Judith Maselli, MSPH

Site and Subjects

We carried out a retrospective, observational study of adult patients discharged from the ICU at University of California San Francisco (UCSF) Medical Center between January 2006 and June 2009. UCSF is a 790‐bed quaternary care academic hospital that admits approximately 17,000 patients annually and has 5 adult ICUs, with 62 beds and 3500 to 4000 ICU admissions annually. Our study was approved by the UCSF Medical Center Committee on Human Research; need for informed consent was waived.

Description of the RRT Before June 1, 2007

Throughout the study, the goal of the RRT was unchanged: to assess, triage, and institute early treatment in patients who experienced an acute decline in their clinical status. From November 2005 to October 2006, the RRT was staffed by an attending hospitalist and medicine resident during daytime, and by a critical care fellow at nighttime and on weekends. The RRT could be activated by any concerned staff member in response to a set of predetermined vital sign abnormalities, decreased urine output, or altered mental status, or simply if the staff member was concerned about the patient's clinical status. Despite extensive educational efforts, utilization of the team was low (2.7 calls per 1000 admissions) and, accordingly, it was discontinued in October 2006. After this time, staff would contact the primary team caring for the patient, should concerns regarding the patient's condition arise.

Description of the RRT After June 1, 2007

In an effort to expand its scope and utility, the RRT was reinstated on June 1, 2007 with a new composition and increased responsibilities. After this date, physician roles were eliminated, and the team composition changed to a dedicated critical care nurse and respiratory therapist, available 24 hours a day. Criteria for calling the team remained unchanged. In addition to responding to acute deteriorations in patients' clinical courses, the RRT began to proactively assess all patients within 12 hours of discharge from the ICU and would continue to round on these patients daily until it was felt that they were clinically stable. During these rounds, the RRT would provide consultation expertise to the bedside nurse and contact the patient's clinicians if concern existed about a patient's clinical trajectory; decisions to transfer a patient back to the ICU ultimately rested with the patient's primary team. During this time period, the RRT received an average of 110.6 calls per 1000 admissions.

Data Sources

Data collected included: demographics, clinical information (all patient refined [APR] severity of illness, APR risk of mortality, and the presence of 29 comorbidities), whether there was a readmission to the ICU, the total ICU LOS, and the vital status at the time of hospital discharge.

Outcomes

Outcomes included: readmission to the ICU, defined as 2 noncontiguous ICU stays during a single hospitalization; ICU LOS, defined as the total number of ICU days accrued during hospitalization; and in‐hospital mortality of patients discharged from the ICU.

Adjustment Variables

Patient age, gender, race, and ethnicity were available from administrative data. We used admission diagnosis code data to classify comorbidities using the method of Elixhauser et al.17

Statistical Analysis

For each of the 3 study outcomes, we assessed the effects of the intervention using multivariable models adjusting for patient‐ and service‐level factors, including a gamma model for ICU LOS and logistic models for ICU readmission and in‐hospital mortality of patients discharged from the ICU. We first compared unadjusted outcome levels before and after implementation. We then used an interrupted time series (ITS) framework to assess the effects of the intervention in terms of 5 measures: 1) the secular trend in the mean of the outcome before the intervention; 2) the change in the mean at the start of the implementation, or immediate effects; 3) the secular trend in the mean after implementation; 4) the change in secular trend, reflecting cumulative intervention effects; and 5) the net effect of the intervention, estimated as the adjusted difference between the fitted mean at the end of the postintervention period and the expected mean if the preintervention trend had continued without interruption or change.

Secondary Analyses

Given the heterogeneity of the RRT in the preintervention period, we assessed potential changes in trend at October 2006, the month in which the RRT was discontinued. We also examined changes in trend midway through the postimplementation period to evaluate for increased efficacy of the RRT with time.

Selection of Covariates

Age, race, and admitting service were included in both the prepost and ITS models by default for face validity. Additional covariates were selected for each outcome using backwards deletion with a retention criterion of P < 0.05, based on models that allowed the outcome rate to vary freely month to month. Because these data were obtained from administrative billing datasets, and the presence of comorbidities could not be definitively linked with time points during hospitalization, only those comorbidities that were likely present prior at ICU discharge were included. For similar reasons, APR severity of illness and risk of mortality scores, which were calculated from billing diagnoses at the end of hospitalization, were excluded from the models.

Patient Characteristics

During the study period, 11,687 patients were admitted to the ICU; 10,288 were discharged from the ICU alive and included in the analysis. In the 17 months prior to the introduction of proactive rounding by the RRT, 4902 (41.9%) patients were admitted, and during the 25 months afterwards, 6785 (58.1%) patients. Patients admitted in the 2 time periods were similar, although there were clinically small but statistically significant differences in race, APR severity of illness, APR risk of mortality, and certain comorbidities between the 2 groups (Table 1).

ICU Readmission Rate

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in the ICU readmission rate (6.7% preintervention vs 7.3% postintervention, P = 0.24; Table 1). In the adjusted ITS model, the intervention had no net effect on the odds of ICU readmission (adjusted odds ratio [AOR] for net intervention effect 0.98, 95% confidence interval [CI] 0.42, 2.28), with similar secular trends both preintervention (AOR 1.00 per year, 95% CI 0.97, 1.03), and afterwards (AOR 0.99 per year, 95% CI 0.98, 01.00), and a nonsignificant increase at implementation (Table 2). Figure 1 uses solid lines to show the fitted readmission rates, a hatched line to show the projection of the preintervention secular trend into the postintervention period, and circles to represent adjusted monthly means. The lack of a net intervention effect is indicated by the convergence of the solid and hatched lines 24 months postintervention.

Figure 1

ICU Average LOS

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in ICU average LOS (5.1 days preintervention vs 4.9 days postintervention, P = 0.24; Table 1). Trends in ICU LOS may have changed in October 2006 (P = 0.07), decreasing in the first half of the study period (adjusted rate ratio [ARR] 0.98 per year, 95% CI 0.961.00), but did not change significantly thereafter. As with the ICU readmission rate, neither the change in estimated secular trend after implementation (ARR 0.98, 95% CI 0.961.01), nor the net effect of the intervention (ARR 0.62, 95% CI 0.321.22) was statistically significant (Table 2); these results are depicted graphically in Figure 2.

Figure 2

In‐Hospital Mortality of Patients Discharged From the ICU

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in the mortality of patients discharged from the ICU (6.0% preintervention vs 5.5% postintervention, P = 0.24; Table 1). Similarly, in the adjusted ITS model, the intervention had no statistically significant net effect on the mortality outcome (Table 2 and Figure 3).

Figure 3

Secondary Analyses

Apart from weak evidence for a change in trend in ICU LOS in October 2006, no other changes in trend were found within the preintervention or postintervention periods (data not shown). This suggests that the heterogeneity of the preintervention RRT had no significant impact on the 3 outcomes examined, and that the RRT intervention failed to gain efficacy with time in the postintervention period. Additionally, we saw no outcome benefit in sensitivity analyses among all ICU patients or in service‐defined analyses (eg, surgical services), where ability to control for illness severity was improved.

Site and Subjects

We carried out a retrospective, observational study of adult patients discharged from the ICU at University of California San Francisco (UCSF) Medical Center between January 2006 and June 2009. UCSF is a 790‐bed quaternary care academic hospital that admits approximately 17,000 patients annually and has 5 adult ICUs, with 62 beds and 3500 to 4000 ICU admissions annually. Our study was approved by the UCSF Medical Center Committee on Human Research; need for informed consent was waived.

Description of the RRT Before June 1, 2007

Throughout the study, the goal of the RRT was unchanged: to assess, triage, and institute early treatment in patients who experienced an acute decline in their clinical status. From November 2005 to October 2006, the RRT was staffed by an attending hospitalist and medicine resident during daytime, and by a critical care fellow at nighttime and on weekends. The RRT could be activated by any concerned staff member in response to a set of predetermined vital sign abnormalities, decreased urine output, or altered mental status, or simply if the staff member was concerned about the patient's clinical status. Despite extensive educational efforts, utilization of the team was low (2.7 calls per 1000 admissions) and, accordingly, it was discontinued in October 2006. After this time, staff would contact the primary team caring for the patient, should concerns regarding the patient's condition arise.

Description of the RRT After June 1, 2007

In an effort to expand its scope and utility, the RRT was reinstated on June 1, 2007 with a new composition and increased responsibilities. After this date, physician roles were eliminated, and the team composition changed to a dedicated critical care nurse and respiratory therapist, available 24 hours a day. Criteria for calling the team remained unchanged. In addition to responding to acute deteriorations in patients' clinical courses, the RRT began to proactively assess all patients within 12 hours of discharge from the ICU and would continue to round on these patients daily until it was felt that they were clinically stable. During these rounds, the RRT would provide consultation expertise to the bedside nurse and contact the patient's clinicians if concern existed about a patient's clinical trajectory; decisions to transfer a patient back to the ICU ultimately rested with the patient's primary team. During this time period, the RRT received an average of 110.6 calls per 1000 admissions.

