Michael B. Rothberg, MD, MPH

Setting and Patients

We conducted a retrospective cohort study using data from 286 US nonfederal acute‐care facilities contributing to the database maintained by Premier (Premier Healthcare Solutions, Inc., Charlotte, NC). This database, created to measure healthcare utilization and quality of care, is drawn from voluntarily participating hospitals and contains data on approximately 1 in every 4 discharges nationwide.[11] Participating hospitals are similar in geographic distribution and metropolitan (urban/rural) status to hospitals nationwide, although large, nonteaching hospitals are slightly overrepresented in Premier. The database contains patient demographics, International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) codes, hospital demographics, and a date‐stamped log of all charges during the course of each hospitalization, including diagnostic tests, therapeutic treatments, and medications with dose and route of administration. The study was approved by the institutional review board at Beth Israel Deaconess Medical Center and granted a waiver of informed consent.

We studied a cohort of all adult nonsurgical admissions to participating hospitals from July 1, 2009, through June 30, 2010. We chose to study nonsurgical admissions, as patients undergoing surgical procedures have a clear indication for, and almost always receive, opioid pain medications. We defined a nonsurgical admission as an admission in which there were no charges for operating‐room procedures (including labor and delivery) and the attending of record was not a surgeon. We excluded admissions with unknown gender, since this is a key demographic variable, and admissions with a length of stay greater than 365 days, as these admissions are not representative of the typical admission to an acute‐care hospital. At the hospital level, we excluded hospitals contributing <100 admissions owing to resultant lack of precision in corresponding hospital prescribing rates, and hospitals that did not prescribe the full range of opioid medications (these hospitals had charges for codeine only), as these facilities seemed likely to have unusual limitations on prescribing or incomplete data capture.

Opioid Exposure

We defined opioid exposure as presence of 1 charge for an opioid medication during the admission. Opioid medications included morphine, hydrocodone, hydromorphone, oxycodone, fentanyl, meperidine, methadone, codeine, tramadol, buprenorphine, levorphanol, oxymorphone, pentazocine, propoxyphene, tapentadol, butorphanol, dezocine, and nalbuphine. We grouped the last 9 into an other category owing to infrequent use and/or differing characteristics from the main opioid drug types, such as synthetic, semisynthetic, and partial agonist qualities.

Severe Opioid‐Related Adverse Events

We defined severe opioid‐related adverse events as either naloxone exposure or an opioid‐related adverse drug event diagnosis code. Naloxone use in an adult patient exposed to opioids is one of the Institute for Healthcare Improvement's Trigger Tools for identifying adverse drug events[12] and previously has been demonstrated to have high positive predictive value for a confirmed adverse drug event.[13] We defined naloxone exposure as presence of 1 charge for naloxone. We excluded charges on hospital day 1 to focus on nosocomial events. We defined opioid‐related adverse drug events using ICD‐9‐CM diagnosis codes for poisoning by opioids (overdose, wrong substance given, or taken in error; ICD‐9‐CM 965.02, 965.09, E850.1, E850.2) and drugs causing adverse effects in therapeutic use (ICD‐9‐CM E935.1, E935.2), as specified in prior analyses by the Agency for Healthcare Research and Quality (AHRQ).[14, 15] To avoid capturing adverse events associated with outpatient use, we required the ICD‐9‐CM code to be qualified as not present on admission using the present on admission indicator required by the Centers for Medicare and Medicaid Services for all discharge diagnosis codes since 2008.[16]

Covariates of Interest

We were interested in the relationship between both patient and hospital characteristics and opioid exposure. Patient characteristics of interest included (1) demographic variables, such as age, sex, race (self‐reported by patients at the time of admission), marital status, and payer; (2) whether or not the patient spent any time in the intensive care unit (ICU); (3) comorbidities, identified via ICD‐9‐CM secondary diagnosis codes and diagnosis‐related groups using Healthcare Cost and Utilization Project Comorbidity Software, version 3.7, based on the work of Elixhauser et al[17, 18]; (4) primary ICD‐9‐CM discharge diagnosis groupings, selected based on hypothesized associations with receipt of opioids, and based on the Clinical Classifications Software (CCS)a diagnosis and procedure categorization scheme maintained by the AHRQ, and defined in the Appendix[19]; (5) and nonoperating‐room‐based procedures potentially necessitating opioids during the admission, selected from the 50 most common ICD‐9‐CM procedure codes in our cohort and grouped as cardiovascular procedures (catheterization and insertion of vascular stents), gastrointestinal procedures (upper and lower endoscopy), and mechanical ventilation, further defined in the Appendix. Hospital characteristics of interest included number of beds, population served (urban vs rural), teaching status, and US census region (Northeast, Midwest, South, West).

Statistical Analysis

We calculated the percent of admissions with exposure to any opioid and the percent exposed to each opioid, along with the total number of different opioid medications used during each admission. We also calculated the percent of admissions with parenteral administration and the percent of admissions with oral administration, among those exposed to the individual categories, and in aggregate. Because medications after discharge were unavailable in Premier's dataset, we report the percent of patients with a charge for opioids on the day of discharge.

We determined the daily dose of an opioid by taking the sum of the doses for that opioid charged on a given day. The average daily dose of an opioid was determined by taking the sum of the daily doses and dividing by the number of days on which 1 dose was charged. To facilitate comparison, all opioids, with the exception of those for which standard equivalences are unavailable (tramadol, other opioid category, oral fentanyl, epidural route for all), were converted to oral morphine equivalents using a standard equivalence conversion table.[20, 21] We excluded from our dosage calculations those charges for which standard morphine equivalence was unavailable, or for which dosage was missing. We also excluded from our dosage calculations any dose that was >3 standard deviations (SD) above the mean dose for that opioid, as such extreme values seemed physiologically implausible and more likely to be a data entry error that could lead to significant overestimation of the mean for that opioid.

All multivariable models used a generalized estimating equation (GEE) via the genmod procedure in SAS, with a Poisson distribution error term and a log link, controlling for repeated patient admissions with an autoregressive correlation structure.

To identify independent predictors of opioid receipt, we used a GEE model of opioid receipt where all patient and hospital characteristics listed in Table 1 were included as independent variables.

To assess hospital variation in opioid prescribing after adjusting for patient characteristics, we used a GEE model of opioid receipt, controlling for all patient characteristics listed in Table 1. We then took the mean of the predicted probabilities of opioid receipt for the patients within each hospital in our cohort to derive the hospital prescribing rate adjusted for patient characteristics. We report the mean, SD, and range of the prescribing rates for the hospitals in our cohort before and after adjustment for patient characteristics.

To assess whether patients admitted to hospitals with higher rates of opioid prescribing have higher relative risk of severe opioid‐related adverse events, we stratified hospitals into opioid‐prescribing rate quartiles and compared the rates of opioid‐related adverse eventsboth overall and among opioid exposedbetween quartiles. To adjust for patient characteristics, we used a GEE model in which severe opioid‐related adverse event (yes/no) was the dependent variable and hospital‐prescribing rate quartile and all patient characteristics in Table 1 were independent variables. We also performed a sensitivity analysis in which we assessed the association between hospital prescribing‐rate quartile and the individual components of our composite outcome. Our results were qualitatively unchanged using this approach, and only the results of our main analysis are presented.

