System version 9.3 (SAS Institute, Inc., Cary, NC) and Stata 13 (StataCorp Inc., College Station, TX).
Ethical and Regulatory Oversight
Permission to conduct this study was obtained from the institutional review board at Baystate Medical Center, Springfield, Massachusetts. The study did not qualify as human subjects research and made use of fully deidentified data.
RESULTS
Between July 2007 and November 2011, 634,285 admissions representing 545,250 unique patients from 495 hospitals met eligibility criteria and were included in the study (Figure 1). Included patients had a median age of 71 years (interquartile range [IQR]: 5782), and 53.0% were female. Most patients were Caucasian (69.2%), unmarried (51.6%), and insured by Medicare (67.9%). Patients were admitted to the hospital by internal medicine providers (43.4%), hospitalists (21.4%), and family practice providers (14.7%); notably, critical care and pulmonary medicine providers admitted 6.5% of patients. The median Gagne comorbidity score was 2 (IQR: 15). Hypertension, chronic obstructive pulmonary disease, diabetes, and congestive heart failure were among the most common comorbidities observed (Table 1).
Among eligible patients, 41,849 (7.7%) received a PICC during hospitalization. Approximately a quarter of all patients who received PICCs did so by hospital day 2; 90% underwent insertion by hospital day 11 (mean = 5.4 days, median = 4 days). Patients who received PICCs were younger (median IQR: 69 years, 5780 years) but otherwise demographically similar to those that did not receive PICCs (median IQR: 72 years, 5782 years). Compared to other specialties, patients admitted by critical care/pulmonary providers were twice as likely to receive PICCs (12.3% vs 6.1%, P < .001). Patients who received PICCs had higher comorbidity scores than those who did not (median Gagne comorbidity score 4 vs 2, P < 0.001) and were more likely to be diagnosed with HCAP (43.1% vs 29.9%, P < 0.001) than CAP (56.9% vs 70.1%, P < 0.001).
PICC recipients were also more likely to receive intensive care unit (ICU) level of care (41.4% vs 16%, P < 0.001) and both noninvasive (17.5% vs 8.1%, P < 0.001) and invasive ventilation (28.6% vs 8.8%, P < 0.001) upon admission. Vasopressor use was also significantly more frequent in patients who received PICCs (24.7% vs 7.6%, P < 0.001) compared to those who did not receive these devices. Patients with PICCs were more often discharged to skilled nursing facilities (34.5% vs 19.3%) than those without PICCs.
Characteristics Associated With PICC Use Following Multivariable Modeling
Using HGLM with a random hospital effect, multiple patient characteristics were associated with PICC use (Table 2). Patients 65 years of age were less likely to receive a PICC compared to younger patients (odds ratio [OR]: 0.81, 95% confidence interval [CI]: 0.79‐0.84). Weight loss (OR: 2.03, 95% CI: 1.97‐2.10), sepsis on admission (OR: 1.80, 95% CI: 1.75‐1.85), and ICU status on hospital day 1 or 2 (OR: 1.70, 95% CI: 1.64‐1.75) represented 3 factors most strongly associated with PICC use.
| Patient Characteristic | Odds Ratio | 95% Confidence Intervals |
|---|---|---|
| ||
| Age group (>66 vs 65 years) | 0.82 | 0.790.84 |
| Race/ethnicity | ||
| Other | 1.02 | 0.971.06 |
| Black | 0.99 | 0.951.03 |
| Hispanic | 0.82 | 0.760.88 |
| White | Referent | |
| Marital status | ||
| Other/missing | 1.07 | 1.011.14 |
| Single | 1.02 | 1.001.05 |
| Married | Referent | |
| Insurance payor | ||
| Other | 0.85 | 0.800.89 |
| Medicaid | 1.13 | 1.081.18 |
| Managed care | 0.95 | 0.910.99 |
| Commercialindemnity | 0.93 | 0.871.00 |
| Medicare | Referent | |
| Admitting physician specialty | ||
| Pulmonary/critical care medicine | 1.18 | 1.131.24 |
| Family practice | 1.01 | 0.971.05 |
| Geriatric medicine (FP and IM) | 1.85 | 1.662.05 |
| Hospitalist | 0.94 | 0.910.98 |
| Other specialties | 1.02 | 0.971.06 |
| Internal medicine | Referent | |
| Comorbidities | ||
| Congestive heart failure | 1.27 | 1.241.31 |
| Valvular disease | 1.11 | 1.071.15 |
| Pulmonary circulation disorders | 1.37 | 1.321.42 |
| Peripheral vascular disease | 1.09 | 1.051.13 |
| Hypertension | 0.94 | 0.920.97 |
| Paralysis | 1.59 | 1.511.67 |
| Other neurological disorders | 1.20 | 1.161.23 |
| Chronic lung disease | 1.10 | 1.071.12 |
| Diabetes | 1.13 | 1.101.16 |
| Hypothyroidism | 1.03 | 1.001.06 |
| Liver disease | 1.16 | 1.101.23 |
| Ulcer | 1.86 | 1.153.02 |
| Lymphoma | 0.88 | 0.810.96 |
| Metastatic cancer | 0.75 | 0.710.80 |
| Solid tumor without metastasis | 0.93 | 0.880.98 |
| Arthritis | 1.22 | 1.161.28 |
| Obesity | 1.47 | 1.421.52 |
| Weight loss | 2.03 | 1.972.10 |
| Blood loss | 1.69 | 1.551.85 |
| Deficiency anemias | 1.40 | 1.371.44 |
| Alcohol abuse | 1.19 | 1.131.26 |
| Drug abuse | 1.31 | 1.231.39 |
| Psychoses | 1.16 | 1.111.21 |
| Depression | 1.10 | 1.061.13 |
| Renal failure | 0.96 | 0.930.98 |
| Type of pneumonia | ||
| HCAP | 1.03 | 1.011.06 |
| CAP | Referent | |
| Sepsis (POA) | 1.80 | 1.751.85 |
| Antibiotic exposure | ||
| Anti‐MRSA agent | 1.72 | 1.671.76 |
| Anti‐Pseudomonal carbapenem | 1.37 | 1.311.44 |
| Non‐Pseudomonal carbapenem | 1.48 | 1.331.66 |
| Third‐generation cephalosporin | 1.04 | 1.011.07 |
| Anti‐Pseudomonal cephalosporin | 1.25 | 1.201.30 |
| Anti‐Pseudomonal ‐lactam | 1.27 | 1.231.31 |
| Aztreonam | 1.31 | 1.231.40 |
| Non‐Pseudomonal ‐lactam | 1.36 | 1.231.50 |
| ‐lactam | 1.55 | 1.261.90 |
| Respiratory quinolone | 0.90 | 0.870.92 |
| Macrolide | 0.85 | 0.820.88 |
| Doxycycline | 0.94 | 0.871.01 |
| Aminoglycoside | 1.21 | 1.141.27 |
| Vasopressors | 1.06 | 1.031.10 |
| Noninvasive ventilation | 1.29 | 1.251.34 |
| Invasive ventilation | 1.66 | 1.611.72 |
| Intensive care unit on admission | 1.70 | 1.641.75 |
| Atrial fibrillation | 1.26 | 1.221.29 |
| Upper extremity chronic DVT | 1.61 | 1.132.28 |
| Nicotine replacement therapy/tobacco abuse | 0.91 | 0.880.94 |
| Aspirin | 0.94 | 0.920.97 |
| Warfarin | 0.90 | 0.860.94 |
| LMWH, prophylactic dose | 1.10 | 1.081.13 |
| LMWH, treatment dose | 1.22 | 1.161.29 |
| Intravenous steroids | 1.05 | 1.021.08 |
| Bacteremia (prior year) | 1.14 | 1.021.27 |
| VTE (prior year) | 1.11 | 1.061.18 |
| Pneumatic compression device | 1.25 | 1.081.45 |
| Invasive ventilation (prior year) | 1.17 | 1.111.24 |
| Irritable bowel disease | 1.19 | 1.051.36 |
Therapy with potent parenteral antimicrobials including antimethicillin‐resistant Staphylococcus aureus agents (OR: 1.72, 95% CI: 1.67‐1.76), antipseudomonal ‐lactamases (OR: 1.27, 95% CI: 1.23‐1.31), and carbapenems (OR: 1.37, 95% CI: 1.31‐1.44) were significantly associated with PICC use. Conversely, use of macrolides (OR: 0.85, 95% CI: 0.82‐0.88) or respiratory fluoroquinolones (OR: 0.90, 95% CI: 0.87‐0.92) were associated with lower likelihood of PICC use. After adjusting for antimicrobial therapy, HCAP was only slightly more likely to result in PICC use than CAP (OR: 1.03, 95% CI: 1.01‐1.06). Compared to internal medicine providers, admission by geriatricians and critical care/pulmonary specialists was associated with greater likelihood of PICC use (OR: 1.85, 95% CI: 1.66‐2.05 and OR: 1.18, 95% CI: =1.13‐1.24, respectively). Admission by hospitalists was associated with a modestly lower likelihood of PICC placement (OR: 0.94, 95% CI: 0.91‐0.98).