Data Sources

Data collected included: demographics, clinical information (all patient refined [APR] severity of illness, APR risk of mortality, and the presence of 29 comorbidities), whether there was a readmission to the ICU, the total ICU LOS, and the vital status at the time of hospital discharge.

Outcomes

Outcomes included: readmission to the ICU, defined as 2 noncontiguous ICU stays during a single hospitalization; ICU LOS, defined as the total number of ICU days accrued during hospitalization; and in‐hospital mortality of patients discharged from the ICU.

Adjustment Variables

Patient age, gender, race, and ethnicity were available from administrative data. We used admission diagnosis code data to classify comorbidities using the method of Elixhauser et al.17

Statistical Analysis

For each of the 3 study outcomes, we assessed the effects of the intervention using multivariable models adjusting for patient‐ and service‐level factors, including a gamma model for ICU LOS and logistic models for ICU readmission and in‐hospital mortality of patients discharged from the ICU. We first compared unadjusted outcome levels before and after implementation. We then used an interrupted time series (ITS) framework to assess the effects of the intervention in terms of 5 measures: 1) the secular trend in the mean of the outcome before the intervention; 2) the change in the mean at the start of the implementation, or immediate effects; 3) the secular trend in the mean after implementation; 4) the change in secular trend, reflecting cumulative intervention effects; and 5) the net effect of the intervention, estimated as the adjusted difference between the fitted mean at the end of the postintervention period and the expected mean if the preintervention trend had continued without interruption or change.

Secondary Analyses

Given the heterogeneity of the RRT in the preintervention period, we assessed potential changes in trend at October 2006, the month in which the RRT was discontinued. We also examined changes in trend midway through the postimplementation period to evaluate for increased efficacy of the RRT with time.

Selection of Covariates

Age, race, and admitting service were included in both the prepost and ITS models by default for face validity. Additional covariates were selected for each outcome using backwards deletion with a retention criterion of P < 0.05, based on models that allowed the outcome rate to vary freely month to month. Because these data were obtained from administrative billing datasets, and the presence of comorbidities could not be definitively linked with time points during hospitalization, only those comorbidities that were likely present prior at ICU discharge were included. For similar reasons, APR severity of illness and risk of mortality scores, which were calculated from billing diagnoses at the end of hospitalization, were excluded from the models.

Patient Characteristics

During the study period, 11,687 patients were admitted to the ICU; 10,288 were discharged from the ICU alive and included in the analysis. In the 17 months prior to the introduction of proactive rounding by the RRT, 4902 (41.9%) patients were admitted, and during the 25 months afterwards, 6785 (58.1%) patients. Patients admitted in the 2 time periods were similar, although there were clinically small but statistically significant differences in race, APR severity of illness, APR risk of mortality, and certain comorbidities between the 2 groups (Table 1).

ICU Readmission Rate

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in the ICU readmission rate (6.7% preintervention vs 7.3% postintervention, P = 0.24; Table 1). In the adjusted ITS model, the intervention had no net effect on the odds of ICU readmission (adjusted odds ratio [AOR] for net intervention effect 0.98, 95% confidence interval [CI] 0.42, 2.28), with similar secular trends both preintervention (AOR 1.00 per year, 95% CI 0.97, 1.03), and afterwards (AOR 0.99 per year, 95% CI 0.98, 01.00), and a nonsignificant increase at implementation (Table 2). Figure 1 uses solid lines to show the fitted readmission rates, a hatched line to show the projection of the preintervention secular trend into the postintervention period, and circles to represent adjusted monthly means. The lack of a net intervention effect is indicated by the convergence of the solid and hatched lines 24 months postintervention.

Figure 1

ICU Average LOS

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in ICU average LOS (5.1 days preintervention vs 4.9 days postintervention, P = 0.24; Table 1). Trends in ICU LOS may have changed in October 2006 (P = 0.07), decreasing in the first half of the study period (adjusted rate ratio [ARR] 0.98 per year, 95% CI 0.961.00), but did not change significantly thereafter. As with the ICU readmission rate, neither the change in estimated secular trend after implementation (ARR 0.98, 95% CI 0.961.01), nor the net effect of the intervention (ARR 0.62, 95% CI 0.321.22) was statistically significant (Table 2); these results are depicted graphically in Figure 2.

Figure 2

In‐Hospital Mortality of Patients Discharged From the ICU

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in the mortality of patients discharged from the ICU (6.0% preintervention vs 5.5% postintervention, P = 0.24; Table 1). Similarly, in the adjusted ITS model, the intervention had no statistically significant net effect on the mortality outcome (Table 2 and Figure 3).

Figure 3

Secondary Analyses

Apart from weak evidence for a change in trend in ICU LOS in October 2006, no other changes in trend were found within the preintervention or postintervention periods (data not shown). This suggests that the heterogeneity of the preintervention RRT had no significant impact on the 3 outcomes examined, and that the RRT intervention failed to gain efficacy with time in the postintervention period. Additionally, we saw no outcome benefit in sensitivity analyses among all ICU patients or in service‐defined analyses (eg, surgical services), where ability to control for illness severity was improved.

Site and Subjects

We carried out a retrospective, observational study of adult patients discharged from the ICU at University of California San Francisco (UCSF) Medical Center between January 2006 and June 2009. UCSF is a 790‐bed quaternary care academic hospital that admits approximately 17,000 patients annually and has 5 adult ICUs, with 62 beds and 3500 to 4000 ICU admissions annually. Our study was approved by the UCSF Medical Center Committee on Human Research; need for informed consent was waived.

Description of the RRT Before June 1, 2007

Throughout the study, the goal of the RRT was unchanged: to assess, triage, and institute early treatment in patients who experienced an acute decline in their clinical status. From November 2005 to October 2006, the RRT was staffed by an attending hospitalist and medicine resident during daytime, and by a critical care fellow at nighttime and on weekends. The RRT could be activated by any concerned staff member in response to a set of predetermined vital sign abnormalities, decreased urine output, or altered mental status, or simply if the staff member was concerned about the patient's clinical status. Despite extensive educational efforts, utilization of the team was low (2.7 calls per 1000 admissions) and, accordingly, it was discontinued in October 2006. After this time, staff would contact the primary team caring for the patient, should concerns regarding the patient's condition arise.

Description of the RRT After June 1, 2007

In an effort to expand its scope and utility, the RRT was reinstated on June 1, 2007 with a new composition and increased responsibilities. After this date, physician roles were eliminated, and the team composition changed to a dedicated critical care nurse and respiratory therapist, available 24 hours a day. Criteria for calling the team remained unchanged. In addition to responding to acute deteriorations in patients' clinical courses, the RRT began to proactively assess all patients within 12 hours of discharge from the ICU and would continue to round on these patients daily until it was felt that they were clinically stable. During these rounds, the RRT would provide consultation expertise to the bedside nurse and contact the patient's clinicians if concern existed about a patient's clinical trajectory; decisions to transfer a patient back to the ICU ultimately rested with the patient's primary team. During this time period, the RRT received an average of 110.6 calls per 1000 admissions.

Data Sources

Data collected included: demographics, clinical information (all patient refined [APR] severity of illness, APR risk of mortality, and the presence of 29 comorbidities), whether there was a readmission to the ICU, the total ICU LOS, and the vital status at the time of hospital discharge.

Outcomes

Outcomes included: readmission to the ICU, defined as 2 noncontiguous ICU stays during a single hospitalization; ICU LOS, defined as the total number of ICU days accrued during hospitalization; and in‐hospital mortality of patients discharged from the ICU.

Adjustment Variables

Patient age, gender, race, and ethnicity were available from administrative data. We used admission diagnosis code data to classify comorbidities using the method of Elixhauser et al.17

Statistical Analysis

For each of the 3 study outcomes, we assessed the effects of the intervention using multivariable models adjusting for patient‐ and service‐level factors, including a gamma model for ICU LOS and logistic models for ICU readmission and in‐hospital mortality of patients discharged from the ICU. We first compared unadjusted outcome levels before and after implementation. We then used an interrupted time series (ITS) framework to assess the effects of the intervention in terms of 5 measures: 1) the secular trend in the mean of the outcome before the intervention; 2) the change in the mean at the start of the implementation, or immediate effects; 3) the secular trend in the mean after implementation; 4) the change in secular trend, reflecting cumulative intervention effects; and 5) the net effect of the intervention, estimated as the adjusted difference between the fitted mean at the end of the postintervention period and the expected mean if the preintervention trend had continued without interruption or change.

Secondary Analyses

Given the heterogeneity of the RRT in the preintervention period, we assessed potential changes in trend at October 2006, the month in which the RRT was discontinued. We also examined changes in trend midway through the postimplementation period to evaluate for increased efficacy of the RRT with time.

Selection of Covariates

Age, race, and admitting service were included in both the prepost and ITS models by default for face validity. Additional covariates were selected for each outcome using backwards deletion with a retention criterion of P < 0.05, based on models that allowed the outcome rate to vary freely month to month. Because these data were obtained from administrative billing datasets, and the presence of comorbidities could not be definitively linked with time points during hospitalization, only those comorbidities that were likely present prior at ICU discharge were included. For similar reasons, APR severity of illness and risk of mortality scores, which were calculated from billing diagnoses at the end of hospitalization, were excluded from the models.