All analyses were carried out using SAS software, version 9.2 (SAS Institute Inc., Cary, NC).

Patient Admission Characteristics

There were 3,190,934 adult admissions to 300 acute‐care hospitals during our study period. After excluding admissions with a length of stay >365 days (n=25), missing patient sex (n=17), and charges for operating‐room procedures or a surgical attending of record (n=2,018,553), 1,172,339 admissions were available for analysis. There were 12 hospitals with incomplete opioid‐prescribing data (n=32,794) and 2 hospitals that contributed <100 admissions each (n=126), leaving 1,139,419 admissions in 286 hospitals in our analytic cohort. The median age of the cohort was 64 years (interquartile range, 4979 years), and 527,062 (46%) were men. Table 1 shows the characteristics of the admissions in the cohort.

Rate, Route, and Dose of Opioid Exposures

Overall, there were 576,373 (51%) admissions with charges for opioid medications. Among those exposed, 244,760 (43%) had charges for multiple opioids during the admission; 172,090 (30%) had charges for 2 different opioids; and 72,670 (13%) had charges for 3 different opioids. Table 2 shows the percent exposed to each opioid, the percent of exposed with parenteral and oral routes of administration, and the mean daily dose received in oral morphine equivalents.

Among the medications/routes for which conversion to morphine equivalents was possible, dosage was missing in 39,728 out of 2,294,673 opioid charges (2%). The average daily dose received in oral morphine equivalents was 68 mg. A total dose of 50 mg per day was received in 39% of exposed, and a total dose of 100 mg per day was received in 23% of exposed. Among those exposed, 52% (26% of overall admissions) had charges for opioids on the day of discharge.

Rates of Opioid Use by Patient and Hospital Characteristics

Table 3 reports the association between admission characteristics and opioid use. Use was highest in patients between the ages of 25 and 54 years. Although use declined with age, 44% of admissions age 65 years had charges for opioid medication. After adjustment for patient demographics, comorbidities, and hospital characteristics, opioid use was more common in females than males, those age 2554 years compared with those older and younger, those of Caucasian race compared with non‐Caucasian race, and those with Medicare or Medicaid primary insurance. Among the primary discharge diagnoses, patients with musculoskeletal injuries, various specific and nonspecific pain‐related diagnoses, and cancer were significantly more likely to receive opioids than patients without these diagnoses, whereas patients with alcohol‐related disorders and psychiatric disorders were significantly less likely to receive opioids than patients without these diagnoses. Patients admitted to hospitals in the Midwest, South, and West were significantly more likely to receive opioid medications than patients in the Northeast.

Variation in Opioid Prescribing

Figure 1 shows the histograms of hospital opioid prescribing rate for the 286 hospitals in our cohort (A) before and (B) after adjustment for patient characteristics. The observed rates ranged from 5% in the lowest‐prescribing hospital to 72% in the highest‐prescribing hospital, with a mean (SD) of 51% (10%). After adjusting for patient characteristics, the adjusted opioid‐prescribing rates ranged from 33% to 64%, with a mean (SD) of 50% (4%).

Figure 1

Severe Opioid‐Related Adverse Events

Among admissions with opioid exposure (n=576,373), naloxone use occurred in 2345 (0.41%) and opioid‐related adverse drug events in 1174 (0.20%), for a total of 3441 (0.60%) severe opioid‐related adverse events (some patients experienced both). Table 4 reports the opioid exposure and severe opioid‐related adverse‐event rates within hospital opioid‐prescribing rate quartiles, along with the adjusted association between the hospital opioid‐prescribing rate quartile and severe opioid‐related adverse events. After adjusting for patient characteristics, the relative risk of a severe opioid‐related adverse event was significantly greater in hospitals with higher opioid‐prescribing rates, both overall and among opioid exposed.

Setting and Patients

We conducted a retrospective cohort study using data from 286 US nonfederal acute‐care facilities contributing to the database maintained by Premier (Premier Healthcare Solutions, Inc., Charlotte, NC). This database, created to measure healthcare utilization and quality of care, is drawn from voluntarily participating hospitals and contains data on approximately 1 in every 4 discharges nationwide.[11] Participating hospitals are similar in geographic distribution and metropolitan (urban/rural) status to hospitals nationwide, although large, nonteaching hospitals are slightly overrepresented in Premier. The database contains patient demographics, International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) codes, hospital demographics, and a date‐stamped log of all charges during the course of each hospitalization, including diagnostic tests, therapeutic treatments, and medications with dose and route of administration. The study was approved by the institutional review board at Beth Israel Deaconess Medical Center and granted a waiver of informed consent.

We studied a cohort of all adult nonsurgical admissions to participating hospitals from July 1, 2009, through June 30, 2010. We chose to study nonsurgical admissions, as patients undergoing surgical procedures have a clear indication for, and almost always receive, opioid pain medications. We defined a nonsurgical admission as an admission in which there were no charges for operating‐room procedures (including labor and delivery) and the attending of record was not a surgeon. We excluded admissions with unknown gender, since this is a key demographic variable, and admissions with a length of stay greater than 365 days, as these admissions are not representative of the typical admission to an acute‐care hospital. At the hospital level, we excluded hospitals contributing <100 admissions owing to resultant lack of precision in corresponding hospital prescribing rates, and hospitals that did not prescribe the full range of opioid medications (these hospitals had charges for codeine only), as these facilities seemed likely to have unusual limitations on prescribing or incomplete data capture.

Opioid Exposure

We defined opioid exposure as presence of 1 charge for an opioid medication during the admission. Opioid medications included morphine, hydrocodone, hydromorphone, oxycodone, fentanyl, meperidine, methadone, codeine, tramadol, buprenorphine, levorphanol, oxymorphone, pentazocine, propoxyphene, tapentadol, butorphanol, dezocine, and nalbuphine. We grouped the last 9 into an other category owing to infrequent use and/or differing characteristics from the main opioid drug types, such as synthetic, semisynthetic, and partial agonist qualities.