Hospital Level Variation in PICC Use
To ensure stable estimates of hospital PICC use, we excluded 152 facilities (31%): 10% had no patients with PICCs and 21% had <5 patients who received a PICC. Therefore, RSPICC was estimated for 343 of 495 facilities (69%) (Figure 2). In these facilities, RSPICC varied from 0.3% to 41.7%. Hospital RSPICC was significantly associated with hospital location (median 11.9% vs 7.8% for urban vs rural hospitals respectively, P = 0.05). RSPICCs were also greater among hospitals in Southern (11.3%), Western (12.7%), and Midwest (12.0%) regions of the nation compared to those in the Northeast (8.4%) (P = 0.02) (Table 3).
| Hospital Characteristic (No.) | Median (IQR), % | P Value |
|---|---|---|
| ||
| Bed size | 0.12 | |
| 200 beds (106) | 9.1 (4.816.3) | |
| 201 beds (237) | 11.6 (5.817.6) | |
| Rural/urban | 0.05 | |
| Urban (275) | 11.9 (5.517.4) | |
| Rural (68) | 7.8 (5.014.0) | |
| Region | 0.02 | |
| Northeast (50) | 8.4 (3.913.0) | |
| Midwest (69) | 12.0 (5.817.4) | |
| West (57) | 12.7 (7.617.0) | |
| South (167) | 11.3 (4.817.8) | |
| Teaching status | 0.77 | |
| Nonteaching (246) | 10.9 (5.017.4) | |
| Teaching (97) | 12.0 (5.816.9) | |
A likelihood ratio test comparing the hierarchical model to a logistic model with patient factors only was highly significant (P < 0.001), indicating that the hospital where the patient was treated had a major impact on receipt of PICC after accounting for patient factors. The MOR was 2.71, which is a larger effect than we found for any of the individual patient characteristics. The proportion of variance explained by hospitals was 25% (95% CI: 22%‐28%), as measured by the ICC.
DISCUSSION
In this study of 545,250 adults hospitalized with pneumonia, we found that approximately 8% of patients received a PICC. Patients who received PICCs had more comorbidities, were more frequently diagnosed with HCAP, and were more often admitted to the ICU, where they experienced greater rates of mechanical ventilation, noninvasive ventilation, and vasopressor use compared to those who did not receive a PICC. Additionally, risk‐adjusted rates of PICC use varied as much as 10‐fold across institutions. In fact, almost 70% of the total variation in rates of PICC use remained unexplained by hospital or patient characteristics. Although use of PICCs is often clinically nuanced in ways that are difficult to capture in large datasets (eg, difficult venous access or inability to tolerate oral medications), the substantial variation of PICC use observed suggests that physician and institutional practice styles are the major determinants of PICC placement during a hospitalization for pneumonia. Because PICCs are associated with serious complications, and evidence regarding discretionary use is accumulating, a research agenda examining reasons for such use and related outcomes appears necessary.
The placement of PICCs has grown substantially in hospitalized patients all over the world.[23, 24] Although originally developed for total parenteral nutrition in surgical patients,[25] contemporary reports of PICC use in critical illness,[26] diseases such as cystic fibrosis,[27] and even pregnancy[28] are now common. Although PICCs are clinically invaluable in many of these conditions, growing use of these devices has led to the realization that benefits may be offset by complications.[9, 10, 29, 30] Additionally, recent data suggest that not all PICCs may be used for appropriate reasons. For instance, in a decade‐long study at a tertiary care center, changes in patterns of PICC use including shortened dwell times, multiple insertions in a single patient, and unclear indications for use were reported.[11] In another study at an academic medical center, a substantial proportion of PICCs were found to be idle or unjustified.[12] It comes as little surprise, then, that a recent multicenter study found that 1 out of every 5 clinicians did not even know that their patient had a PICC.[29] Although calls to improve PICC use in the hospital setting have emerged, strategies to do so are limited by data that emanate from single‐center reports or retrospective designs. No other studies reporting use of PICCs across US hospitals for any clinical condition currently exist.[31]
We found that patients with weight loss, those with greater combined comorbidity scores, and those who were critically ill or diagnosed with sepsis were more likely to receive PICCs than others. These observations suggest that PICC use may reflect underlying severity of illness, as advanced care such as ventilator support was often associated with PICC use. Additionally, discharge to a skilled nursing facility was frequently associated with PICC placement, a finding consistent with a recent study evaluating the use of PICCs in these settings.[32] However, a substantial proportion of PICC use remained unexplained by available patient or hospital factors. Although our study was not specifically designed to examine this question, a possible reason may relate to unmeasured institutional factors that influence the propensity to use a PICC, recently termed as PICC culture.[33] For example, it is plausible that hospitals with nursing‐led PICC teams or interventional radiology (such as teaching hospitals) are more likely to use PICCs than those without such operators. This hypothesis may explain why urban, larger, and teaching hospitals exhibited higher rates of PICC use. Conversely, providers may have an affinity toward PICC use that is predicated not just by operator availability, but also local hospital norms. Understanding why some facilities use PICCs at higher rates than others and implications of such variation with respect to patient safety, cost, and outcomes is important. Study designs that use mixed‐methods approaches or seek to qualitatively understand reasons behind PICC use are likely to be valuable in this enquiry.
Our study has limitations. First, we used an administrative dataset and ICD‐9‐CM codes rather than clinical data from medical records to identify cases of pneumonia or comorbidities. Our estimates of PICC use across hospitals thus may not fully account for differences in severity of illness, and it is possible that patients needed a PICC for reasons that we could not observe. However, the substantial variation observed in rates of PICC use across hospitals is unlikely to be explained by differences in patient severity of illness, documentation, or coding practices. Second, as PICC removal codes were not available, we are unable to comment on how often hospitalized pneumonia patients were discharged with PICCs or received antimicrobial therapy beyond their inpatient stay. Third, although we observed that a number of patient and hospital factors were associated with PICC receipt, our study was not designed to determine the reasons underlying these patterns.
These limitations aside, our study has important strengths. To our knowledge, this is the first study to report utilization and outcomes associated with PICC use among those hospitalized with pneumonia across the United States. The inclusion of a large number of patients receiving care in diverse facilities lends a high degree of external validity to our findings. Second, we used advanced modeling to identify factors associated with PICC use in hospitalized patients with pneumonia, producing innovative and novel findings. Third, our study is the first to show the existence of substantial variation in rates of PICC use across US hospitals within the single disease state of pneumonia. Understanding the drivers of this variability is important as it may inform future studies, policies, and practices to improve PICC use in hospitalized patients.