Patient Characteristics

During the study period, 11,687 patients were admitted to the ICU; 10,288 were discharged from the ICU alive and included in the analysis. In the 17 months prior to the introduction of proactive rounding by the RRT, 4902 (41.9%) patients were admitted, and during the 25 months afterwards, 6785 (58.1%) patients. Patients admitted in the 2 time periods were similar, although there were clinically small but statistically significant differences in race, APR severity of illness, APR risk of mortality, and certain comorbidities between the 2 groups (Table 1).

ICU Readmission Rate

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in the ICU readmission rate (6.7% preintervention vs 7.3% postintervention, P = 0.24; Table 1). In the adjusted ITS model, the intervention had no net effect on the odds of ICU readmission (adjusted odds ratio [AOR] for net intervention effect 0.98, 95% confidence interval [CI] 0.42, 2.28), with similar secular trends both preintervention (AOR 1.00 per year, 95% CI 0.97, 1.03), and afterwards (AOR 0.99 per year, 95% CI 0.98, 01.00), and a nonsignificant increase at implementation (Table 2). Figure 1 uses solid lines to show the fitted readmission rates, a hatched line to show the projection of the preintervention secular trend into the postintervention period, and circles to represent adjusted monthly means. The lack of a net intervention effect is indicated by the convergence of the solid and hatched lines 24 months postintervention.

Figure 1

ICU Average LOS

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in ICU average LOS (5.1 days preintervention vs 4.9 days postintervention, P = 0.24; Table 1). Trends in ICU LOS may have changed in October 2006 (P = 0.07), decreasing in the first half of the study period (adjusted rate ratio [ARR] 0.98 per year, 95% CI 0.961.00), but did not change significantly thereafter. As with the ICU readmission rate, neither the change in estimated secular trend after implementation (ARR 0.98, 95% CI 0.961.01), nor the net effect of the intervention (ARR 0.62, 95% CI 0.321.22) was statistically significant (Table 2); these results are depicted graphically in Figure 2.

Figure 2

In‐Hospital Mortality of Patients Discharged From the ICU

Introduction of proactive rounding by the RRT was not associated with unadjusted differences in the mortality of patients discharged from the ICU (6.0% preintervention vs 5.5% postintervention, P = 0.24; Table 1). Similarly, in the adjusted ITS model, the intervention had no statistically significant net effect on the mortality outcome (Table 2 and Figure 3).

Figure 3

Secondary Analyses

Apart from weak evidence for a change in trend in ICU LOS in October 2006, no other changes in trend were found within the preintervention or postintervention periods (data not shown). This suggests that the heterogeneity of the preintervention RRT had no significant impact on the 3 outcomes examined, and that the RRT intervention failed to gain efficacy with time in the postintervention period. Additionally, we saw no outcome benefit in sensitivity analyses among all ICU patients or in service‐defined analyses (eg, surgical services), where ability to control for illness severity was improved.

Site

Our survey was carried out at UCSF Moffitt‐Long Hospital, a 560‐bed university teaching hospital and the primary university hospital for the University of California San Francisco. This study was reviewed and approved by the Committee on Human Research at UCSF.

Procedure Service

The HPS is composed of two interns who rotate for 2 weeks on a mandatory rotation performing the majority of the procedures done by the service. Every procedure is supervised by an attending hospitalist who has received extended training from interventional radiologists and emergency department ultrasound faculty. Patients are referred to the service by their primary admitting team. Interns receive procedure‐specific didactics, demonstration, and practice with procedure kits, supplemental readings, computer‐based procedure modules, and evidence‐based summaries of procedure‐related considerations. All interns also attend a half‐day procedure simulation session to review procedural and ultrasound techniques.

While interns obtain informed consent and prepare the patient for the procedure, the attending and intern team communicate the following points with each patient: 1) identification as the dedicated procedure team, separate from the primary team caring for the patient; 2) attending self‐identification as the supervisor; 3) attention to stepwise communication with the patient during the procedure; 4) attention to patient comfort throughout the procedure; 5) emphasis on patient safety through the use of time‐outs, sterile technique, and ultrasound when appropriate; and 6) the intention to discuss best practice and teach during the procedure.

All paracentesis and thoracentesis sites are marked by using bedside ultrasound (S‐Cath, SonoSite, Bothell, WA) guidance prior to and, if needed, during the procedure. Ultrasound is occasionally used for marking joint aspiration and lumbar puncture.11 Interns are responsible for making an initial site marking, which is then confirmed by the attending physician. Although not systematized, our service encourages the intern and attending to communicate about proper technique during the procedure itself. For example, attendings ask questions about technique based on evidence in the literature (eg, Why do you replace the stylet in a lumbar puncture needle prior to removal?) or about trouble shooting (eg, What would you do if the flow of ascites stops during this paracentesis?) and also correct any errors in technique (Recall the angle you intended to use based on the ultrasound view).

Patients

Patients are referred to the procedure service by their primary team; referrals are accepted for patients on all services at all levels of care, including the emergency department (ED) and the intensive care unit (ICU). Participants in this study were referred for one of our target procedures (paracentesis, thoracentesis, or lumbar puncture) between November 2008 and July 2009. Patients gave written consent for the supplemental survey independent of consent for the procedure. All consents and procedures were performed in a patient's hospital room and one family member was allowed to stay in the room if desired by the patient. After the completion of the procedure, the attending on the procedure service at the time, which included study authors D.S. and M.M., approached consecutive patients who spoke and read English and were deemed to have capacity to consent for their own procedure to be surveyed. Patients were considered to have capacity to consent based on commonly accepted criteria described in the literature.12, 13 Patients were also excluded if their procedure was performed by the attending alone, if they had repeated procedures done by the service, or if they were too altered or critically ill to participate in the survey.

Survey

Our survey was developed through identification of items reported in the literature,1421 as well as items newly developed for purposes of examining our primary aims. Newly developed questions focused on patients' satisfaction with major aspects of procedure performance as well as the quality and impact of communication with the patient and between members of the team. Two open‐text questions were included to allow patients to share what went well with the procedure as well as areas for improvement. The research team developed a pool of question items for potential inclusion in a patient satisfaction questionnaire. These items were then shown to a group of research‐oriented health professionals, who meet regularly to review academic research protocols. The group provided their opinions about the content and comprehension of the questions, and the written survey employed was a result of their revisions (see Appendix in Supporting Information online).

Written surveys were distributed to patients by the hospitalist attending on service following the procedure as permitted by patients' severity of illness and availability. Surveys were anonymous and self‐administered by the patient or a family member who was in the room for the procedure; all questions were voluntary. A nurse was made responsible for collecting the survey when possible. Survey results were entered into a database without identifiers, with limited demographic information; patient gender, age, and procedure type were included by the attending hospitalist at the end of the survey. A separate and more detailed procedure database was kept of all procedures performed and was used to record patient consent or reason for not consenting as well as documented receipt of a completed survey. This non‐anonymous database contained detailed supplemental information including patient age, level of care, referring service, presence of bloody fluid at any point during the procedure, and physician‐reported immediate complications at the bedside in free text.

Analysis

Reported immediate complications were classified into major and minor based on reported definitions in the literature.2226 Similar to previous studies, major immediate complications were defined as those requiring further procedural intervention, medical therapy, or both.27 Major complications were defined as: bleeding requiring transfusion, pneumothorax requiring a chest tube, respiratory failure, bowel perforation, cerebral herniation or shock, cerebrospinal fluid (CSF) leak requiring intervention, and transfer to a higher level of care. For patients receiving a thoracentesis, chart review was performed to determine the presence of a follow‐up chest x‐ray, the presence of a pneumothorax, or clinical evidence for re‐expansion pulmonary edema. We analyzed differences between respondents and non‐respondents using Chi‐square tests for categorical variables (gender, level of care, referring service, procedure type, bloody fluid, and immediate reported complications) and independent t tests for continuous variables (age).

After review of the open‐ended fields, responses were classified into the following categories: pain control, physician skill, professionalism, communication, symptom relief, procedure duration, and miscellaneous comments. Responses regarding patient perceptions of physician communication were dichotomized into positive (1 = Strongly Agree, 2 = Agree) and negative (3 = Neutral, 4 = Disagree, and 5 = Strongly Disagree), and independent t tests were used to determine the contribution of factors, such as age, while Chi‐square tests were used for the contribution of gender and procedure type. All statistical tests were performed by using the SAS statistical application program (version 9.2).

Respondent Characteristics

Of 324 procedures performed by the HPS during the study period, 95 (29%) were eligible for consent. Of the 229 patients not eligible for consent, 32 (10%) were excluded because the procedure was performed by the attending alone, 76 (23%) lacked English proficiency or literacy, 66 (20%) had altered mental status, 32 (10%) were intubated and/or had severe illness precluding consent, and 23 (7%) were repeat procedures on patients who had previously completed the survey. Only two patients specifically requested an attending to perform the procedure after an introduction to the service. Of the 95 patients eligible for consent, 89 were consented for the survey, and 65 (68%) completed the survey. Of the six eligible, non‐consented patients, all were leaving the floor immediately following the procedure, and time did not allow for consent and survey distribution. There were no differences between eligible responders and nonresponders in age, gender, procedure, requesting service, presence of bloody fluid, or physician‐reported immediate complications (Table 1).