Severe Opioid‐Related Adverse Events

We defined severe opioid‐related adverse events as either naloxone exposure or an opioid‐related adverse drug event diagnosis code. Naloxone use in an adult patient exposed to opioids is one of the Institute for Healthcare Improvement's Trigger Tools for identifying adverse drug events[12] and previously has been demonstrated to have high positive predictive value for a confirmed adverse drug event.[13] We defined naloxone exposure as presence of 1 charge for naloxone. We excluded charges on hospital day 1 to focus on nosocomial events. We defined opioid‐related adverse drug events using ICD‐9‐CM diagnosis codes for poisoning by opioids (overdose, wrong substance given, or taken in error; ICD‐9‐CM 965.02, 965.09, E850.1, E850.2) and drugs causing adverse effects in therapeutic use (ICD‐9‐CM E935.1, E935.2), as specified in prior analyses by the Agency for Healthcare Research and Quality (AHRQ).[14, 15] To avoid capturing adverse events associated with outpatient use, we required the ICD‐9‐CM code to be qualified as not present on admission using the present on admission indicator required by the Centers for Medicare and Medicaid Services for all discharge diagnosis codes since 2008.[16]

Covariates of Interest

We were interested in the relationship between both patient and hospital characteristics and opioid exposure. Patient characteristics of interest included (1) demographic variables, such as age, sex, race (self‐reported by patients at the time of admission), marital status, and payer; (2) whether or not the patient spent any time in the intensive care unit (ICU); (3) comorbidities, identified via ICD‐9‐CM secondary diagnosis codes and diagnosis‐related groups using Healthcare Cost and Utilization Project Comorbidity Software, version 3.7, based on the work of Elixhauser et al[17, 18]; (4) primary ICD‐9‐CM discharge diagnosis groupings, selected based on hypothesized associations with receipt of opioids, and based on the Clinical Classifications Software (CCS)a diagnosis and procedure categorization scheme maintained by the AHRQ, and defined in the Appendix[19]; (5) and nonoperating‐room‐based procedures potentially necessitating opioids during the admission, selected from the 50 most common ICD‐9‐CM procedure codes in our cohort and grouped as cardiovascular procedures (catheterization and insertion of vascular stents), gastrointestinal procedures (upper and lower endoscopy), and mechanical ventilation, further defined in the Appendix. Hospital characteristics of interest included number of beds, population served (urban vs rural), teaching status, and US census region (Northeast, Midwest, South, West).

Statistical Analysis

We calculated the percent of admissions with exposure to any opioid and the percent exposed to each opioid, along with the total number of different opioid medications used during each admission. We also calculated the percent of admissions with parenteral administration and the percent of admissions with oral administration, among those exposed to the individual categories, and in aggregate. Because medications after discharge were unavailable in Premier's dataset, we report the percent of patients with a charge for opioids on the day of discharge.

We determined the daily dose of an opioid by taking the sum of the doses for that opioid charged on a given day. The average daily dose of an opioid was determined by taking the sum of the daily doses and dividing by the number of days on which 1 dose was charged. To facilitate comparison, all opioids, with the exception of those for which standard equivalences are unavailable (tramadol, other opioid category, oral fentanyl, epidural route for all), were converted to oral morphine equivalents using a standard equivalence conversion table.[20, 21] We excluded from our dosage calculations those charges for which standard morphine equivalence was unavailable, or for which dosage was missing. We also excluded from our dosage calculations any dose that was >3 standard deviations (SD) above the mean dose for that opioid, as such extreme values seemed physiologically implausible and more likely to be a data entry error that could lead to significant overestimation of the mean for that opioid.

All multivariable models used a generalized estimating equation (GEE) via the genmod procedure in SAS, with a Poisson distribution error term and a log link, controlling for repeated patient admissions with an autoregressive correlation structure.

To identify independent predictors of opioid receipt, we used a GEE model of opioid receipt where all patient and hospital characteristics listed in Table 1 were included as independent variables.

To assess hospital variation in opioid prescribing after adjusting for patient characteristics, we used a GEE model of opioid receipt, controlling for all patient characteristics listed in Table 1. We then took the mean of the predicted probabilities of opioid receipt for the patients within each hospital in our cohort to derive the hospital prescribing rate adjusted for patient characteristics. We report the mean, SD, and range of the prescribing rates for the hospitals in our cohort before and after adjustment for patient characteristics.

To assess whether patients admitted to hospitals with higher rates of opioid prescribing have higher relative risk of severe opioid‐related adverse events, we stratified hospitals into opioid‐prescribing rate quartiles and compared the rates of opioid‐related adverse eventsboth overall and among opioid exposedbetween quartiles. To adjust for patient characteristics, we used a GEE model in which severe opioid‐related adverse event (yes/no) was the dependent variable and hospital‐prescribing rate quartile and all patient characteristics in Table 1 were independent variables. We also performed a sensitivity analysis in which we assessed the association between hospital prescribing‐rate quartile and the individual components of our composite outcome. Our results were qualitatively unchanged using this approach, and only the results of our main analysis are presented.

All analyses were carried out using SAS software, version 9.2 (SAS Institute Inc., Cary, NC).

Patient Admission Characteristics

There were 3,190,934 adult admissions to 300 acute‐care hospitals during our study period. After excluding admissions with a length of stay >365 days (n=25), missing patient sex (n=17), and charges for operating‐room procedures or a surgical attending of record (n=2,018,553), 1,172,339 admissions were available for analysis. There were 12 hospitals with incomplete opioid‐prescribing data (n=32,794) and 2 hospitals that contributed <100 admissions each (n=126), leaving 1,139,419 admissions in 286 hospitals in our analytic cohort. The median age of the cohort was 64 years (interquartile range, 4979 years), and 527,062 (46%) were men. Table 1 shows the characteristics of the admissions in the cohort.

Rate, Route, and Dose of Opioid Exposures

Overall, there were 576,373 (51%) admissions with charges for opioid medications. Among those exposed, 244,760 (43%) had charges for multiple opioids during the admission; 172,090 (30%) had charges for 2 different opioids; and 72,670 (13%) had charges for 3 different opioids. Table 2 shows the percent exposed to each opioid, the percent of exposed with parenteral and oral routes of administration, and the mean daily dose received in oral morphine equivalents.

Among the medications/routes for which conversion to morphine equivalents was possible, dosage was missing in 39,728 out of 2,294,673 opioid charges (2%). The average daily dose received in oral morphine equivalents was 68 mg. A total dose of 50 mg per day was received in 39% of exposed, and a total dose of 100 mg per day was received in 23% of exposed. Among those exposed, 52% (26% of overall admissions) had charges for opioids on the day of discharge.

Rates of Opioid Use by Patient and Hospital Characteristics

Table 3 reports the association between admission characteristics and opioid use. Use was highest in patients between the ages of 25 and 54 years. Although use declined with age, 44% of admissions age 65 years had charges for opioid medication. After adjustment for patient demographics, comorbidities, and hospital characteristics, opioid use was more common in females than males, those age 2554 years compared with those older and younger, those of Caucasian race compared with non‐Caucasian race, and those with Medicare or Medicaid primary insurance. Among the primary discharge diagnoses, patients with musculoskeletal injuries, various specific and nonspecific pain‐related diagnoses, and cancer were significantly more likely to receive opioids than patients without these diagnoses, whereas patients with alcohol‐related disorders and psychiatric disorders were significantly less likely to receive opioids than patients without these diagnoses. Patients admitted to hospitals in the Midwest, South, and West were significantly more likely to receive opioid medications than patients in the Northeast.

Variation in Opioid Prescribing

Figure 1 shows the histograms of hospital opioid prescribing rate for the 286 hospitals in our cohort (A) before and (B) after adjustment for patient characteristics. The observed rates ranged from 5% in the lowest‐prescribing hospital to 72% in the highest‐prescribing hospital, with a mean (SD) of 51% (10%). After adjusting for patient characteristics, the adjusted opioid‐prescribing rates ranged from 33% to 64%, with a mean (SD) of 50% (4%).