In conclusion, we found that PICC use in patients hospitalized with pneumonia is common and highly variable. Future studies examining the contextual factors behind PICC use and their association with outcomes are needed to facilitate efforts to standardize PICC use across hospitals.
Disclosures
Dr. Chopra is supported by a career development award (1‐K08‐HS022835‐01) from the Agency of Healthcare Research and Quality. The authors report no conflicts of interest.
- , . Reasons for being admitted to the hospital through the emergency department, 2003. Healthcare Cost and Utilization Project Statistical Brief 2. Rockville, MD: Agency for Healthcare Research and Quality. Available at: www.hcup‐us.ahrq.gov/reports/statbriefs/sb2.pdf. Published February 2006. Accessed June 27, 2014.
- , , , et al. National patterns of risk‐standardized mortality and readmission after hospitalization for acute myocardial infarction, heart failure, and pneumonia: update on publicly reported outcomes measures based on the 2013 release. J Gen Intern Med. 2014;29(10):1333–1340.
- , , , et al. Quality of care for elderly patients hospitalized for pneumonia in the United States, 2006 to 2010. JAMA Intern Med. 2014;174(11):1806–1814.
- , . PICC lines: the latest home care challenge. RN. 1990;53(1):44–51.
- , , , . Peripherally inserted central catheters in an acute‐care hospital. Arch Intern Med. 1994;154(16):1833–1837.
- , . The peripherally inserted central catheter: a retrospective look at three years of insertions. J Intraven Nurs. 1993;16(2):92–103.
- , , , . Peripherally inserted central catheters in general medicine. Mayo Clin Proc. 1997;72(3):225–233.
- , , . Two‐year trends of peripherally inserted central catheter‐line complications at a tertiary‐care hospital: role of nursing expertise. Infect Control Hosp Epidemiol. 2001;22(6):377–379.
- , , , , , . PICC‐associated bloodstream infections: prevalence, patterns, and predictors. Am J Med. 2014;127(4):319–328.
- , , , , . The risk of bloodstream infection associated with peripherally inserted central catheters compared with central venous catheters in adults: a systematic review and meta‐analysis. Infect Control Hosp Epidemiol. 2013;34(9):908–918.
- , , , , , . Peripherally inserted central catheters: use at a tertiary care pediatric center. J Vasc Interv Radiol. 2013;24(9):1323–1331.
- , , , et al. Temporary central venous catheter utilization patterns in a large tertiary care center: tracking the “idle central venous catheter”. Infect Control Hosp Epidemiol. 2012;33(1):50–57.
- , , , , . Inappropriate intravascular device use: a prospective study. Journal Hosp Infect. 2011;78(2):128–132.
- , , , et al. Enhancing patient‐centered care: SGIM and choosing wisely. J Gen Intern Med. 2014;29(3):432–433.
- , , , et al. Critical and honest conversations: the evidence behind the “Choosing Wisely” campaign recommendations by the American Society of Nephrology. Clin J Am Soc Nephrol. 2012;7(10):1664–1672.
- , , , et al. Using highly detailed administrative data to predict pneumonia mortality. PLoS One. 2014;9(1):e87382.
- , , , , . A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–759.
- , , , , . Hospitals with the highest intensive care utilization provide lower quality pneumonia care to the elderly. Crit Care Med. 2015;43(6):1178–1186.
- , . Statistical and clinical aspects of hospital outcomes profiling. Stat Sci. 2007;22(2):206–226.
- , . Appropriate assessment of neighborhood effects on individual health: integrating random and fixed effects in multilevel logistic regression. Am J Epidemiol. 2005;161(1):81–88.
- , , , . Interpreting parameters in the logistic regression model with random effects. Biometrics. 2000;56(3):909–914.
- , , , . Hospital‐level associations with 30‐day patient mortality after cardiac surgery: a tutorial on the application and interpretation of marginal and multilevel logistic regression. BMC Med Res Methodol. 2012;12:28.
- , , , . Experiences of the first PICC team in the Czech Republic. Br J Nurs. 2015;24(suppl 2):S4–S10.
- , , , et al. Greece reports prototype intervention with first peripherally inserted central catheter: case report and literature review. J Vasc Nurs. 2012;30(3):88–93.
- Total intravenous nutrition with peripherally inserted silicone elastomer central venous catheters. Arch Surg. 1975;110(5):644–646.
- , . Focus on peripherally inserted central catheters in critically ill patients. World J Crit Care Med. 2014;3(4):80–94.
- , , , et al. Quality improvement initiative to reduce deep vein thrombosis associated with peripherally inserted central catheters in adults with cystic fibrosis. Ann Am Thorac Soc. 2014;11(9):1404–1410.
- , , , . Peripherally Inserted central catheter (PICC) complications during pregnancy. JPEN J Parenter Enteral Nutr. 2013;38(5):595–601.
- , , , et al. Do clinicians know which of their patients have central venous catheters?: a multicenter observational study. Ann Intern Med. 2014;161(8):562–567.
- , , , et al. Risk of venous thromboembolism associated with peripherally inserted central catheters: a systematic review and meta‐analysis. Lancet. 2013;382(9889):311–325.
- , , . The problem with peripherally inserted central catheters. JAMA. 2012;308(15):1527–1528.
- , , , et al. Peripherally inserted central catheter use in skilled nursing facilities: a pilot study. J Am Geriatr Soc. 2015;63(9):1894–1899.
- , , , , . Inpatient venous access practices: PICC culture and the kidney patient. J Vasc Access. 2015;16(3):206–210.
- , . Reasons for being admitted to the hospital through the emergency department, 2003. Healthcare Cost and Utilization Project Statistical Brief 2. Rockville, MD: Agency for Healthcare Research and Quality. Available at: www.hcup‐us.ahrq.gov/reports/statbriefs/sb2.pdf. Published February 2006. Accessed June 27, 2014.
- , , , et al. National patterns of risk‐standardized mortality and readmission after hospitalization for acute myocardial infarction, heart failure, and pneumonia: update on publicly reported outcomes measures based on the 2013 release. J Gen Intern Med. 2014;29(10):1333–1340.
- , , , et al. Quality of care for elderly patients hospitalized for pneumonia in the United States, 2006 to 2010. JAMA Intern Med. 2014;174(11):1806–1814.
- , . PICC lines: the latest home care challenge. RN. 1990;53(1):44–51.
- , , , . Peripherally inserted central catheters in an acute‐care hospital. Arch Intern Med. 1994;154(16):1833–1837.
- , . The peripherally inserted central catheter: a retrospective look at three years of insertions. J Intraven Nurs. 1993;16(2):92–103.
- , , , . Peripherally inserted central catheters in general medicine. Mayo Clin Proc. 1997;72(3):225–233.
- , , . Two‐year trends of peripherally inserted central catheter‐line complications at a tertiary‐care hospital: role of nursing expertise. Infect Control Hosp Epidemiol. 2001;22(6):377–379.
- , , , , , . PICC‐associated bloodstream infections: prevalence, patterns, and predictors. Am J Med. 2014;127(4):319–328.
- , , , , . The risk of bloodstream infection associated with peripherally inserted central catheters compared with central venous catheters in adults: a systematic review and meta‐analysis. Infect Control Hosp Epidemiol. 2013;34(9):908–918.
- , , , , , . Peripherally inserted central catheters: use at a tertiary care pediatric center. J Vasc Interv Radiol. 2013;24(9):1323–1331.
- , , , et al. Temporary central venous catheter utilization patterns in a large tertiary care center: tracking the “idle central venous catheter”. Infect Control Hosp Epidemiol. 2012;33(1):50–57.
- , , , , . Inappropriate intravascular device use: a prospective study. Journal Hosp Infect. 2011;78(2):128–132.