Complications

As complications would likely play a role in procedure satisfaction, we describe immediate complications for the study population. Of the 324 procedures performed during the study period, no patient had predefined major immediate complications. Upon further chart review of the 96 patients that had a thoracentesis performed, all had a follow‐up chest x‐ray and none suffered an iatrogenic pneumothorax or re‐expansion pulmonary edema. Minor immediate complications for the 324 procedures were reported as follows: postprocedure pain in four patients (1.2%), cough in nine patients (2.8%), five equipment malfunctions (1.5%), four ascites leaks (1.2%), and one incisional bleed requiring a suture for hemostasis (0.3%). There was no significant difference in complications between those consented for the survey and the total study population.

Procedure Satisfaction

More than 90% of patients were satisfied or very satisfied with most aspects of the procedure, including the informed consent process, pain control, expertise, and courtesy of physicians (Table 2). The percentage of patients satisfied with the duration of procedure (88%) was lower than for other measures of satisfaction. Of the 38 patients receiving therapeutic procedures, 34 (89%) were satisfied or highly satisfied with the improvement in symptoms following the procedure.

When asked what went well with the procedure, 59 (91%) respondents provided additional comments and feedback. Each response was classified as described in the Methods section. Of the free text responses, 8 of the 59 patients (14%) commented on the attention to pain control (eg, The caring and attention to my pain was most important to me), 5 (8%) on the skills of the operators (Great examination of the entire stomach region with the ultrasound to ensure the best position of the catheter), 6 (10%) on the courtesy and professionalism of the team (eg, Courteous, team‐feeling, addressed my concerns), 9 (15%) on their communication with the team (eg, The doctors made me feel very comfortable before the procedure by laying out the plan and explaining each part of the procedure), and 8 (14%) on relief of their symptoms (eg, There was an almost immediate and significant improvement in my breathing, bloating, and pain). Twenty‐three of the 59 comments (39%) were categorized as miscellaneous (eg, All went great. I fell asleep).

When asked areas for improvement, 55 (85%) patients responded. Thirty‐three patients (60%) reported that nothing could be improved or they instructed the team to just keep doing what you are doing, while 22 (40%) patients expressed a concern. Responses were categorized in a similar fashion to the positive responses. Five of the 22 negative comments (23%) reported that the procedure took too long (eg, Procedure could have been shorter. I got tired sitting up), 4 (18%) commented on pain control (eg, The poke for marking my skin hurt more than the anesthetic. I was surprised), 6 (27%) felt communication was a problem (eg, Discuss the steps with the patient audibly, no whispering, speak clearly), and 7 (32%) had miscellaneous concerns (eg, Try not to do this procedure right after another one).

Physician Communication

Sixty‐four patients (98%) reported that the physicians performing their procedure communicated with each other during the procedure (Table 3). Although one patient did not feel that the physicians communicated with each other, he or she still answered the follow‐up questions regarding perceptions of physician communication. We excluded this patient from our analysis as his or her answers may not be reliable. The majority of patients (84%) reported this communication as reassuring and felt it was a normal part of procedure performance (94%). Those that did not agree that physician communication was reassuring did not differ in average age (P = 0.307), gender (P = 0.511), or procedure type (P = 0.562).

Of all positive and negative comments, five specifically addressed communication between physicians. Most (four) reflected satisfaction with bedside teaching (eg, They discussed the procedure in a professional manner and eased my mind at all times) and with having an expert in the room (eg, [The team] discussed things like needle placement, which was nice because there was a second opinion right there in the room). Patients also felt that it was good to experience the teaching, with one patient reporting that the best part of the procedure was watching doctors learn from each other. Patients did not express specific reservations about bedside teaching, resident technique, or fear of complications in free text.

Site

Our survey was carried out at UCSF Moffitt‐Long Hospital, a 560‐bed university teaching hospital and the primary university hospital for the University of California San Francisco. This study was reviewed and approved by the Committee on Human Research at UCSF.

Procedure Service

The HPS is composed of two interns who rotate for 2 weeks on a mandatory rotation performing the majority of the procedures done by the service. Every procedure is supervised by an attending hospitalist who has received extended training from interventional radiologists and emergency department ultrasound faculty. Patients are referred to the service by their primary admitting team. Interns receive procedure‐specific didactics, demonstration, and practice with procedure kits, supplemental readings, computer‐based procedure modules, and evidence‐based summaries of procedure‐related considerations. All interns also attend a half‐day procedure simulation session to review procedural and ultrasound techniques.

While interns obtain informed consent and prepare the patient for the procedure, the attending and intern team communicate the following points with each patient: 1) identification as the dedicated procedure team, separate from the primary team caring for the patient; 2) attending self‐identification as the supervisor; 3) attention to stepwise communication with the patient during the procedure; 4) attention to patient comfort throughout the procedure; 5) emphasis on patient safety through the use of time‐outs, sterile technique, and ultrasound when appropriate; and 6) the intention to discuss best practice and teach during the procedure.

All paracentesis and thoracentesis sites are marked by using bedside ultrasound (S‐Cath, SonoSite, Bothell, WA) guidance prior to and, if needed, during the procedure. Ultrasound is occasionally used for marking joint aspiration and lumbar puncture.11 Interns are responsible for making an initial site marking, which is then confirmed by the attending physician. Although not systematized, our service encourages the intern and attending to communicate about proper technique during the procedure itself. For example, attendings ask questions about technique based on evidence in the literature (eg, Why do you replace the stylet in a lumbar puncture needle prior to removal?) or about trouble shooting (eg, What would you do if the flow of ascites stops during this paracentesis?) and also correct any errors in technique (Recall the angle you intended to use based on the ultrasound view).

Patients

Patients are referred to the procedure service by their primary team; referrals are accepted for patients on all services at all levels of care, including the emergency department (ED) and the intensive care unit (ICU). Participants in this study were referred for one of our target procedures (paracentesis, thoracentesis, or lumbar puncture) between November 2008 and July 2009. Patients gave written consent for the supplemental survey independent of consent for the procedure. All consents and procedures were performed in a patient's hospital room and one family member was allowed to stay in the room if desired by the patient. After the completion of the procedure, the attending on the procedure service at the time, which included study authors D.S. and M.M., approached consecutive patients who spoke and read English and were deemed to have capacity to consent for their own procedure to be surveyed. Patients were considered to have capacity to consent based on commonly accepted criteria described in the literature.12, 13 Patients were also excluded if their procedure was performed by the attending alone, if they had repeated procedures done by the service, or if they were too altered or critically ill to participate in the survey.

Survey

Our survey was developed through identification of items reported in the literature,1421 as well as items newly developed for purposes of examining our primary aims. Newly developed questions focused on patients' satisfaction with major aspects of procedure performance as well as the quality and impact of communication with the patient and between members of the team. Two open‐text questions were included to allow patients to share what went well with the procedure as well as areas for improvement. The research team developed a pool of question items for potential inclusion in a patient satisfaction questionnaire. These items were then shown to a group of research‐oriented health professionals, who meet regularly to review academic research protocols. The group provided their opinions about the content and comprehension of the questions, and the written survey employed was a result of their revisions (see Appendix in Supporting Information online).

Written surveys were distributed to patients by the hospitalist attending on service following the procedure as permitted by patients' severity of illness and availability. Surveys were anonymous and self‐administered by the patient or a family member who was in the room for the procedure; all questions were voluntary. A nurse was made responsible for collecting the survey when possible. Survey results were entered into a database without identifiers, with limited demographic information; patient gender, age, and procedure type were included by the attending hospitalist at the end of the survey. A separate and more detailed procedure database was kept of all procedures performed and was used to record patient consent or reason for not consenting as well as documented receipt of a completed survey. This non‐anonymous database contained detailed supplemental information including patient age, level of care, referring service, presence of bloody fluid at any point during the procedure, and physician‐reported immediate complications at the bedside in free text.

Analysis

Reported immediate complications were classified into major and minor based on reported definitions in the literature.2226 Similar to previous studies, major immediate complications were defined as those requiring further procedural intervention, medical therapy, or both.27 Major complications were defined as: bleeding requiring transfusion, pneumothorax requiring a chest tube, respiratory failure, bowel perforation, cerebral herniation or shock, cerebrospinal fluid (CSF) leak requiring intervention, and transfer to a higher level of care. For patients receiving a thoracentesis, chart review was performed to determine the presence of a follow‐up chest x‐ray, the presence of a pneumothorax, or clinical evidence for re‐expansion pulmonary edema. We analyzed differences between respondents and non‐respondents using Chi‐square tests for categorical variables (gender, level of care, referring service, procedure type, bloody fluid, and immediate reported complications) and independent t tests for continuous variables (age).

After review of the open‐ended fields, responses were classified into the following categories: pain control, physician skill, professionalism, communication, symptom relief, procedure duration, and miscellaneous comments. Responses regarding patient perceptions of physician communication were dichotomized into positive (1 = Strongly Agree, 2 = Agree) and negative (3 = Neutral, 4 = Disagree, and 5 = Strongly Disagree), and independent t tests were used to determine the contribution of factors, such as age, while Chi‐square tests were used for the contribution of gender and procedure type. All statistical tests were performed by using the SAS statistical application program (version 9.2).