Figure 1

Severe Opioid‐Related Adverse Events

Among admissions with opioid exposure (n=576,373), naloxone use occurred in 2345 (0.41%) and opioid‐related adverse drug events in 1174 (0.20%), for a total of 3441 (0.60%) severe opioid‐related adverse events (some patients experienced both). Table 4 reports the opioid exposure and severe opioid‐related adverse‐event rates within hospital opioid‐prescribing rate quartiles, along with the adjusted association between the hospital opioid‐prescribing rate quartile and severe opioid‐related adverse events. After adjusting for patient characteristics, the relative risk of a severe opioid‐related adverse event was significantly greater in hospitals with higher opioid‐prescribing rates, both overall and among opioid exposed.

Setting and Patients

We conducted a retrospective cohort study using data from 286 US nonfederal acute‐care facilities contributing to the database maintained by Premier (Premier Healthcare Solutions, Inc., Charlotte, NC). This database, created to measure healthcare utilization and quality of care, is drawn from voluntarily participating hospitals and contains data on approximately 1 in every 4 discharges nationwide.[11] Participating hospitals are similar in geographic distribution and metropolitan (urban/rural) status to hospitals nationwide, although large, nonteaching hospitals are slightly overrepresented in Premier. The database contains patient demographics, International Classification of Diseases, Ninth Revision, Clinical Modification (ICD‐9‐CM) codes, hospital demographics, and a date‐stamped log of all charges during the course of each hospitalization, including diagnostic tests, therapeutic treatments, and medications with dose and route of administration. The study was approved by the institutional review board at Beth Israel Deaconess Medical Center and granted a waiver of informed consent.

We studied a cohort of all adult nonsurgical admissions to participating hospitals from July 1, 2009, through June 30, 2010. We chose to study nonsurgical admissions, as patients undergoing surgical procedures have a clear indication for, and almost always receive, opioid pain medications. We defined a nonsurgical admission as an admission in which there were no charges for operating‐room procedures (including labor and delivery) and the attending of record was not a surgeon. We excluded admissions with unknown gender, since this is a key demographic variable, and admissions with a length of stay greater than 365 days, as these admissions are not representative of the typical admission to an acute‐care hospital. At the hospital level, we excluded hospitals contributing <100 admissions owing to resultant lack of precision in corresponding hospital prescribing rates, and hospitals that did not prescribe the full range of opioid medications (these hospitals had charges for codeine only), as these facilities seemed likely to have unusual limitations on prescribing or incomplete data capture.

Opioid Exposure

We defined opioid exposure as presence of 1 charge for an opioid medication during the admission. Opioid medications included morphine, hydrocodone, hydromorphone, oxycodone, fentanyl, meperidine, methadone, codeine, tramadol, buprenorphine, levorphanol, oxymorphone, pentazocine, propoxyphene, tapentadol, butorphanol, dezocine, and nalbuphine. We grouped the last 9 into an other category owing to infrequent use and/or differing characteristics from the main opioid drug types, such as synthetic, semisynthetic, and partial agonist qualities.

Severe Opioid‐Related Adverse Events

We defined severe opioid‐related adverse events as either naloxone exposure or an opioid‐related adverse drug event diagnosis code. Naloxone use in an adult patient exposed to opioids is one of the Institute for Healthcare Improvement's Trigger Tools for identifying adverse drug events[12] and previously has been demonstrated to have high positive predictive value for a confirmed adverse drug event.[13] We defined naloxone exposure as presence of 1 charge for naloxone. We excluded charges on hospital day 1 to focus on nosocomial events. We defined opioid‐related adverse drug events using ICD‐9‐CM diagnosis codes for poisoning by opioids (overdose, wrong substance given, or taken in error; ICD‐9‐CM 965.02, 965.09, E850.1, E850.2) and drugs causing adverse effects in therapeutic use (ICD‐9‐CM E935.1, E935.2), as specified in prior analyses by the Agency for Healthcare Research and Quality (AHRQ).[14, 15] To avoid capturing adverse events associated with outpatient use, we required the ICD‐9‐CM code to be qualified as not present on admission using the present on admission indicator required by the Centers for Medicare and Medicaid Services for all discharge diagnosis codes since 2008.[16]

Covariates of Interest

We were interested in the relationship between both patient and hospital characteristics and opioid exposure. Patient characteristics of interest included (1) demographic variables, such as age, sex, race (self‐reported by patients at the time of admission), marital status, and payer; (2) whether or not the patient spent any time in the intensive care unit (ICU); (3) comorbidities, identified via ICD‐9‐CM secondary diagnosis codes and diagnosis‐related groups using Healthcare Cost and Utilization Project Comorbidity Software, version 3.7, based on the work of Elixhauser et al[17, 18]; (4) primary ICD‐9‐CM discharge diagnosis groupings, selected based on hypothesized associations with receipt of opioids, and based on the Clinical Classifications Software (CCS)a diagnosis and procedure categorization scheme maintained by the AHRQ, and defined in the Appendix[19]; (5) and nonoperating‐room‐based procedures potentially necessitating opioids during the admission, selected from the 50 most common ICD‐9‐CM procedure codes in our cohort and grouped as cardiovascular procedures (catheterization and insertion of vascular stents), gastrointestinal procedures (upper and lower endoscopy), and mechanical ventilation, further defined in the Appendix. Hospital characteristics of interest included number of beds, population served (urban vs rural), teaching status, and US census region (Northeast, Midwest, South, West).

Statistical Analysis

We calculated the percent of admissions with exposure to any opioid and the percent exposed to each opioid, along with the total number of different opioid medications used during each admission. We also calculated the percent of admissions with parenteral administration and the percent of admissions with oral administration, among those exposed to the individual categories, and in aggregate. Because medications after discharge were unavailable in Premier's dataset, we report the percent of patients with a charge for opioids on the day of discharge.

We determined the daily dose of an opioid by taking the sum of the doses for that opioid charged on a given day. The average daily dose of an opioid was determined by taking the sum of the daily doses and dividing by the number of days on which 1 dose was charged. To facilitate comparison, all opioids, with the exception of those for which standard equivalences are unavailable (tramadol, other opioid category, oral fentanyl, epidural route for all), were converted to oral morphine equivalents using a standard equivalence conversion table.[20, 21] We excluded from our dosage calculations those charges for which standard morphine equivalence was unavailable, or for which dosage was missing. We also excluded from our dosage calculations any dose that was >3 standard deviations (SD) above the mean dose for that opioid, as such extreme values seemed physiologically implausible and more likely to be a data entry error that could lead to significant overestimation of the mean for that opioid.

All multivariable models used a generalized estimating equation (GEE) via the genmod procedure in SAS, with a Poisson distribution error term and a log link, controlling for repeated patient admissions with an autoregressive correlation structure.

To identify independent predictors of opioid receipt, we used a GEE model of opioid receipt where all patient and hospital characteristics listed in Table 1 were included as independent variables.

To assess hospital variation in opioid prescribing after adjusting for patient characteristics, we used a GEE model of opioid receipt, controlling for all patient characteristics listed in Table 1. We then took the mean of the predicted probabilities of opioid receipt for the patients within each hospital in our cohort to derive the hospital prescribing rate adjusted for patient characteristics. We report the mean, SD, and range of the prescribing rates for the hospitals in our cohort before and after adjustment for patient characteristics.