- , , , et al. Enhancing patient‐centered care: SGIM and choosing wisely. J Gen Intern Med. 2014;29(3):432–433.
- , , , et al. Critical and honest conversations: the evidence behind the “Choosing Wisely” campaign recommendations by the American Society of Nephrology. Clin J Am Soc Nephrol. 2012;7(10):1664–1672.
- , , , et al. Using highly detailed administrative data to predict pneumonia mortality. PLoS One. 2014;9(1):e87382.
- , , , , . A combined comorbidity score predicted mortality in elderly patients better than existing scores. J Clin Epidemiol. 2011;64(7):749–759.
- , , , , . Hospitals with the highest intensive care utilization provide lower quality pneumonia care to the elderly. Crit Care Med. 2015;43(6):1178–1186.
- , . Statistical and clinical aspects of hospital outcomes profiling. Stat Sci. 2007;22(2):206–226.
- , . Appropriate assessment of neighborhood effects on individual health: integrating random and fixed effects in multilevel logistic regression. Am J Epidemiol. 2005;161(1):81–88.
- , , , . Interpreting parameters in the logistic regression model with random effects. Biometrics. 2000;56(3):909–914.
- , , , . Hospital‐level associations with 30‐day patient mortality after cardiac surgery: a tutorial on the application and interpretation of marginal and multilevel logistic regression. BMC Med Res Methodol. 2012;12:28.
- , , , . Experiences of the first PICC team in the Czech Republic. Br J Nurs. 2015;24(suppl 2):S4–S10.
- , , , et al. Greece reports prototype intervention with first peripherally inserted central catheter: case report and literature review. J Vasc Nurs. 2012;30(3):88–93.
- Total intravenous nutrition with peripherally inserted silicone elastomer central venous catheters. Arch Surg. 1975;110(5):644–646.
- , . Focus on peripherally inserted central catheters in critically ill patients. World J Crit Care Med. 2014;3(4):80–94.
- , , , et al. Quality improvement initiative to reduce deep vein thrombosis associated with peripherally inserted central catheters in adults with cystic fibrosis. Ann Am Thorac Soc. 2014;11(9):1404–1410.
- , , , . Peripherally Inserted central catheter (PICC) complications during pregnancy. JPEN J Parenter Enteral Nutr. 2013;38(5):595–601.
- , , , et al. Do clinicians know which of their patients have central venous catheters?: a multicenter observational study. Ann Intern Med. 2014;161(8):562–567.
- , , , et al. Risk of venous thromboembolism associated with peripherally inserted central catheters: a systematic review and meta‐analysis. Lancet. 2013;382(9889):311–325.
- , , . The problem with peripherally inserted central catheters. JAMA. 2012;308(15):1527–1528.
- , , , et al. Peripherally inserted central catheter use in skilled nursing facilities: a pilot study. J Am Geriatr Soc. 2015;63(9):1894–1899.
- , , , , . Inpatient venous access practices: PICC culture and the kidney patient. J Vasc Access. 2015;16(3):206–210.
Using Social Media as a Hospital QI Tool
Patient experience has become a major component of the Center for Medicare and Medicaid Services Value‐Based Purchasing initiative.[1] Hospitals have therefore focused quality improvement (QI) efforts on this area.[2] Hospital performance in the realm of patient experience is generally determined using systematic surveys with closed‐ended questions, but patient‐generated narrative feedback can help hospitals identify the components of care that contribute to patient satisfaction and or are in need of improvement.[3] Online narrative responses posted by patients on rating websites or social media have been criticized because they may not be representative of the population,[4] but they also have some advantages.[5] Any patient may leave a comment, not just those who are selected for a survey. Patients may also experience benefits through the act of sharing their story with others. Moreover, most US hospitals use some form of social media,[6] which they can theoretically use to self‐collect narrative data online. To realize the full potential of patient‐generated online narratives, we need a clearer understanding of the best practices for collecting and using these narratives. We therefore solicited patient feedback on the Facebook page of a large tertiary academic medical center to determine whether it is feasible to use social media platforms for learning about and improving hospital quality.
METHODS
Baystate Medical Center (BMC) is a tertiary care medical center in western Massachusetts. We identified key BMC stakeholders in the areas of QI and public affairs. Noting that patients have expressed interest in leaving comments via social media,[7] the group opted to perform a pilot study to obtain patient narratives via a Facebook prompt (Facebook is a social media site used by an estimated 58% of US adults[8]). The BMC public affairs department delivered a press release to the local media describing a 3‐week period during which patients were invited to leave narrative feedback on the BMC Facebook wall. The BMC Institutional Review Board deemed that this study did not constitute human subjects research.
During March 2014 (March 10, 2014March 24, 2014), we posted (once a week) an open‐ended prompt on BMC's Facebook wall. The prompt was designed to elicit novel descriptions of patient experience that could help to drive QI. It read: We want to hear about your experiences. In the comment section below, please tell us what we do well and how we can improve your care. Because of concerns about the potential reputational risks of allowing open feedback on a public social media page, the prompt also reminded patients of the social media ground rules: there should be no mention of specific physicians, nurses, or other caregivers by name (for liability reasons); and patients should not include details about their medical history (for privacy reasons).
We collected all posts to preserve comments and used directed qualitative content analysis to examine them.[9] Two research team members[3, 10, 11] independently coded the responses. Starting with an a priori codebook that was developed during a previous study,[3] they amended the codebook through an iterative process to incorporate new concepts. After independently coding all blocks of text, the coders reviewed their coding selections and resolved discrepancies through discussion. We then performed second‐level coding, in which codes were organized into major pertinent themes. We reviewed the coded text after applying secondary codes in order to check for accuracy of coding and theme assignment as well as completeness of second‐level coding. We calculated percent agreement, defined as both raters scoring a block of text with the same code divided by total number of codes. We also calculated the Spearman correlation between the 2 reviewers. We used descriptive statistics to assess the frequency of select codes and themes (see Supporting Information, Appendix 1 and Appendix 2, in the online version of this article).[9, 12, 13]
RESULTS
Over a 3‐week study period, 47 comments were submitted by 37 respondents. This yielded 148 codable statements (Table 1). Despite limited information on respondents, we ascertained from Facebook that 32 (86%) were women and 5 (14%) were men.
| Theme | Total Respondents, N (%) | % Positive | Positive Quotation | % Negative | Negative Quotation |
|---|---|---|---|---|---|
| |||||
| Staff | 17 (46) | 45% | The nurses in the pediatric unit, as well as the doctors in radiology and x‐ray department were AMAZING! | 55% | My 24‐year‐old daughter had to go for 5 days of IV treatmentwhile getting her infusion there was a fire alarm. She has a video showing the flashing of the light and the sound of the alarm and the closing of doors and NOT A SINGLE staff member to be found. Her infusions take about 2 hours. They set it and forget it. Luckily there wasn't a fire and someone did finally come to disconnect her. |
| Had a fabulous experience with Wesson women's this week! Had a C section and 3‐day admission. All staff from preoperative to inpatient were so helpful and really anticipated my needs before I could even ask for things. | My mother was hospitalized for at least 3 weeks right after the cardiovascular center openedwhen she went into cardiac arrest and in acute care and the step unit the care was great, very attentive nurses and doctors. When she was starting to recover and moved upstairs, downhill it went. She'd ring for assistance because she wanted to walk to the bathrooms and more times she was left to her own devices because no one would respond. | ||||
| Facility | 9 (24) | 25% | New buildings are beautiful and the new signs are way better. | 75% | The parking situation was disappointing and the waiting room was also very dirty. |
| I really like the individual pods in the ER. | I could have used a single room as my roommate was very annoying and demanding. | ||||
| Departments | 22 (60) | 44% | The NICU was great when my son was in there. The children's unit was great with my daughter and respected my needs. | 56% | Revamp maternity; it needs it desperately. |
| Labor and delivery was a great place. | Love Baystate but hate the ER. | ||||
| Technical aspects of care (eg, errors) | 9 (24) | 0 | 100% | Day 2 of my 24 year old getting her 2‐hour IV infusion....she was set up with her IV. When checked 2 hours later, the staff member was very upset to find that only the saline had run. She never opened the medication clamp. So now they gave her the medication in 1 hour instead of 2. | |
| If I had 1 suggestion it would be to re‐evaluate patient comfort when patients are waiting to be admitted. | |||||
From coded text, several broad themes were identified (see Table 1 for representative quotes): (1) comments about staff (17/37 respondents, 45.9%). These included positive descriptions of efficiency, caring behavior, good training, and good communication, whereas negative comments included descriptions of unfriendliness, apparent lack of caring, inattentiveness, poor training, unprofessional behavior, and poor communication; (2) comments about specific departments (22/37, 59.5%); (3) comments on technical aspects of care, including perceived errors, incorrect diagnoses, and inattention to pain control (9/37, 24.3%); and (4) comments describing the hospital physical plant, parking, and amenities (9/37, 24.3%). There were a few miscellaneous comments that did not fit into these broad themes, such as expressions of gratitude for our solicitation of narratives. Percent agreement between coders was 80% and Spearman's Rho was 0.82 (p<0.001).