Respondent Characteristics

Of 324 procedures performed by the HPS during the study period, 95 (29%) were eligible for consent. Of the 229 patients not eligible for consent, 32 (10%) were excluded because the procedure was performed by the attending alone, 76 (23%) lacked English proficiency or literacy, 66 (20%) had altered mental status, 32 (10%) were intubated and/or had severe illness precluding consent, and 23 (7%) were repeat procedures on patients who had previously completed the survey. Only two patients specifically requested an attending to perform the procedure after an introduction to the service. Of the 95 patients eligible for consent, 89 were consented for the survey, and 65 (68%) completed the survey. Of the six eligible, non‐consented patients, all were leaving the floor immediately following the procedure, and time did not allow for consent and survey distribution. There were no differences between eligible responders and nonresponders in age, gender, procedure, requesting service, presence of bloody fluid, or physician‐reported immediate complications (Table 1).

Complications

As complications would likely play a role in procedure satisfaction, we describe immediate complications for the study population. Of the 324 procedures performed during the study period, no patient had predefined major immediate complications. Upon further chart review of the 96 patients that had a thoracentesis performed, all had a follow‐up chest x‐ray and none suffered an iatrogenic pneumothorax or re‐expansion pulmonary edema. Minor immediate complications for the 324 procedures were reported as follows: postprocedure pain in four patients (1.2%), cough in nine patients (2.8%), five equipment malfunctions (1.5%), four ascites leaks (1.2%), and one incisional bleed requiring a suture for hemostasis (0.3%). There was no significant difference in complications between those consented for the survey and the total study population.

Procedure Satisfaction

More than 90% of patients were satisfied or very satisfied with most aspects of the procedure, including the informed consent process, pain control, expertise, and courtesy of physicians (Table 2). The percentage of patients satisfied with the duration of procedure (88%) was lower than for other measures of satisfaction. Of the 38 patients receiving therapeutic procedures, 34 (89%) were satisfied or highly satisfied with the improvement in symptoms following the procedure.

When asked what went well with the procedure, 59 (91%) respondents provided additional comments and feedback. Each response was classified as described in the Methods section. Of the free text responses, 8 of the 59 patients (14%) commented on the attention to pain control (eg, The caring and attention to my pain was most important to me), 5 (8%) on the skills of the operators (Great examination of the entire stomach region with the ultrasound to ensure the best position of the catheter), 6 (10%) on the courtesy and professionalism of the team (eg, Courteous, team‐feeling, addressed my concerns), 9 (15%) on their communication with the team (eg, The doctors made me feel very comfortable before the procedure by laying out the plan and explaining each part of the procedure), and 8 (14%) on relief of their symptoms (eg, There was an almost immediate and significant improvement in my breathing, bloating, and pain). Twenty‐three of the 59 comments (39%) were categorized as miscellaneous (eg, All went great. I fell asleep).

When asked areas for improvement, 55 (85%) patients responded. Thirty‐three patients (60%) reported that nothing could be improved or they instructed the team to just keep doing what you are doing, while 22 (40%) patients expressed a concern. Responses were categorized in a similar fashion to the positive responses. Five of the 22 negative comments (23%) reported that the procedure took too long (eg, Procedure could have been shorter. I got tired sitting up), 4 (18%) commented on pain control (eg, The poke for marking my skin hurt more than the anesthetic. I was surprised), 6 (27%) felt communication was a problem (eg, Discuss the steps with the patient audibly, no whispering, speak clearly), and 7 (32%) had miscellaneous concerns (eg, Try not to do this procedure right after another one).

Physician Communication

Sixty‐four patients (98%) reported that the physicians performing their procedure communicated with each other during the procedure (Table 3). Although one patient did not feel that the physicians communicated with each other, he or she still answered the follow‐up questions regarding perceptions of physician communication. We excluded this patient from our analysis as his or her answers may not be reliable. The majority of patients (84%) reported this communication as reassuring and felt it was a normal part of procedure performance (94%). Those that did not agree that physician communication was reassuring did not differ in average age (P = 0.307), gender (P = 0.511), or procedure type (P = 0.562).

Of all positive and negative comments, five specifically addressed communication between physicians. Most (four) reflected satisfaction with bedside teaching (eg, They discussed the procedure in a professional manner and eased my mind at all times) and with having an expert in the room (eg, [The team] discussed things like needle placement, which was nice because there was a second opinion right there in the room). Patients also felt that it was good to experience the teaching, with one patient reporting that the best part of the procedure was watching doctors learn from each other. Patients did not express specific reservations about bedside teaching, resident technique, or fear of complications in free text.

Site

Our survey was carried out at UCSF Moffitt‐Long Hospital, a 560‐bed university teaching hospital and the primary university hospital for the University of California San Francisco. This study was reviewed and approved by the Committee on Human Research at UCSF.

Procedure Service

The HPS is composed of two interns who rotate for 2 weeks on a mandatory rotation performing the majority of the procedures done by the service. Every procedure is supervised by an attending hospitalist who has received extended training from interventional radiologists and emergency department ultrasound faculty. Patients are referred to the service by their primary admitting team. Interns receive procedure‐specific didactics, demonstration, and practice with procedure kits, supplemental readings, computer‐based procedure modules, and evidence‐based summaries of procedure‐related considerations. All interns also attend a half‐day procedure simulation session to review procedural and ultrasound techniques.

While interns obtain informed consent and prepare the patient for the procedure, the attending and intern team communicate the following points with each patient: 1) identification as the dedicated procedure team, separate from the primary team caring for the patient; 2) attending self‐identification as the supervisor; 3) attention to stepwise communication with the patient during the procedure; 4) attention to patient comfort throughout the procedure; 5) emphasis on patient safety through the use of time‐outs, sterile technique, and ultrasound when appropriate; and 6) the intention to discuss best practice and teach during the procedure.

All paracentesis and thoracentesis sites are marked by using bedside ultrasound (S‐Cath, SonoSite, Bothell, WA) guidance prior to and, if needed, during the procedure. Ultrasound is occasionally used for marking joint aspiration and lumbar puncture.11 Interns are responsible for making an initial site marking, which is then confirmed by the attending physician. Although not systematized, our service encourages the intern and attending to communicate about proper technique during the procedure itself. For example, attendings ask questions about technique based on evidence in the literature (eg, Why do you replace the stylet in a lumbar puncture needle prior to removal?) or about trouble shooting (eg, What would you do if the flow of ascites stops during this paracentesis?) and also correct any errors in technique (Recall the angle you intended to use based on the ultrasound view).

Patients

Patients are referred to the procedure service by their primary team; referrals are accepted for patients on all services at all levels of care, including the emergency department (ED) and the intensive care unit (ICU). Participants in this study were referred for one of our target procedures (paracentesis, thoracentesis, or lumbar puncture) between November 2008 and July 2009. Patients gave written consent for the supplemental survey independent of consent for the procedure. All consents and procedures were performed in a patient's hospital room and one family member was allowed to stay in the room if desired by the patient. After the completion of the procedure, the attending on the procedure service at the time, which included study authors D.S. and M.M., approached consecutive patients who spoke and read English and were deemed to have capacity to consent for their own procedure to be surveyed. Patients were considered to have capacity to consent based on commonly accepted criteria described in the literature.12, 13 Patients were also excluded if their procedure was performed by the attending alone, if they had repeated procedures done by the service, or if they were too altered or critically ill to participate in the survey.

Survey

Our survey was developed through identification of items reported in the literature,1421 as well as items newly developed for purposes of examining our primary aims. Newly developed questions focused on patients' satisfaction with major aspects of procedure performance as well as the quality and impact of communication with the patient and between members of the team. Two open‐text questions were included to allow patients to share what went well with the procedure as well as areas for improvement. The research team developed a pool of question items for potential inclusion in a patient satisfaction questionnaire. These items were then shown to a group of research‐oriented health professionals, who meet regularly to review academic research protocols. The group provided their opinions about the content and comprehension of the questions, and the written survey employed was a result of their revisions (see Appendix in Supporting Information online).

Written surveys were distributed to patients by the hospitalist attending on service following the procedure as permitted by patients' severity of illness and availability. Surveys were anonymous and self‐administered by the patient or a family member who was in the room for the procedure; all questions were voluntary. A nurse was made responsible for collecting the survey when possible. Survey results were entered into a database without identifiers, with limited demographic information; patient gender, age, and procedure type were included by the attending hospitalist at the end of the survey. A separate and more detailed procedure database was kept of all procedures performed and was used to record patient consent or reason for not consenting as well as documented receipt of a completed survey. This non‐anonymous database contained detailed supplemental information including patient age, level of care, referring service, presence of bloody fluid at any point during the procedure, and physician‐reported immediate complications at the bedside in free text.