To assess whether patients admitted to hospitals with higher rates of opioid prescribing have higher relative risk of severe opioid‐related adverse events, we stratified hospitals into opioid‐prescribing rate quartiles and compared the rates of opioid‐related adverse eventsboth overall and among opioid exposedbetween quartiles. To adjust for patient characteristics, we used a GEE model in which severe opioid‐related adverse event (yes/no) was the dependent variable and hospital‐prescribing rate quartile and all patient characteristics in Table 1 were independent variables. We also performed a sensitivity analysis in which we assessed the association between hospital prescribing‐rate quartile and the individual components of our composite outcome. Our results were qualitatively unchanged using this approach, and only the results of our main analysis are presented.

All analyses were carried out using SAS software, version 9.2 (SAS Institute Inc., Cary, NC).

Patient Admission Characteristics

There were 3,190,934 adult admissions to 300 acute‐care hospitals during our study period. After excluding admissions with a length of stay >365 days (n=25), missing patient sex (n=17), and charges for operating‐room procedures or a surgical attending of record (n=2,018,553), 1,172,339 admissions were available for analysis. There were 12 hospitals with incomplete opioid‐prescribing data (n=32,794) and 2 hospitals that contributed <100 admissions each (n=126), leaving 1,139,419 admissions in 286 hospitals in our analytic cohort. The median age of the cohort was 64 years (interquartile range, 4979 years), and 527,062 (46%) were men. Table 1 shows the characteristics of the admissions in the cohort.

Rate, Route, and Dose of Opioid Exposures

Overall, there were 576,373 (51%) admissions with charges for opioid medications. Among those exposed, 244,760 (43%) had charges for multiple opioids during the admission; 172,090 (30%) had charges for 2 different opioids; and 72,670 (13%) had charges for 3 different opioids. Table 2 shows the percent exposed to each opioid, the percent of exposed with parenteral and oral routes of administration, and the mean daily dose received in oral morphine equivalents.

Among the medications/routes for which conversion to morphine equivalents was possible, dosage was missing in 39,728 out of 2,294,673 opioid charges (2%). The average daily dose received in oral morphine equivalents was 68 mg. A total dose of 50 mg per day was received in 39% of exposed, and a total dose of 100 mg per day was received in 23% of exposed. Among those exposed, 52% (26% of overall admissions) had charges for opioids on the day of discharge.

Rates of Opioid Use by Patient and Hospital Characteristics

Table 3 reports the association between admission characteristics and opioid use. Use was highest in patients between the ages of 25 and 54 years. Although use declined with age, 44% of admissions age 65 years had charges for opioid medication. After adjustment for patient demographics, comorbidities, and hospital characteristics, opioid use was more common in females than males, those age 2554 years compared with those older and younger, those of Caucasian race compared with non‐Caucasian race, and those with Medicare or Medicaid primary insurance. Among the primary discharge diagnoses, patients with musculoskeletal injuries, various specific and nonspecific pain‐related diagnoses, and cancer were significantly more likely to receive opioids than patients without these diagnoses, whereas patients with alcohol‐related disorders and psychiatric disorders were significantly less likely to receive opioids than patients without these diagnoses. Patients admitted to hospitals in the Midwest, South, and West were significantly more likely to receive opioid medications than patients in the Northeast.

Variation in Opioid Prescribing

Figure 1 shows the histograms of hospital opioid prescribing rate for the 286 hospitals in our cohort (A) before and (B) after adjustment for patient characteristics. The observed rates ranged from 5% in the lowest‐prescribing hospital to 72% in the highest‐prescribing hospital, with a mean (SD) of 51% (10%). After adjusting for patient characteristics, the adjusted opioid‐prescribing rates ranged from 33% to 64%, with a mean (SD) of 50% (4%).

Figure 1

Severe Opioid‐Related Adverse Events

Among admissions with opioid exposure (n=576,373), naloxone use occurred in 2345 (0.41%) and opioid‐related adverse drug events in 1174 (0.20%), for a total of 3441 (0.60%) severe opioid‐related adverse events (some patients experienced both). Table 4 reports the opioid exposure and severe opioid‐related adverse‐event rates within hospital opioid‐prescribing rate quartiles, along with the adjusted association between the hospital opioid‐prescribing rate quartile and severe opioid‐related adverse events. After adjusting for patient characteristics, the relative risk of a severe opioid‐related adverse event was significantly greater in hospitals with higher opioid‐prescribing rates, both overall and among opioid exposed.

Study Population

The hospital's billing database was used to obtain the data. Inclusion criteria included age 18 years or older with index hospitalization between January 2007and July 2009 with International Classification of Diseases, 9th Revision admitting diagnoses of chest pain (786.5), chest pain NOSnot otherwise specified (786.50), chest pain NECnot elsewhere classified (786.59) or angina pectoris (413.9). All eligible patients were admitted under observation status. Although observation patients are technically outpatients, they are cared for by inpatient physicians on inpatient units and are otherwise indistinguishable from inpatients. Patients with a discharge diagnosis of acute myocardial infarction at index admission were excluded. Also, patients who had a chest pain admission or an outpatient stress test within the previous 12 months of index admission were excluded.

Data Collection and Outcomes

All data were extracted electronically from the hospital's billing database. For each patient we noted age, sex, race, insurance status, and cardiovascular comorbidities (current smoker, congestive heart failure, valvular disease, pulmonary/circulatory disorders, peripheral vascular disease, obesity, diabetes mellitus, and hypertension). For each admission we ascertained whether or not any type of stress test was performed. We obtained ED and hospitalization costs for chest pain visits within 12 months of index admission from the hospital's cost accounting system. We also obtained corresponding physician charges as well as collection rate from the health system's clinical decision support system.

The primary outcome was the rate of ED visits and readmissions for chest pain within 1 year of the index visit. Secondary outcomes included total annual hospitalization and ED costs. Total annual costs were calculated by summing index costs and follow‐up costs for subsequent ED visits and readmissions.

Statistical Analysis

Fisher exact (categorical) and unpaired t tests/Wilcoxon rank sum (continuous) tests were used to compare the baseline characteristics of patients who received a stress test at index admission to those who did not. To address possible confounding by indication (allocation bias), the association between stress testing and various outcomes was quantified using multivariable logistic (ED visits and readmissions) or linear regression (costs).[8, 9] In addition, we developed a propensity model using conditional logistic regression and matched patients on propensity score using 1:1 greedy matching algorithm with a caliper tolerance of 0.05.[10, 11] For cost analyses, the annual collection rate was applied to all physician charges, and these were added to hospital or ED costs to obtain the total cost of each visit. The average cost of ED visits or readmissions for each group was calculated by dividing the total ED or readmission cost by the number of ED visits and readmissions, respectively. Physician charges were unavailable for approximately one‐third (1487/5163 or 29%) of all hospitalizations; missing charges were estimated using mean imputation, and sensitivity analyses were conducted to ensure consistency of inferences between full (imputed) and restricted models.[12, 13, 14] Stata/MP 12.1 for Windows (StataCorp, College Station, TX) was used for all analyses.