A small number (n=3) of respondents repeatedly made comments over the 3‐week period, accounting for 30% (45/148) of codes. These repetitive commenters tended to dominate the Facebook conversation, at times describing the same experience more than once.
DISCUSSION
In this study evaluating the potential utility of social media as a hospital QI tool, several broad themes emerged. From these themes, we identified several areas that could be deemed as QI targets, including: training staff to be more responsive and sensitive to patients needs and concerns, improving patient and visitor parking, and reducing emergency department waiting times. However, the insight gained from solicited Facebook comments was similar to feedback gained from more traditional approaches of soliciting patient perspectives on care, such as patient experience surveys.[14]
Our findings should be viewed in the context of prior work focused on patient narratives in healthcare. Greaves et al. used sentiment analysis to describe the content of nearly 200,000 tweets (comments posted on the social networking website Twitter) sent to National Health Service (NHS) hospitals.[15] Themes were similar to those found in our study: (1) interaction with staff, (2) environment and facilities, and (3) issues of access and timeliness of service. Notably, these themes mirrored prior work examining narratives at NHS hospitals[3] and were similar to domains of commonly used surveys of patient experience.[14] The authors noted that there were issues with the signal to noise ratio (only about 10% of tweets were about quality) and the enforced brevity of Twitter (tweets must be 140 characters or less). These limitations suggest that using Twitter to identify QI targets would be difficult.
In contrast to Greaves et al., we chose to solicit feedback on our hospital's Facebook page. Facebook does not have Twitter's enforced brevity, allowing for more detailed narratives. In addition, we did not encounter the signal‐to‐noise problem, because our prompt was designed to request feedback that was relevant to recent experiences of care. However, a few respondents dominated the conversation, supporting the hypothesis that those most likely to comment may be the patients or families who have had the best or worst experiences. In the future, we will attempt to address this limitation and reduce the influence of repeat commenters by changing our prompt (eg, Please tell us about your experience, but please do not post descriptions of the same experience more than once.).
This pilot demonstrated some of the previously described benefits of online narratives.[5] First, there appears to be value in allowing patients to share their experiences and to read the experiences of others (as indicated in a few grateful patients comments). Second, soliciting online narratives offers a way for hospitals to demonstrate a commitment to transparency. Third, in contrast to closed‐ended survey questions, narrative comments help to identify why patients were satisfied or unsatisfied with their care. Although some surveys with closed‐ended questions also allow for narratives, these comments may or may not be carefully reviewed by the hospital. Using social media to solicit and respond to comments enhances existing methods for evaluating patient experience by engaging patients in a public space, which increases the likelihood that hospitals will attempt to improve care in response.
Notably, none of the identified areas for improvement could be considered novel QI targets for BMC. For example, our hospital has been very focused on training staff around patient experience, and emergency department wait times are the focus of a system‐wide improvement effort called Patient Progress.
This study has other limitations. We conducted this study over a 3‐week time period in a single center and on a single social media site whose members may not be representative of the overall patient population at BMC. Although we do not know how generalizable our findings are (in terms of identifying QI targets), we feel that we have demonstrated how using social media to collect data on patient experience is feasible and could be informative for other hospitals in other locations. It is possible that we did not allow the experiment to run long enough; a longer time or broader outreach (eg, a handout given to every discharged patient over a longer period) may be needed to allow patients adequate opportunity to comment. Of note, we did not specifically examine responses by time period, but it does seem, in post hoc analysis, that after 2 weeks of data collection we reached theoretical saturation with no new themes emerging in the third week (eg, third‐week comments included I heart your nurses. and Love Baystate but hate the ER.). More work is also needed that includes a broader range of social media platforms and more participating hospitals.
In conclusion, the opportunity to provide feedback on Facebook has the potential to engage and empower patients, and hospitals can use these online narratives to help to drive improvement efforts. Yet potential benefits must be weighed against reputational risks, a lack of representative respondents, and the paucity of novel QI targets obtained in this study.
Disclosures: Dr. Lagu is supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. The authors report no conflicts of interest.
- Centers for Medicare 47(2):193–219.
- , , , , . A mixed‐methods analysis of patient reviews of hospital care in England: implications for public reporting of health care quality data in the United States. Jt Comm J Qual Patient Saf. 2013;39(1):7–15.
- , , , , , , , . Taking Patients' Narratives about Clinicians from Anecdote to Science. NEJM. 2015;373(7):675–679.
- , . Putting the public back in public reporting of health care quality. JAMA. 2010;304(15):1711–1712.
- , , , et al. Use of social media across US hospitals: descriptive analysis of adoption and utilization. J Med Internet Res. 2014;16(11):e264.
- , , , , , . Patient use of email, Facebook, and physician websites to communicate with physicians: a national online survey of retail pharmacy users [published online June 24, 2015]. J Gen Intern Med. doi:10.1007/s11606-015-3374-7.
- Pew Research Center. Social networking fact sheet. Available at: http://www.pewinternet.org/fact‐sheets/social‐networking‐fact‐sheet. Accessed March 4, 2015.
- , . Three approaches to qualitative content analysis. Qual Health Res. 2005;15(9):1277–1288.
- , , , . Patients’ evaluations of health care providers in the era of social networking: an analysis of physician‐rating websites. J Gen Intern Med. 2010;25(9):942–946.
- , , , , . Vaccine counseling: a content analysis of patient‐physician discussions regarding human papilloma virus vaccine. Vaccine. 2011;29(43):7343–7349.
- . Qualitative research methods. Int J Qual Health Care. 2002;14(4):329–336.
- , . Doing Qualitative Research. Vol 2. Thousand Oaks, CA: Sage Publications; 1999.
- , , , . Patients’ perception of hospital care in the United States. N Engl J Med. 2008;359(18):1921–1931.
- , , , et al. Tweets about hospital quality: a mixed methods study. BMJ Qual Saf. 2014;23(10):838–846.
Patient experience has become a major component of the Center for Medicare and Medicaid Services Value‐Based Purchasing initiative.[1] Hospitals have therefore focused quality improvement (QI) efforts on this area.[2] Hospital performance in the realm of patient experience is generally determined using systematic surveys with closed‐ended questions, but patient‐generated narrative feedback can help hospitals identify the components of care that contribute to patient satisfaction and or are in need of improvement.[3] Online narrative responses posted by patients on rating websites or social media have been criticized because they may not be representative of the population,[4] but they also have some advantages.[5] Any patient may leave a comment, not just those who are selected for a survey. Patients may also experience benefits through the act of sharing their story with others. Moreover, most US hospitals use some form of social media,[6] which they can theoretically use to self‐collect narrative data online. To realize the full potential of patient‐generated online narratives, we need a clearer understanding of the best practices for collecting and using these narratives. We therefore solicited patient feedback on the Facebook page of a large tertiary academic medical center to determine whether it is feasible to use social media platforms for learning about and improving hospital quality.