Analysis

Reported immediate complications were classified into major and minor based on reported definitions in the literature.2226 Similar to previous studies, major immediate complications were defined as those requiring further procedural intervention, medical therapy, or both.27 Major complications were defined as: bleeding requiring transfusion, pneumothorax requiring a chest tube, respiratory failure, bowel perforation, cerebral herniation or shock, cerebrospinal fluid (CSF) leak requiring intervention, and transfer to a higher level of care. For patients receiving a thoracentesis, chart review was performed to determine the presence of a follow‐up chest x‐ray, the presence of a pneumothorax, or clinical evidence for re‐expansion pulmonary edema. We analyzed differences between respondents and non‐respondents using Chi‐square tests for categorical variables (gender, level of care, referring service, procedure type, bloody fluid, and immediate reported complications) and independent t tests for continuous variables (age).

After review of the open‐ended fields, responses were classified into the following categories: pain control, physician skill, professionalism, communication, symptom relief, procedure duration, and miscellaneous comments. Responses regarding patient perceptions of physician communication were dichotomized into positive (1 = Strongly Agree, 2 = Agree) and negative (3 = Neutral, 4 = Disagree, and 5 = Strongly Disagree), and independent t tests were used to determine the contribution of factors, such as age, while Chi‐square tests were used for the contribution of gender and procedure type. All statistical tests were performed by using the SAS statistical application program (version 9.2).

Respondent Characteristics

Of 324 procedures performed by the HPS during the study period, 95 (29%) were eligible for consent. Of the 229 patients not eligible for consent, 32 (10%) were excluded because the procedure was performed by the attending alone, 76 (23%) lacked English proficiency or literacy, 66 (20%) had altered mental status, 32 (10%) were intubated and/or had severe illness precluding consent, and 23 (7%) were repeat procedures on patients who had previously completed the survey. Only two patients specifically requested an attending to perform the procedure after an introduction to the service. Of the 95 patients eligible for consent, 89 were consented for the survey, and 65 (68%) completed the survey. Of the six eligible, non‐consented patients, all were leaving the floor immediately following the procedure, and time did not allow for consent and survey distribution. There were no differences between eligible responders and nonresponders in age, gender, procedure, requesting service, presence of bloody fluid, or physician‐reported immediate complications (Table 1).

Complications

As complications would likely play a role in procedure satisfaction, we describe immediate complications for the study population. Of the 324 procedures performed during the study period, no patient had predefined major immediate complications. Upon further chart review of the 96 patients that had a thoracentesis performed, all had a follow‐up chest x‐ray and none suffered an iatrogenic pneumothorax or re‐expansion pulmonary edema. Minor immediate complications for the 324 procedures were reported as follows: postprocedure pain in four patients (1.2%), cough in nine patients (2.8%), five equipment malfunctions (1.5%), four ascites leaks (1.2%), and one incisional bleed requiring a suture for hemostasis (0.3%). There was no significant difference in complications between those consented for the survey and the total study population.

Procedure Satisfaction

More than 90% of patients were satisfied or very satisfied with most aspects of the procedure, including the informed consent process, pain control, expertise, and courtesy of physicians (Table 2). The percentage of patients satisfied with the duration of procedure (88%) was lower than for other measures of satisfaction. Of the 38 patients receiving therapeutic procedures, 34 (89%) were satisfied or highly satisfied with the improvement in symptoms following the procedure.

When asked what went well with the procedure, 59 (91%) respondents provided additional comments and feedback. Each response was classified as described in the Methods section. Of the free text responses, 8 of the 59 patients (14%) commented on the attention to pain control (eg, The caring and attention to my pain was most important to me), 5 (8%) on the skills of the operators (Great examination of the entire stomach region with the ultrasound to ensure the best position of the catheter), 6 (10%) on the courtesy and professionalism of the team (eg, Courteous, team‐feeling, addressed my concerns), 9 (15%) on their communication with the team (eg, The doctors made me feel very comfortable before the procedure by laying out the plan and explaining each part of the procedure), and 8 (14%) on relief of their symptoms (eg, There was an almost immediate and significant improvement in my breathing, bloating, and pain). Twenty‐three of the 59 comments (39%) were categorized as miscellaneous (eg, All went great. I fell asleep).

When asked areas for improvement, 55 (85%) patients responded. Thirty‐three patients (60%) reported that nothing could be improved or they instructed the team to just keep doing what you are doing, while 22 (40%) patients expressed a concern. Responses were categorized in a similar fashion to the positive responses. Five of the 22 negative comments (23%) reported that the procedure took too long (eg, Procedure could have been shorter. I got tired sitting up), 4 (18%) commented on pain control (eg, The poke for marking my skin hurt more than the anesthetic. I was surprised), 6 (27%) felt communication was a problem (eg, Discuss the steps with the patient audibly, no whispering, speak clearly), and 7 (32%) had miscellaneous concerns (eg, Try not to do this procedure right after another one).

Physician Communication

Sixty‐four patients (98%) reported that the physicians performing their procedure communicated with each other during the procedure (Table 3). Although one patient did not feel that the physicians communicated with each other, he or she still answered the follow‐up questions regarding perceptions of physician communication. We excluded this patient from our analysis as his or her answers may not be reliable. The majority of patients (84%) reported this communication as reassuring and felt it was a normal part of procedure performance (94%). Those that did not agree that physician communication was reassuring did not differ in average age (P = 0.307), gender (P = 0.511), or procedure type (P = 0.562).

Of all positive and negative comments, five specifically addressed communication between physicians. Most (four) reflected satisfaction with bedside teaching (eg, They discussed the procedure in a professional manner and eased my mind at all times) and with having an expert in the room (eg, [The team] discussed things like needle placement, which was nice because there was a second opinion right there in the room). Patients also felt that it was good to experience the teaching, with one patient reporting that the best part of the procedure was watching doctors learn from each other. Patients did not express specific reservations about bedside teaching, resident technique, or fear of complications in free text.

Sites and Subjects

Our data were collected on general medicine patients during hospitalization between June 1, 2006 and May 31, 2008, at the University of California San Francisco. The University of California, San Francisco (UCSF) Medical Center is composed of Moffitt‐Long Hospital (a 400‐bed center) and UCSF‐Mount Zion Hospital (a 200‐bed facility) located in San Francisco, CA.

Medical patients at Moffitt‐Long Hospital are admitted to 1 of 8 medical teams composed of a resident, 1 to 2 interns, and 0 to 3 medical students, supervised by an attending physician who is most often a hospitalist. At Moffitt‐Long Hospital, housestaff write all orders and provide 24‐hour coverage to inpatients. Mount Zion medical patients are cared for by 1 of 2 teams and staffed by a hospitalist on each team who is responsible for all elements of care. Both services care for common inpatient diagnoses, as well as specialty‐associated diagnoses such as cancer, pneumonia, and chronic obstructive pulmonary disease (COPD). Of note, at Moffitt‐Long Hospital, those patients with primary cardiac diagnoses are cared for by a separate team composed of housestaff and students supervised by a cardiologist.

The discharge process at both sites utilizes a multidisciplinary teamincluding physicians, case managers, nurses, pharmacists, and discharge coordinatorsworking in concert. Key components include arranging follow‐up care, faxing the discharge summary to the primary care provider, and educating the patient and caregivers, especially regarding medications. While these goals are clearly delineated, significant variability exists in how these tasks are actually accomplished. The multidisciplinary approach, components of the discharge process, and lack of a systematic approach are representative of the discharge process around the country.22

Data

Data regarding patient demographics, age, comorbidities, and insurance status were collected from administrative data systems at UCSF, reflecting the patient's status at the time of index admission. These same systems were used to collect a date‐stamped log of all medications (eg, anticoagulants) for which the patient was billed during the last 48 hours of hospitalization. Specifically, data were obtained for medications previously shown to cause adverse drug events following hospital discharge.23, 24 These medication groups include corticosteroids, anticoagulants, antibiotics, narcotics, nonsteroidal anti‐inflammatory drugs (NSAIDs), cardiovascular medications, antiepileptics, anticholinergics, antidepressants, and antidiabetics. Operational factors that we hypothesized would affect readmission risk included admission source, discharge disposition, and weekday vs. weekend discharge. Case management, social work, and pharmacy services operate with limited staffing on weekends. Likewise, resident and intern physicians are more likely to be off on a weekend day than a weekday; covering attending physicians care for about half of patients during the weekend. Data were obtained from Transition Systems International (TSI, Boston, MA) administrative databases, a cost‐accounting system that collects data abstracted from patient charts upon discharge from UCSF.

Definition of Readmission Measure

Using TSI, we detected readmission to UCSF by screening for any inpatient encounters on any service (not just medicine) within the 30 days following discharge from the general medicine service at the 2 UCSF campuses. We excluded elective readmissions, such as for scheduled chemotherapy. Patients who died at the index admission were excluded from the cohort.

Adjustment Variables

Age, gender, payer status and APR risk of mortality (3M Health Information Systems, St. Paul, MN) were collected from administrative data. The All Patient Refined (APR) risk of mortality is the all patients risk of mortality score developed by 3M which divides patients into 4 subclasses of risk based on clinical problems and comorbidities.25 We used secondary diagnosis codes in administrative data to classify comorbidities using the method of Elixhauser.26

Using the log of medication charges, we grouped high‐risk medications according to the classification scheme of Forster et al.23 and Hanlon et al.24 We then created a count representing the total number of medications administered to patients within the final 48 hours of stay.