Subsequent ED Visits for Chest Pain

Within 1 year, 1279 (38.6%) of all patients returned to the ED, and 256 (7.7%) returned at least once for chest pain. Patients who had a stress test at index admission were less likely to return to ED for chest pain, compared to those who did not get a stress test at admission (6.2% vs 11.5%; P<0.001). The median time to the first subsequent ED visit for any complaint was greater among patients who had a stress test at index admission (108 days vs 67 days, P<0.001), but no effect was noted on time to return for chest complaint (67 days vs 71 days, P=0.86).

In a multivariable model, return to the ED for chest pain was positively associated with self‐reported nonwhite race, insurance with Medicare or Medicaid, and earlier year of index admission (Table 2). Return ED visit was negatively associated with stress testing at index admission (adjusted odds ratio [OR]: 0.5, 95% CI: 0.4 to 0.7; propensity‐matched analysis OR: 0.6, 95% CI: 0.5 to 0.9).

Subsequent Readmissions for Chest Pain

Of the 256 patients who returned to the ED for chest pain, 112 (43.8%) were readmitted during the first return visit. There was no statistically significant difference in the proportion admitted from the ED by prior stress test status. In a multivariable model, readmission after returning to the ED for chest pain was positively associated with cardiac comorbidities and earlier year of index admission (Table 3). The decision to readmit was not significantly associated with prior stress testing (adjusted OR: 0.8, 95% CI: 0.5 to 1.4; propensity‐matched analysis OR: 0.8, 95% CI: 0.4 to 1.4).

Cost Analysis

The average multivariable‐adjusted cost (hospital+physician costs) for a patient at index chest pain admission was $3462 if a stress test was performed compared to $2374 without a stress test (+$1088, 95% CI: $972 to $1203). In the propensity‐matched sample the difference was +$1211(95% CI: $1084 to $1338). There were 155 occasions on which a patient returned to the ED for chest pain but was not readmitted. The average per‐visit cost did not differ based on prior stress test status in the overall sample ($763 if stress testing done previously vs $722 if not [+$41, 95% CI: $43 to+$125]) or in the propensity‐matched sample ($787 if stress testing was done vs $744 if not [$43, 95% CI: $54 to +$140]). Because ED visits were less frequent among patients who had a stress test at index admission, the average annual cost of ED visits was significantly lower for this group ($32 vs $52; $20, 95% CI: $36 to $4) or ($42 vs $54; $12 (95% CI: $32 to +$8) in the propensity‐matched sample. For the 117 occasions on which a patient returned with chest pain and was readmitted, the average cost per readmission also did not differ based on whether a stress test was performed at index admission or not ($2912 vs $2806, P=0.85). Again, because readmissions were less common after stress testing, the average cost of readmissions was lower for patients with stress tests than for those without ($88 vs $180; $92, 95% CI: $176 to $8) or $137 vs $194 ( $57, 95% CI: $161 to $47) in the propensity‐matched sample. The total cost of all visits (index, ED, and readmissions) was higher for patients who had a stress test at index admission than for those who did not ($3582 vs $2606; +$975, 95% CI: $829 to $1122) or ($3833 vs $2690; +$1142, 95% CI: $970, $1315) in the propensity‐matched sample.

Study Population

The hospital's billing database was used to obtain the data. Inclusion criteria included age 18 years or older with index hospitalization between January 2007and July 2009 with International Classification of Diseases, 9th Revision admitting diagnoses of chest pain (786.5), chest pain NOSnot otherwise specified (786.50), chest pain NECnot elsewhere classified (786.59) or angina pectoris (413.9). All eligible patients were admitted under observation status. Although observation patients are technically outpatients, they are cared for by inpatient physicians on inpatient units and are otherwise indistinguishable from inpatients. Patients with a discharge diagnosis of acute myocardial infarction at index admission were excluded. Also, patients who had a chest pain admission or an outpatient stress test within the previous 12 months of index admission were excluded.

Data Collection and Outcomes

All data were extracted electronically from the hospital's billing database. For each patient we noted age, sex, race, insurance status, and cardiovascular comorbidities (current smoker, congestive heart failure, valvular disease, pulmonary/circulatory disorders, peripheral vascular disease, obesity, diabetes mellitus, and hypertension). For each admission we ascertained whether or not any type of stress test was performed. We obtained ED and hospitalization costs for chest pain visits within 12 months of index admission from the hospital's cost accounting system. We also obtained corresponding physician charges as well as collection rate from the health system's clinical decision support system.

The primary outcome was the rate of ED visits and readmissions for chest pain within 1 year of the index visit. Secondary outcomes included total annual hospitalization and ED costs. Total annual costs were calculated by summing index costs and follow‐up costs for subsequent ED visits and readmissions.

Statistical Analysis

Fisher exact (categorical) and unpaired t tests/Wilcoxon rank sum (continuous) tests were used to compare the baseline characteristics of patients who received a stress test at index admission to those who did not. To address possible confounding by indication (allocation bias), the association between stress testing and various outcomes was quantified using multivariable logistic (ED visits and readmissions) or linear regression (costs).[8, 9] In addition, we developed a propensity model using conditional logistic regression and matched patients on propensity score using 1:1 greedy matching algorithm with a caliper tolerance of 0.05.[10, 11] For cost analyses, the annual collection rate was applied to all physician charges, and these were added to hospital or ED costs to obtain the total cost of each visit. The average cost of ED visits or readmissions for each group was calculated by dividing the total ED or readmission cost by the number of ED visits and readmissions, respectively. Physician charges were unavailable for approximately one‐third (1487/5163 or 29%) of all hospitalizations; missing charges were estimated using mean imputation, and sensitivity analyses were conducted to ensure consistency of inferences between full (imputed) and restricted models.[12, 13, 14] Stata/MP 12.1 for Windows (StataCorp, College Station, TX) was used for all analyses.

Subsequent ED Visits for Chest Pain

Within 1 year, 1279 (38.6%) of all patients returned to the ED, and 256 (7.7%) returned at least once for chest pain. Patients who had a stress test at index admission were less likely to return to ED for chest pain, compared to those who did not get a stress test at admission (6.2% vs 11.5%; P<0.001). The median time to the first subsequent ED visit for any complaint was greater among patients who had a stress test at index admission (108 days vs 67 days, P<0.001), but no effect was noted on time to return for chest complaint (67 days vs 71 days, P=0.86).

In a multivariable model, return to the ED for chest pain was positively associated with self‐reported nonwhite race, insurance with Medicare or Medicaid, and earlier year of index admission (Table 2). Return ED visit was negatively associated with stress testing at index admission (adjusted odds ratio [OR]: 0.5, 95% CI: 0.4 to 0.7; propensity‐matched analysis OR: 0.6, 95% CI: 0.5 to 0.9).