METHODS
Baystate Medical Center (BMC) is a tertiary care medical center in western Massachusetts. We identified key BMC stakeholders in the areas of QI and public affairs. Noting that patients have expressed interest in leaving comments via social media,[7] the group opted to perform a pilot study to obtain patient narratives via a Facebook prompt (Facebook is a social media site used by an estimated 58% of US adults[8]). The BMC public affairs department delivered a press release to the local media describing a 3‐week period during which patients were invited to leave narrative feedback on the BMC Facebook wall. The BMC Institutional Review Board deemed that this study did not constitute human subjects research.
During March 2014 (March 10, 2014March 24, 2014), we posted (once a week) an open‐ended prompt on BMC's Facebook wall. The prompt was designed to elicit novel descriptions of patient experience that could help to drive QI. It read: We want to hear about your experiences. In the comment section below, please tell us what we do well and how we can improve your care. Because of concerns about the potential reputational risks of allowing open feedback on a public social media page, the prompt also reminded patients of the social media ground rules: there should be no mention of specific physicians, nurses, or other caregivers by name (for liability reasons); and patients should not include details about their medical history (for privacy reasons).
We collected all posts to preserve comments and used directed qualitative content analysis to examine them.[9] Two research team members[3, 10, 11] independently coded the responses. Starting with an a priori codebook that was developed during a previous study,[3] they amended the codebook through an iterative process to incorporate new concepts. After independently coding all blocks of text, the coders reviewed their coding selections and resolved discrepancies through discussion. We then performed second‐level coding, in which codes were organized into major pertinent themes. We reviewed the coded text after applying secondary codes in order to check for accuracy of coding and theme assignment as well as completeness of second‐level coding. We calculated percent agreement, defined as both raters scoring a block of text with the same code divided by total number of codes. We also calculated the Spearman correlation between the 2 reviewers. We used descriptive statistics to assess the frequency of select codes and themes (see Supporting Information, Appendix 1 and Appendix 2, in the online version of this article).[9, 12, 13]
RESULTS
Over a 3‐week study period, 47 comments were submitted by 37 respondents. This yielded 148 codable statements (Table 1). Despite limited information on respondents, we ascertained from Facebook that 32 (86%) were women and 5 (14%) were men.
| Theme | Total Respondents, N (%) | % Positive | Positive Quotation | % Negative | Negative Quotation |
|---|---|---|---|---|---|
| |||||
| Staff | 17 (46) | 45% | The nurses in the pediatric unit, as well as the doctors in radiology and x‐ray department were AMAZING! | 55% | My 24‐year‐old daughter had to go for 5 days of IV treatmentwhile getting her infusion there was a fire alarm. She has a video showing the flashing of the light and the sound of the alarm and the closing of doors and NOT A SINGLE staff member to be found. Her infusions take about 2 hours. They set it and forget it. Luckily there wasn't a fire and someone did finally come to disconnect her. |
| Had a fabulous experience with Wesson women's this week! Had a C section and 3‐day admission. All staff from preoperative to inpatient were so helpful and really anticipated my needs before I could even ask for things. | My mother was hospitalized for at least 3 weeks right after the cardiovascular center openedwhen she went into cardiac arrest and in acute care and the step unit the care was great, very attentive nurses and doctors. When she was starting to recover and moved upstairs, downhill it went. She'd ring for assistance because she wanted to walk to the bathrooms and more times she was left to her own devices because no one would respond. | ||||
| Facility | 9 (24) | 25% | New buildings are beautiful and the new signs are way better. | 75% | The parking situation was disappointing and the waiting room was also very dirty. |
| I really like the individual pods in the ER. | I could have used a single room as my roommate was very annoying and demanding. | ||||
| Departments | 22 (60) | 44% | The NICU was great when my son was in there. The children's unit was great with my daughter and respected my needs. | 56% | Revamp maternity; it needs it desperately. |
| Labor and delivery was a great place. | Love Baystate but hate the ER. | ||||
| Technical aspects of care (eg, errors) | 9 (24) | 0 | 100% | Day 2 of my 24 year old getting her 2‐hour IV infusion....she was set up with her IV. When checked 2 hours later, the staff member was very upset to find that only the saline had run. She never opened the medication clamp. So now they gave her the medication in 1 hour instead of 2. | |
| If I had 1 suggestion it would be to re‐evaluate patient comfort when patients are waiting to be admitted. | |||||
From coded text, several broad themes were identified (see Table 1 for representative quotes): (1) comments about staff (17/37 respondents, 45.9%). These included positive descriptions of efficiency, caring behavior, good training, and good communication, whereas negative comments included descriptions of unfriendliness, apparent lack of caring, inattentiveness, poor training, unprofessional behavior, and poor communication; (2) comments about specific departments (22/37, 59.5%); (3) comments on technical aspects of care, including perceived errors, incorrect diagnoses, and inattention to pain control (9/37, 24.3%); and (4) comments describing the hospital physical plant, parking, and amenities (9/37, 24.3%). There were a few miscellaneous comments that did not fit into these broad themes, such as expressions of gratitude for our solicitation of narratives. Percent agreement between coders was 80% and Spearman's Rho was 0.82 (p<0.001).
A small number (n=3) of respondents repeatedly made comments over the 3‐week period, accounting for 30% (45/148) of codes. These repetitive commenters tended to dominate the Facebook conversation, at times describing the same experience more than once.
DISCUSSION
In this study evaluating the potential utility of social media as a hospital QI tool, several broad themes emerged. From these themes, we identified several areas that could be deemed as QI targets, including: training staff to be more responsive and sensitive to patients needs and concerns, improving patient and visitor parking, and reducing emergency department waiting times. However, the insight gained from solicited Facebook comments was similar to feedback gained from more traditional approaches of soliciting patient perspectives on care, such as patient experience surveys.[14]
Our findings should be viewed in the context of prior work focused on patient narratives in healthcare. Greaves et al. used sentiment analysis to describe the content of nearly 200,000 tweets (comments posted on the social networking website Twitter) sent to National Health Service (NHS) hospitals.[15] Themes were similar to those found in our study: (1) interaction with staff, (2) environment and facilities, and (3) issues of access and timeliness of service. Notably, these themes mirrored prior work examining narratives at NHS hospitals[3] and were similar to domains of commonly used surveys of patient experience.[14] The authors noted that there were issues with the signal to noise ratio (only about 10% of tweets were about quality) and the enforced brevity of Twitter (tweets must be 140 characters or less). These limitations suggest that using Twitter to identify QI targets would be difficult.
In contrast to Greaves et al., we chose to solicit feedback on our hospital's Facebook page. Facebook does not have Twitter's enforced brevity, allowing for more detailed narratives. In addition, we did not encounter the signal‐to‐noise problem, because our prompt was designed to request feedback that was relevant to recent experiences of care. However, a few respondents dominated the conversation, supporting the hypothesis that those most likely to comment may be the patients or families who have had the best or worst experiences. In the future, we will attempt to address this limitation and reduce the influence of repeat commenters by changing our prompt (eg, Please tell us about your experience, but please do not post descriptions of the same experience more than once.).