Analysis

We first described study patients and hospitals using univariable methods. Multivariable generalized estimating equations (SAS PROC GENMOD) were used to account for clustering of patients within physicians and calculate adjusted odds ratios (ORs). As there were 2 sites within UCSF Medical Center (Moffitt‐Long and Mount Zion hospitals), we included site as a fixed effect in our model. Models were constructed using manual variable selection methods with final selection being made based on whether the covariate was associated with readmission at P < 0.05. All analyses were carried out using SAS version 9.2 (SAS Institute, Inc. Cary, NC).

Baseline Characteristics

During the 2‐year accrual period, 295 attending physicians admitted 6805 unique patients for a total of 10,359 admissions. Seventeen percent of these 10,359 admissions were readmitted within 30 days. The cohort of all patients had a mean age of 59.6 years 19.5 standard deviation (SD), with 52.8% women. The mean length of stay was 5.6 days 10.4 SD. Medicare was the payer source for approximately half of the admissions. The majority of admissions (90.4%) were billed for at least 1 high risk medication, with narcotics, cardiac medications, and antibiotics being the most common. Regarding disposition, 79.5% of admissions were discharged to home; 9.1% were discharged to a skilled nursing facility (SNF).

Baseline sociodemographic, operational, and clinical characteristics for patients readmitted and not readmitted are shown in Table 1. Demographic characteristics with significant differences (P < 0.05) between readmitted and nonreadmitted groups included mean age, race, payer status, and primary language other than English. Regarding operational characteristics, readmitted patients had a higher median length of stay and were more likely to be admitted through the emergency room during their index admission. Discharge to an SNF was higher in the readmitted group versus the nonreadmitted group (9.7% vs. 9.0%). Several clinical factors were more prevalent in the readmitted group: high‐risk medications, specifically steroids, narcotics, and cardiovascular medications; high‐risk medication count of 3 or greater; and comorbidities including congestive heart failure, renal disease, cancer, anemia, and depression.

Frequency of Readmission

The 30‐day readmission rate was 17.0% (1762 patients), with 49.7% (875 patients) of the readmissions occurring within 10 days of discharge. Of patients readmitted, the general medicine service was the readmitting team in 78.2%. A quarter of readmissions (26.2%) had the same primary diagnosis on initial and repeat admission.

Factors Associated With Readmission

Factors associated with readmission were categorized as sociodemographic, operational, and clinical. Factors associated with readmission with P < 0.05 and present in at least 5% of admissions are presented in Table 2. Of the sociodemographic factors, black race was significantly associated with readmission. Within the Medicare cohort, risk for readmission was similar for white vs. nonwhite race, with relative risk of 1.0 (95% confidence interval [CI], 0.86‐1.18). Medicaid as payer status was significantly associated in the unadjusted model, and in the adjusted model showed a trend toward readmission. Mean age was significantly different in the readmitted and nonreadmitted groups, but the difference was small (1.0 year). Moreover, when we evaluated age in 5‐year categories (ex. 65‐70, 71‐75, etc.), age was not associated with readmission. In the adjusted model, none of the operational factors were significantly associated with readmission, including discharge to SNF, weekend discharge, or admit source.

Of the clinical factors, high‐risk medications and 6 comorbidities were associated with readmission. High‐risk medication categories associated with readmission were steroids and narcotics; anticholinergics medications were protective. The 6 comorbidities associated with readmission were congestive heart failure, renal disease, cancer (with and without metastasis), weight loss, and iron deficiency anemia. While APR risk of mortality was associated with readmission at P < 0.05, including APR in our final model did not alter which other factors were significantly associated with readmission. When site (Moffitt‐Long vs. Mount Zion Hospitals) was added to the model, the ORs for factors associated with readmission did not change appreciably (0.01).

Sites and Subjects

Our data were collected on general medicine patients during hospitalization between June 1, 2006 and May 31, 2008, at the University of California San Francisco. The University of California, San Francisco (UCSF) Medical Center is composed of Moffitt‐Long Hospital (a 400‐bed center) and UCSF‐Mount Zion Hospital (a 200‐bed facility) located in San Francisco, CA.

Medical patients at Moffitt‐Long Hospital are admitted to 1 of 8 medical teams composed of a resident, 1 to 2 interns, and 0 to 3 medical students, supervised by an attending physician who is most often a hospitalist. At Moffitt‐Long Hospital, housestaff write all orders and provide 24‐hour coverage to inpatients. Mount Zion medical patients are cared for by 1 of 2 teams and staffed by a hospitalist on each team who is responsible for all elements of care. Both services care for common inpatient diagnoses, as well as specialty‐associated diagnoses such as cancer, pneumonia, and chronic obstructive pulmonary disease (COPD). Of note, at Moffitt‐Long Hospital, those patients with primary cardiac diagnoses are cared for by a separate team composed of housestaff and students supervised by a cardiologist.

The discharge process at both sites utilizes a multidisciplinary teamincluding physicians, case managers, nurses, pharmacists, and discharge coordinatorsworking in concert. Key components include arranging follow‐up care, faxing the discharge summary to the primary care provider, and educating the patient and caregivers, especially regarding medications. While these goals are clearly delineated, significant variability exists in how these tasks are actually accomplished. The multidisciplinary approach, components of the discharge process, and lack of a systematic approach are representative of the discharge process around the country.22

Data

Data regarding patient demographics, age, comorbidities, and insurance status were collected from administrative data systems at UCSF, reflecting the patient's status at the time of index admission. These same systems were used to collect a date‐stamped log of all medications (eg, anticoagulants) for which the patient was billed during the last 48 hours of hospitalization. Specifically, data were obtained for medications previously shown to cause adverse drug events following hospital discharge.23, 24 These medication groups include corticosteroids, anticoagulants, antibiotics, narcotics, nonsteroidal anti‐inflammatory drugs (NSAIDs), cardiovascular medications, antiepileptics, anticholinergics, antidepressants, and antidiabetics. Operational factors that we hypothesized would affect readmission risk included admission source, discharge disposition, and weekday vs. weekend discharge. Case management, social work, and pharmacy services operate with limited staffing on weekends. Likewise, resident and intern physicians are more likely to be off on a weekend day than a weekday; covering attending physicians care for about half of patients during the weekend. Data were obtained from Transition Systems International (TSI, Boston, MA) administrative databases, a cost‐accounting system that collects data abstracted from patient charts upon discharge from UCSF.

Definition of Readmission Measure

Using TSI, we detected readmission to UCSF by screening for any inpatient encounters on any service (not just medicine) within the 30 days following discharge from the general medicine service at the 2 UCSF campuses. We excluded elective readmissions, such as for scheduled chemotherapy. Patients who died at the index admission were excluded from the cohort.

Adjustment Variables

Age, gender, payer status and APR risk of mortality (3M Health Information Systems, St. Paul, MN) were collected from administrative data. The All Patient Refined (APR) risk of mortality is the all patients risk of mortality score developed by 3M which divides patients into 4 subclasses of risk based on clinical problems and comorbidities.25 We used secondary diagnosis codes in administrative data to classify comorbidities using the method of Elixhauser.26

Using the log of medication charges, we grouped high‐risk medications according to the classification scheme of Forster et al.23 and Hanlon et al.24 We then created a count representing the total number of medications administered to patients within the final 48 hours of stay.

Analysis

We first described study patients and hospitals using univariable methods. Multivariable generalized estimating equations (SAS PROC GENMOD) were used to account for clustering of patients within physicians and calculate adjusted odds ratios (ORs). As there were 2 sites within UCSF Medical Center (Moffitt‐Long and Mount Zion hospitals), we included site as a fixed effect in our model. Models were constructed using manual variable selection methods with final selection being made based on whether the covariate was associated with readmission at P < 0.05. All analyses were carried out using SAS version 9.2 (SAS Institute, Inc. Cary, NC).

Baseline Characteristics

During the 2‐year accrual period, 295 attending physicians admitted 6805 unique patients for a total of 10,359 admissions. Seventeen percent of these 10,359 admissions were readmitted within 30 days. The cohort of all patients had a mean age of 59.6 years 19.5 standard deviation (SD), with 52.8% women. The mean length of stay was 5.6 days 10.4 SD. Medicare was the payer source for approximately half of the admissions. The majority of admissions (90.4%) were billed for at least 1 high risk medication, with narcotics, cardiac medications, and antibiotics being the most common. Regarding disposition, 79.5% of admissions were discharged to home; 9.1% were discharged to a skilled nursing facility (SNF).

Baseline sociodemographic, operational, and clinical characteristics for patients readmitted and not readmitted are shown in Table 1. Demographic characteristics with significant differences (P < 0.05) between readmitted and nonreadmitted groups included mean age, race, payer status, and primary language other than English. Regarding operational characteristics, readmitted patients had a higher median length of stay and were more likely to be admitted through the emergency room during their index admission. Discharge to an SNF was higher in the readmitted group versus the nonreadmitted group (9.7% vs. 9.0%). Several clinical factors were more prevalent in the readmitted group: high‐risk medications, specifically steroids, narcotics, and cardiovascular medications; high‐risk medication count of 3 or greater; and comorbidities including congestive heart failure, renal disease, cancer, anemia, and depression.