Subsequent Readmissions for Chest Pain

Of the 256 patients who returned to the ED for chest pain, 112 (43.8%) were readmitted during the first return visit. There was no statistically significant difference in the proportion admitted from the ED by prior stress test status. In a multivariable model, readmission after returning to the ED for chest pain was positively associated with cardiac comorbidities and earlier year of index admission (Table 3). The decision to readmit was not significantly associated with prior stress testing (adjusted OR: 0.8, 95% CI: 0.5 to 1.4; propensity‐matched analysis OR: 0.8, 95% CI: 0.4 to 1.4).

Cost Analysis

The average multivariable‐adjusted cost (hospital+physician costs) for a patient at index chest pain admission was $3462 if a stress test was performed compared to $2374 without a stress test (+$1088, 95% CI: $972 to $1203). In the propensity‐matched sample the difference was +$1211(95% CI: $1084 to $1338). There were 155 occasions on which a patient returned to the ED for chest pain but was not readmitted. The average per‐visit cost did not differ based on prior stress test status in the overall sample ($763 if stress testing done previously vs $722 if not [+$41, 95% CI: $43 to+$125]) or in the propensity‐matched sample ($787 if stress testing was done vs $744 if not [$43, 95% CI: $54 to +$140]). Because ED visits were less frequent among patients who had a stress test at index admission, the average annual cost of ED visits was significantly lower for this group ($32 vs $52; $20, 95% CI: $36 to $4) or ($42 vs $54; $12 (95% CI: $32 to +$8) in the propensity‐matched sample. For the 117 occasions on which a patient returned with chest pain and was readmitted, the average cost per readmission also did not differ based on whether a stress test was performed at index admission or not ($2912 vs $2806, P=0.85). Again, because readmissions were less common after stress testing, the average cost of readmissions was lower for patients with stress tests than for those without ($88 vs $180; $92, 95% CI: $176 to $8) or $137 vs $194 ( $57, 95% CI: $161 to $47) in the propensity‐matched sample. The total cost of all visits (index, ED, and readmissions) was higher for patients who had a stress test at index admission than for those who did not ($3582 vs $2606; +$975, 95% CI: $829 to $1122) or ($3833 vs $2690; +$1142, 95% CI: $970, $1315) in the propensity‐matched sample.

Study Population

The hospital's billing database was used to obtain the data. Inclusion criteria included age 18 years or older with index hospitalization between January 2007and July 2009 with International Classification of Diseases, 9th Revision admitting diagnoses of chest pain (786.5), chest pain NOSnot otherwise specified (786.50), chest pain NECnot elsewhere classified (786.59) or angina pectoris (413.9). All eligible patients were admitted under observation status. Although observation patients are technically outpatients, they are cared for by inpatient physicians on inpatient units and are otherwise indistinguishable from inpatients. Patients with a discharge diagnosis of acute myocardial infarction at index admission were excluded. Also, patients who had a chest pain admission or an outpatient stress test within the previous 12 months of index admission were excluded.

Data Collection and Outcomes

All data were extracted electronically from the hospital's billing database. For each patient we noted age, sex, race, insurance status, and cardiovascular comorbidities (current smoker, congestive heart failure, valvular disease, pulmonary/circulatory disorders, peripheral vascular disease, obesity, diabetes mellitus, and hypertension). For each admission we ascertained whether or not any type of stress test was performed. We obtained ED and hospitalization costs for chest pain visits within 12 months of index admission from the hospital's cost accounting system. We also obtained corresponding physician charges as well as collection rate from the health system's clinical decision support system.

The primary outcome was the rate of ED visits and readmissions for chest pain within 1 year of the index visit. Secondary outcomes included total annual hospitalization and ED costs. Total annual costs were calculated by summing index costs and follow‐up costs for subsequent ED visits and readmissions.

Statistical Analysis

Fisher exact (categorical) and unpaired t tests/Wilcoxon rank sum (continuous) tests were used to compare the baseline characteristics of patients who received a stress test at index admission to those who did not. To address possible confounding by indication (allocation bias), the association between stress testing and various outcomes was quantified using multivariable logistic (ED visits and readmissions) or linear regression (costs).[8, 9] In addition, we developed a propensity model using conditional logistic regression and matched patients on propensity score using 1:1 greedy matching algorithm with a caliper tolerance of 0.05.[10, 11] For cost analyses, the annual collection rate was applied to all physician charges, and these were added to hospital or ED costs to obtain the total cost of each visit. The average cost of ED visits or readmissions for each group was calculated by dividing the total ED or readmission cost by the number of ED visits and readmissions, respectively. Physician charges were unavailable for approximately one‐third (1487/5163 or 29%) of all hospitalizations; missing charges were estimated using mean imputation, and sensitivity analyses were conducted to ensure consistency of inferences between full (imputed) and restricted models.[12, 13, 14] Stata/MP 12.1 for Windows (StataCorp, College Station, TX) was used for all analyses.

Subsequent ED Visits for Chest Pain

Within 1 year, 1279 (38.6%) of all patients returned to the ED, and 256 (7.7%) returned at least once for chest pain. Patients who had a stress test at index admission were less likely to return to ED for chest pain, compared to those who did not get a stress test at admission (6.2% vs 11.5%; P<0.001). The median time to the first subsequent ED visit for any complaint was greater among patients who had a stress test at index admission (108 days vs 67 days, P<0.001), but no effect was noted on time to return for chest complaint (67 days vs 71 days, P=0.86).

In a multivariable model, return to the ED for chest pain was positively associated with self‐reported nonwhite race, insurance with Medicare or Medicaid, and earlier year of index admission (Table 2). Return ED visit was negatively associated with stress testing at index admission (adjusted odds ratio [OR]: 0.5, 95% CI: 0.4 to 0.7; propensity‐matched analysis OR: 0.6, 95% CI: 0.5 to 0.9).

Subsequent Readmissions for Chest Pain

Of the 256 patients who returned to the ED for chest pain, 112 (43.8%) were readmitted during the first return visit. There was no statistically significant difference in the proportion admitted from the ED by prior stress test status. In a multivariable model, readmission after returning to the ED for chest pain was positively associated with cardiac comorbidities and earlier year of index admission (Table 3). The decision to readmit was not significantly associated with prior stress testing (adjusted OR: 0.8, 95% CI: 0.5 to 1.4; propensity‐matched analysis OR: 0.8, 95% CI: 0.4 to 1.4).

Cost Analysis

The average multivariable‐adjusted cost (hospital+physician costs) for a patient at index chest pain admission was $3462 if a stress test was performed compared to $2374 without a stress test (+$1088, 95% CI: $972 to $1203). In the propensity‐matched sample the difference was +$1211(95% CI: $1084 to $1338). There were 155 occasions on which a patient returned to the ED for chest pain but was not readmitted. The average per‐visit cost did not differ based on prior stress test status in the overall sample ($763 if stress testing done previously vs $722 if not [+$41, 95% CI: $43 to+$125]) or in the propensity‐matched sample ($787 if stress testing was done vs $744 if not [$43, 95% CI: $54 to +$140]). Because ED visits were less frequent among patients who had a stress test at index admission, the average annual cost of ED visits was significantly lower for this group ($32 vs $52; $20, 95% CI: $36 to $4) or ($42 vs $54; $12 (95% CI: $32 to +$8) in the propensity‐matched sample. For the 117 occasions on which a patient returned with chest pain and was readmitted, the average cost per readmission also did not differ based on whether a stress test was performed at index admission or not ($2912 vs $2806, P=0.85). Again, because readmissions were less common after stress testing, the average cost of readmissions was lower for patients with stress tests than for those without ($88 vs $180; $92, 95% CI: $176 to $8) or $137 vs $194 ( $57, 95% CI: $161 to $47) in the propensity‐matched sample. The total cost of all visits (index, ED, and readmissions) was higher for patients who had a stress test at index admission than for those who did not ($3582 vs $2606; +$975, 95% CI: $829 to $1122) or ($3833 vs $2690; +$1142, 95% CI: $970, $1315) in the propensity‐matched sample.