This pilot demonstrated some of the previously described benefits of online narratives.[5] First, there appears to be value in allowing patients to share their experiences and to read the experiences of others (as indicated in a few grateful patients comments). Second, soliciting online narratives offers a way for hospitals to demonstrate a commitment to transparency. Third, in contrast to closed‐ended survey questions, narrative comments help to identify why patients were satisfied or unsatisfied with their care. Although some surveys with closed‐ended questions also allow for narratives, these comments may or may not be carefully reviewed by the hospital. Using social media to solicit and respond to comments enhances existing methods for evaluating patient experience by engaging patients in a public space, which increases the likelihood that hospitals will attempt to improve care in response.
Notably, none of the identified areas for improvement could be considered novel QI targets for BMC. For example, our hospital has been very focused on training staff around patient experience, and emergency department wait times are the focus of a system‐wide improvement effort called Patient Progress.
This study has other limitations. We conducted this study over a 3‐week time period in a single center and on a single social media site whose members may not be representative of the overall patient population at BMC. Although we do not know how generalizable our findings are (in terms of identifying QI targets), we feel that we have demonstrated how using social media to collect data on patient experience is feasible and could be informative for other hospitals in other locations. It is possible that we did not allow the experiment to run long enough; a longer time or broader outreach (eg, a handout given to every discharged patient over a longer period) may be needed to allow patients adequate opportunity to comment. Of note, we did not specifically examine responses by time period, but it does seem, in post hoc analysis, that after 2 weeks of data collection we reached theoretical saturation with no new themes emerging in the third week (eg, third‐week comments included I heart your nurses. and Love Baystate but hate the ER.). More work is also needed that includes a broader range of social media platforms and more participating hospitals.
In conclusion, the opportunity to provide feedback on Facebook has the potential to engage and empower patients, and hospitals can use these online narratives to help to drive improvement efforts. Yet potential benefits must be weighed against reputational risks, a lack of representative respondents, and the paucity of novel QI targets obtained in this study.
Disclosures: Dr. Lagu is supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. The authors report no conflicts of interest.
Patient experience has become a major component of the Center for Medicare and Medicaid Services Value‐Based Purchasing initiative.[1] Hospitals have therefore focused quality improvement (QI) efforts on this area.[2] Hospital performance in the realm of patient experience is generally determined using systematic surveys with closed‐ended questions, but patient‐generated narrative feedback can help hospitals identify the components of care that contribute to patient satisfaction and or are in need of improvement.[3] Online narrative responses posted by patients on rating websites or social media have been criticized because they may not be representative of the population,[4] but they also have some advantages.[5] Any patient may leave a comment, not just those who are selected for a survey. Patients may also experience benefits through the act of sharing their story with others. Moreover, most US hospitals use some form of social media,[6] which they can theoretically use to self‐collect narrative data online. To realize the full potential of patient‐generated online narratives, we need a clearer understanding of the best practices for collecting and using these narratives. We therefore solicited patient feedback on the Facebook page of a large tertiary academic medical center to determine whether it is feasible to use social media platforms for learning about and improving hospital quality.
METHODS
Baystate Medical Center (BMC) is a tertiary care medical center in western Massachusetts. We identified key BMC stakeholders in the areas of QI and public affairs. Noting that patients have expressed interest in leaving comments via social media,[7] the group opted to perform a pilot study to obtain patient narratives via a Facebook prompt (Facebook is a social media site used by an estimated 58% of US adults[8]). The BMC public affairs department delivered a press release to the local media describing a 3‐week period during which patients were invited to leave narrative feedback on the BMC Facebook wall. The BMC Institutional Review Board deemed that this study did not constitute human subjects research.
During March 2014 (March 10, 2014March 24, 2014), we posted (once a week) an open‐ended prompt on BMC's Facebook wall. The prompt was designed to elicit novel descriptions of patient experience that could help to drive QI. It read: We want to hear about your experiences. In the comment section below, please tell us what we do well and how we can improve your care. Because of concerns about the potential reputational risks of allowing open feedback on a public social media page, the prompt also reminded patients of the social media ground rules: there should be no mention of specific physicians, nurses, or other caregivers by name (for liability reasons); and patients should not include details about their medical history (for privacy reasons).
We collected all posts to preserve comments and used directed qualitative content analysis to examine them.[9] Two research team members[3, 10, 11] independently coded the responses. Starting with an a priori codebook that was developed during a previous study,[3] they amended the codebook through an iterative process to incorporate new concepts. After independently coding all blocks of text, the coders reviewed their coding selections and resolved discrepancies through discussion. We then performed second‐level coding, in which codes were organized into major pertinent themes. We reviewed the coded text after applying secondary codes in order to check for accuracy of coding and theme assignment as well as completeness of second‐level coding. We calculated percent agreement, defined as both raters scoring a block of text with the same code divided by total number of codes. We also calculated the Spearman correlation between the 2 reviewers. We used descriptive statistics to assess the frequency of select codes and themes (see Supporting Information, Appendix 1 and Appendix 2, in the online version of this article).[9, 12, 13]
RESULTS
Over a 3‐week study period, 47 comments were submitted by 37 respondents. This yielded 148 codable statements (Table 1). Despite limited information on respondents, we ascertained from Facebook that 32 (86%) were women and 5 (14%) were men.
| Theme | Total Respondents, N (%) | % Positive | Positive Quotation | % Negative | Negative Quotation |
|---|---|---|---|---|---|
| |||||
| Staff | 17 (46) | 45% | The nurses in the pediatric unit, as well as the doctors in radiology and x‐ray department were AMAZING! | 55% | My 24‐year‐old daughter had to go for 5 days of IV treatmentwhile getting her infusion there was a fire alarm. She has a video showing the flashing of the light and the sound of the alarm and the closing of doors and NOT A SINGLE staff member to be found. Her infusions take about 2 hours. They set it and forget it. Luckily there wasn't a fire and someone did finally come to disconnect her. |
| Had a fabulous experience with Wesson women's this week! Had a C section and 3‐day admission. All staff from preoperative to inpatient were so helpful and really anticipated my needs before I could even ask for things. | My mother was hospitalized for at least 3 weeks right after the cardiovascular center openedwhen she went into cardiac arrest and in acute care and the step unit the care was great, very attentive nurses and doctors. When she was starting to recover and moved upstairs, downhill it went. She'd ring for assistance because she wanted to walk to the bathrooms and more times she was left to her own devices because no one would respond. | ||||
| Facility | 9 (24) | 25% | New buildings are beautiful and the new signs are way better. | 75% | The parking situation was disappointing and the waiting room was also very dirty. |
| I really like the individual pods in the ER. | I could have used a single room as my roommate was very annoying and demanding. | ||||
| Departments | 22 (60) | 44% | The NICU was great when my son was in there. The children's unit was great with my daughter and respected my needs. | 56% | Revamp maternity; it needs it desperately. |
| Labor and delivery was a great place. | Love Baystate but hate the ER. | ||||
| Technical aspects of care (eg, errors) | 9 (24) | 0 | 100% | Day 2 of my 24 year old getting her 2‐hour IV infusion....she was set up with her IV. When checked 2 hours later, the staff member was very upset to find that only the saline had run. She never opened the medication clamp. So now they gave her the medication in 1 hour instead of 2. | |
| If I had 1 suggestion it would be to re‐evaluate patient comfort when patients are waiting to be admitted. | |||||
From coded text, several broad themes were identified (see Table 1 for representative quotes): (1) comments about staff (17/37 respondents, 45.9%). These included positive descriptions of efficiency, caring behavior, good training, and good communication, whereas negative comments included descriptions of unfriendliness, apparent lack of caring, inattentiveness, poor training, unprofessional behavior, and poor communication; (2) comments about specific departments (22/37, 59.5%); (3) comments on technical aspects of care, including perceived errors, incorrect diagnoses, and inattention to pain control (9/37, 24.3%); and (4) comments describing the hospital physical plant, parking, and amenities (9/37, 24.3%). There were a few miscellaneous comments that did not fit into these broad themes, such as expressions of gratitude for our solicitation of narratives. Percent agreement between coders was 80% and Spearman's Rho was 0.82 (p<0.001).