Data Source

We utilized data from the Nationwide Inpatient Sample (NIS) of the Health Care Cost and Utilization Project,[16] which is a 20% stratified probability sample of all US acute‐care hospitals each year. These data are drawn from a sampling frame that contains close to 95% of all discharges in the United States, with the hospital discharge record as the unit of analysis. The NIS has been used to study trends in many different diagnoses.[17, 18, 19] The database contains demographic information, payer information, principal and secondary diagnoses, cost, discharge disposition, and death during hospitalization. It also contains information on hospital characteristics including ownership, size, teaching status, and geographic region.

Definitions

We included patients 18 years old discharged between 2001 and 2009 with a primary or secondary diagnosis of ARF. We identified cases of ARF using diagnostic codes (International Classification of Diseases, Ninth Revision, Clinical Modification [ICD‐9‐CM]) previously used in studies of acute organ dysfunction in sepsis (518.81, 518.82, 518.84, 518.4, 799.1, 786.09).[17, 20, 21] To define ARDS we relied on ICD‐9‐CM codes (518.4, 518.82, 518.5, 786.09) used in prior studies that showed good sensitivity and specificity.[22, 23] The use of ventilatory support was identified using the ICD‐9‐CM procedure codes[24] (93.90, 93.70, 93.71, 93.76). Comorbidities were classified using the Agency for Healthcare Research and Quality's (Rockville, MD) Healthcare Cost and Utilization Project's (HCUP) Comorbidity Software version 3.103.5.[25]

Outcomes

The primary outcomes included the annual number of hospitalizations, population incidence, hospital mortality, and costs of care. Secondary outcomes included length of stay, most common diagnoses associated with ARF, disposition at discharge, and use and type of ventilatory support.

Analysis

We estimated the number of hospitalizations with a diagnosis of ARF/year, and we calculated the weighted frequencies following HCUP‐NIS recommendations using SAS/STAT survey procedures. Using population estimates for the years 2001 to 2009 from the US Census Bureau, we employed direct standardization to calculate age‐, gender‐, and race‐adjusted population incidence and mortality rates of ARF per 100,000 population. Hospital mortality was defined as the ratio of ARF hospitalizations ending in death divided by total number of ARF hospitalizations. Mechanical ventilation rates and rates of selected comorbidities were similarly defined.

We employed indirect standardization to adjust hospital mortality rates for age, sex, race/ethnicity, comorbidities, and hospital characteristics using logistic regression models from 2001 to predict hospital mortality for 2002 to 2009. We used linear regression models to test whether the slope of year was significant for trends in outcomes overtime. Costs were calculated using hospital‐specific cost‐to‐charge ratios when available and a weighted group average at the state level for remaining hospitals. We converted all costs to 2009 US dollars using the Consumer Price Index. Costs and lengths of stay were not normally distributed, so we calculated weighted geometric means (the average of all logarithmic values), then converted back to a base‐10 number. Using a Taylor series expansion, we then calculated standard errors. All analyses were performed using SAS version 9.2 (SAS Institute, Inc., Cary, NC).

The Baystate Medical Center institutional review board determined that the project did not constitute human subjects research.

Hospitalization Trends

The number of hospitalizations with an ARF diagnosis code increased at an average annual rate of 11.3% from 1,007,549 (standard deviation [SD] = 19,268) in 2001 to 1,917,910 (SD = 47,558) in 2009. More than two‐thirds of ARF admissions were associated with medical, rather than surgical, conditions (69.5% in 2001 and 71.2% in 2009). The median age, racial make‐up, and gender did not change significantly. Over the study period we observed an increase in ARF‐related hospitalizations in large, urban, teaching hospitals and in hospitals located in the Midwest (Table 1).

After adjusting for age and sex, the population incidence of ARF increased from 502 (standard error [SE] = 10) cases per 100,000 in 2001 to 784 (SE = 19) cases per 100,000 in 2009 (a 56% increase, P < 0.0001). Hispanics had the lowest rates of ARF, with both black and white groups having similar rates (Table 2).

The most common etiologies of ARF among medical patients were pneumonia, CHF, ARDS, COPD exacerbation, and sepsis. Over the 9‐year study, the proportion of cases secondary to pneumonia and sepsis rose significantly: from 39% to 46% and 13% to 21%, respectively (Figure 1).

Figure 1

Mortality and Other Outcomes

The number of in‐hospital deaths related to ARF increased from 277,407 deaths in 2001 to 381,155 in 2009 (a 37% increase, P < 0.001). Standardized to the population, deaths increased from 140 in 2001 to 154 cases per 100,000 in 2009 (a 10% increase, P = 0.027). Despite slightly increasing mortality rates at a population level, adjusted in‐hospital mortality improved from 27.6% in 2001 to 20.6% in 2009 (P < 0.001). Mortality declined for both IMV and NIV patients from 35.3% in 2001 to 30.2% in 2009 and from 23.5% to 19%, respectively, but increased for those who required both NIV and IMV (from 26.9% in 2001 to 28% in 2009).

Adjusted hospital length of stay decreased from 7.8 days per patient in 2001 to 7.1 days in 2009 (P < 0.001), with a concomitant increase in discharges to nursing facilities, from 24% in 2001 to 29% in 2009. There was no linear trend in adjusted cost per case, with $15,818 in 2001 and $15,987 in 2009 (in 2009 US dollars) (Table 1).

Ventilation Practices

Overall, 50.9% patients received ventilatory support (NIV or IMV or both) in 2001 and 49.7% in 2009 (P= 0.25). The use of NIV increased from 3.8% to 10.1% (P < 0.001), a 169% increase, whereas the utilization of IMV decreased from 48.5% in 2001 to 42.1% in 2009 (P for trend < 0.0001), a 13% decrease. Uses of both NIV and IMV during hospitalization were seen in 1.4% of cases in 2001 and 2.5% of cases in 2009.

2009 Data Analysis

In 2009 the 1,917,910 hospitalizations with ARF resulted in 381,155 (SD = 8965) deaths and a total inpatient cost of $54 billion. The most common etiologies in patients over 65 years old were pneumonia, CHF, COPD, ARDS, and sepsis. In patients younger than 45 years the most frequent diagnoses were pneumonia, ARDS, sepsis, asthma, drug ingestion, and trauma. Stratified analysis by gender and by age groups showed that mortality rates among men were higher than for women and were highest in patients older than 85 years (Table 3).