A small number (n=3) of respondents repeatedly made comments over the 3‐week period, accounting for 30% (45/148) of codes. These repetitive commenters tended to dominate the Facebook conversation, at times describing the same experience more than once.
DISCUSSION
In this study evaluating the potential utility of social media as a hospital QI tool, several broad themes emerged. From these themes, we identified several areas that could be deemed as QI targets, including: training staff to be more responsive and sensitive to patients needs and concerns, improving patient and visitor parking, and reducing emergency department waiting times. However, the insight gained from solicited Facebook comments was similar to feedback gained from more traditional approaches of soliciting patient perspectives on care, such as patient experience surveys.[14]
Our findings should be viewed in the context of prior work focused on patient narratives in healthcare. Greaves et al. used sentiment analysis to describe the content of nearly 200,000 tweets (comments posted on the social networking website Twitter) sent to National Health Service (NHS) hospitals.[15] Themes were similar to those found in our study: (1) interaction with staff, (2) environment and facilities, and (3) issues of access and timeliness of service. Notably, these themes mirrored prior work examining narratives at NHS hospitals[3] and were similar to domains of commonly used surveys of patient experience.[14] The authors noted that there were issues with the signal to noise ratio (only about 10% of tweets were about quality) and the enforced brevity of Twitter (tweets must be 140 characters or less). These limitations suggest that using Twitter to identify QI targets would be difficult.
In contrast to Greaves et al., we chose to solicit feedback on our hospital's Facebook page. Facebook does not have Twitter's enforced brevity, allowing for more detailed narratives. In addition, we did not encounter the signal‐to‐noise problem, because our prompt was designed to request feedback that was relevant to recent experiences of care. However, a few respondents dominated the conversation, supporting the hypothesis that those most likely to comment may be the patients or families who have had the best or worst experiences. In the future, we will attempt to address this limitation and reduce the influence of repeat commenters by changing our prompt (eg, Please tell us about your experience, but please do not post descriptions of the same experience more than once.).
This pilot demonstrated some of the previously described benefits of online narratives.[5] First, there appears to be value in allowing patients to share their experiences and to read the experiences of others (as indicated in a few grateful patients comments). Second, soliciting online narratives offers a way for hospitals to demonstrate a commitment to transparency. Third, in contrast to closed‐ended survey questions, narrative comments help to identify why patients were satisfied or unsatisfied with their care. Although some surveys with closed‐ended questions also allow for narratives, these comments may or may not be carefully reviewed by the hospital. Using social media to solicit and respond to comments enhances existing methods for evaluating patient experience by engaging patients in a public space, which increases the likelihood that hospitals will attempt to improve care in response.
Notably, none of the identified areas for improvement could be considered novel QI targets for BMC. For example, our hospital has been very focused on training staff around patient experience, and emergency department wait times are the focus of a system‐wide improvement effort called Patient Progress.
This study has other limitations. We conducted this study over a 3‐week time period in a single center and on a single social media site whose members may not be representative of the overall patient population at BMC. Although we do not know how generalizable our findings are (in terms of identifying QI targets), we feel that we have demonstrated how using social media to collect data on patient experience is feasible and could be informative for other hospitals in other locations. It is possible that we did not allow the experiment to run long enough; a longer time or broader outreach (eg, a handout given to every discharged patient over a longer period) may be needed to allow patients adequate opportunity to comment. Of note, we did not specifically examine responses by time period, but it does seem, in post hoc analysis, that after 2 weeks of data collection we reached theoretical saturation with no new themes emerging in the third week (eg, third‐week comments included I heart your nurses. and Love Baystate but hate the ER.). More work is also needed that includes a broader range of social media platforms and more participating hospitals.
In conclusion, the opportunity to provide feedback on Facebook has the potential to engage and empower patients, and hospitals can use these online narratives to help to drive improvement efforts. Yet potential benefits must be weighed against reputational risks, a lack of representative respondents, and the paucity of novel QI targets obtained in this study.
Disclosures: Dr. Lagu is supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number K01HL114745. The authors report no conflicts of interest.
- Centers for Medicare 47(2):193–219.
- , , , , . A mixed‐methods analysis of patient reviews of hospital care in England: implications for public reporting of health care quality data in the United States. Jt Comm J Qual Patient Saf. 2013;39(1):7–15.
- , , , , , , , . Taking Patients' Narratives about Clinicians from Anecdote to Science. NEJM. 2015;373(7):675–679.
- , . Putting the public back in public reporting of health care quality. JAMA. 2010;304(15):1711–1712.
- , , , et al. Use of social media across US hospitals: descriptive analysis of adoption and utilization. J Med Internet Res. 2014;16(11):e264.
- , , , , , . Patient use of email, Facebook, and physician websites to communicate with physicians: a national online survey of retail pharmacy users [published online June 24, 2015]. J Gen Intern Med. doi:10.1007/s11606-015-3374-7.
- Pew Research Center. Social networking fact sheet. Available at: http://www.pewinternet.org/fact‐sheets/social‐networking‐fact‐sheet. Accessed March 4, 2015.
- , . Three approaches to qualitative content analysis. Qual Health Res. 2005;15(9):1277–1288.
- , , , . Patients’ evaluations of health care providers in the era of social networking: an analysis of physician‐rating websites. J Gen Intern Med. 2010;25(9):942–946.
- , , , , . Vaccine counseling: a content analysis of patient‐physician discussions regarding human papilloma virus vaccine. Vaccine. 2011;29(43):7343–7349.
- . Qualitative research methods. Int J Qual Health Care. 2002;14(4):329–336.
- , . Doing Qualitative Research. Vol 2. Thousand Oaks, CA: Sage Publications; 1999.
- , , , . Patients’ perception of hospital care in the United States. N Engl J Med. 2008;359(18):1921–1931.
- , , , et al. Tweets about hospital quality: a mixed methods study. BMJ Qual Saf. 2014;23(10):838–846.
- Centers for Medicare 47(2):193–219.
- , , , , . A mixed‐methods analysis of patient reviews of hospital care in England: implications for public reporting of health care quality data in the United States. Jt Comm J Qual Patient Saf. 2013;39(1):7–15.
- , , , , , , , . Taking Patients' Narratives about Clinicians from Anecdote to Science. NEJM. 2015;373(7):675–679.
- , . Putting the public back in public reporting of health care quality. JAMA. 2010;304(15):1711–1712.
- , , , et al. Use of social media across US hospitals: descriptive analysis of adoption and utilization. J Med Internet Res. 2014;16(11):e264.
- , , , , , . Patient use of email, Facebook, and physician websites to communicate with physicians: a national online survey of retail pharmacy users [published online June 24, 2015]. J Gen Intern Med. doi:10.1007/s11606-015-3374-7.
- Pew Research Center. Social networking fact sheet. Available at: http://www.pewinternet.org/fact‐sheets/social‐networking‐fact‐sheet. Accessed March 4, 2015.
- , . Three approaches to qualitative content analysis. Qual Health Res. 2005;15(9):1277–1288.
- , , , . Patients’ evaluations of health care providers in the era of social networking: an analysis of physician‐rating websites. J Gen Intern Med. 2010;25(9):942–946.
- , , , , . Vaccine counseling: a content analysis of patient‐physician discussions regarding human papilloma virus vaccine. Vaccine. 2011;29(43):7343–7349.
- . Qualitative research methods. Int J Qual Health Care. 2002;14(4):329–336.
- , . Doing Qualitative Research. Vol 2. Thousand Oaks, CA: Sage Publications; 1999.
- , , , . Patients’ perception of hospital care in the United States. N Engl J Med. 2008;359(18):1921–1931.
- , , , et al. Tweets about hospital quality: a mixed methods study. BMJ Qual Saf. 2014;23(10):838–846.