Frequency of Readmission

The 30‐day readmission rate was 17.0% (1762 patients), with 49.7% (875 patients) of the readmissions occurring within 10 days of discharge. Of patients readmitted, the general medicine service was the readmitting team in 78.2%. A quarter of readmissions (26.2%) had the same primary diagnosis on initial and repeat admission.

Factors Associated With Readmission

Factors associated with readmission were categorized as sociodemographic, operational, and clinical. Factors associated with readmission with P < 0.05 and present in at least 5% of admissions are presented in Table 2. Of the sociodemographic factors, black race was significantly associated with readmission. Within the Medicare cohort, risk for readmission was similar for white vs. nonwhite race, with relative risk of 1.0 (95% confidence interval [CI], 0.86‐1.18). Medicaid as payer status was significantly associated in the unadjusted model, and in the adjusted model showed a trend toward readmission. Mean age was significantly different in the readmitted and nonreadmitted groups, but the difference was small (1.0 year). Moreover, when we evaluated age in 5‐year categories (ex. 65‐70, 71‐75, etc.), age was not associated with readmission. In the adjusted model, none of the operational factors were significantly associated with readmission, including discharge to SNF, weekend discharge, or admit source.

Of the clinical factors, high‐risk medications and 6 comorbidities were associated with readmission. High‐risk medication categories associated with readmission were steroids and narcotics; anticholinergics medications were protective. The 6 comorbidities associated with readmission were congestive heart failure, renal disease, cancer (with and without metastasis), weight loss, and iron deficiency anemia. While APR risk of mortality was associated with readmission at P < 0.05, including APR in our final model did not alter which other factors were significantly associated with readmission. When site (Moffitt‐Long vs. Mount Zion Hospitals) was added to the model, the ORs for factors associated with readmission did not change appreciably (0.01).

Sites and Subjects

Our data were collected on general medicine patients during hospitalization between June 1, 2006 and May 31, 2008, at the University of California San Francisco. The University of California, San Francisco (UCSF) Medical Center is composed of Moffitt‐Long Hospital (a 400‐bed center) and UCSF‐Mount Zion Hospital (a 200‐bed facility) located in San Francisco, CA.

Medical patients at Moffitt‐Long Hospital are admitted to 1 of 8 medical teams composed of a resident, 1 to 2 interns, and 0 to 3 medical students, supervised by an attending physician who is most often a hospitalist. At Moffitt‐Long Hospital, housestaff write all orders and provide 24‐hour coverage to inpatients. Mount Zion medical patients are cared for by 1 of 2 teams and staffed by a hospitalist on each team who is responsible for all elements of care. Both services care for common inpatient diagnoses, as well as specialty‐associated diagnoses such as cancer, pneumonia, and chronic obstructive pulmonary disease (COPD). Of note, at Moffitt‐Long Hospital, those patients with primary cardiac diagnoses are cared for by a separate team composed of housestaff and students supervised by a cardiologist.

The discharge process at both sites utilizes a multidisciplinary teamincluding physicians, case managers, nurses, pharmacists, and discharge coordinatorsworking in concert. Key components include arranging follow‐up care, faxing the discharge summary to the primary care provider, and educating the patient and caregivers, especially regarding medications. While these goals are clearly delineated, significant variability exists in how these tasks are actually accomplished. The multidisciplinary approach, components of the discharge process, and lack of a systematic approach are representative of the discharge process around the country.22

Data

Data regarding patient demographics, age, comorbidities, and insurance status were collected from administrative data systems at UCSF, reflecting the patient's status at the time of index admission. These same systems were used to collect a date‐stamped log of all medications (eg, anticoagulants) for which the patient was billed during the last 48 hours of hospitalization. Specifically, data were obtained for medications previously shown to cause adverse drug events following hospital discharge.23, 24 These medication groups include corticosteroids, anticoagulants, antibiotics, narcotics, nonsteroidal anti‐inflammatory drugs (NSAIDs), cardiovascular medications, antiepileptics, anticholinergics, antidepressants, and antidiabetics. Operational factors that we hypothesized would affect readmission risk included admission source, discharge disposition, and weekday vs. weekend discharge. Case management, social work, and pharmacy services operate with limited staffing on weekends. Likewise, resident and intern physicians are more likely to be off on a weekend day than a weekday; covering attending physicians care for about half of patients during the weekend. Data were obtained from Transition Systems International (TSI, Boston, MA) administrative databases, a cost‐accounting system that collects data abstracted from patient charts upon discharge from UCSF.

Definition of Readmission Measure

Using TSI, we detected readmission to UCSF by screening for any inpatient encounters on any service (not just medicine) within the 30 days following discharge from the general medicine service at the 2 UCSF campuses. We excluded elective readmissions, such as for scheduled chemotherapy. Patients who died at the index admission were excluded from the cohort.

Adjustment Variables

Age, gender, payer status and APR risk of mortality (3M Health Information Systems, St. Paul, MN) were collected from administrative data. The All Patient Refined (APR) risk of mortality is the all patients risk of mortality score developed by 3M which divides patients into 4 subclasses of risk based on clinical problems and comorbidities.25 We used secondary diagnosis codes in administrative data to classify comorbidities using the method of Elixhauser.26

Using the log of medication charges, we grouped high‐risk medications according to the classification scheme of Forster et al.23 and Hanlon et al.24 We then created a count representing the total number of medications administered to patients within the final 48 hours of stay.

Analysis

We first described study patients and hospitals using univariable methods. Multivariable generalized estimating equations (SAS PROC GENMOD) were used to account for clustering of patients within physicians and calculate adjusted odds ratios (ORs). As there were 2 sites within UCSF Medical Center (Moffitt‐Long and Mount Zion hospitals), we included site as a fixed effect in our model. Models were constructed using manual variable selection methods with final selection being made based on whether the covariate was associated with readmission at P < 0.05. All analyses were carried out using SAS version 9.2 (SAS Institute, Inc. Cary, NC).

Baseline Characteristics

During the 2‐year accrual period, 295 attending physicians admitted 6805 unique patients for a total of 10,359 admissions. Seventeen percent of these 10,359 admissions were readmitted within 30 days. The cohort of all patients had a mean age of 59.6 years 19.5 standard deviation (SD), with 52.8% women. The mean length of stay was 5.6 days 10.4 SD. Medicare was the payer source for approximately half of the admissions. The majority of admissions (90.4%) were billed for at least 1 high risk medication, with narcotics, cardiac medications, and antibiotics being the most common. Regarding disposition, 79.5% of admissions were discharged to home; 9.1% were discharged to a skilled nursing facility (SNF).

Baseline sociodemographic, operational, and clinical characteristics for patients readmitted and not readmitted are shown in Table 1. Demographic characteristics with significant differences (P < 0.05) between readmitted and nonreadmitted groups included mean age, race, payer status, and primary language other than English. Regarding operational characteristics, readmitted patients had a higher median length of stay and were more likely to be admitted through the emergency room during their index admission. Discharge to an SNF was higher in the readmitted group versus the nonreadmitted group (9.7% vs. 9.0%). Several clinical factors were more prevalent in the readmitted group: high‐risk medications, specifically steroids, narcotics, and cardiovascular medications; high‐risk medication count of 3 or greater; and comorbidities including congestive heart failure, renal disease, cancer, anemia, and depression.

Frequency of Readmission

The 30‐day readmission rate was 17.0% (1762 patients), with 49.7% (875 patients) of the readmissions occurring within 10 days of discharge. Of patients readmitted, the general medicine service was the readmitting team in 78.2%. A quarter of readmissions (26.2%) had the same primary diagnosis on initial and repeat admission.

Factors Associated With Readmission

Factors associated with readmission were categorized as sociodemographic, operational, and clinical. Factors associated with readmission with P < 0.05 and present in at least 5% of admissions are presented in Table 2. Of the sociodemographic factors, black race was significantly associated with readmission. Within the Medicare cohort, risk for readmission was similar for white vs. nonwhite race, with relative risk of 1.0 (95% confidence interval [CI], 0.86‐1.18). Medicaid as payer status was significantly associated in the unadjusted model, and in the adjusted model showed a trend toward readmission. Mean age was significantly different in the readmitted and nonreadmitted groups, but the difference was small (1.0 year). Moreover, when we evaluated age in 5‐year categories (ex. 65‐70, 71‐75, etc.), age was not associated with readmission. In the adjusted model, none of the operational factors were significantly associated with readmission, including discharge to SNF, weekend discharge, or admit source.

Of the clinical factors, high‐risk medications and 6 comorbidities were associated with readmission. High‐risk medication categories associated with readmission were steroids and narcotics; anticholinergics medications were protective. The 6 comorbidities associated with readmission were congestive heart failure, renal disease, cancer (with and without metastasis), weight loss, and iron deficiency anemia. While APR risk of mortality was associated with readmission at P < 0.05, including APR in our final model did not alter which other factors were significantly associated with readmission. When site (Moffitt‐Long vs. Mount Zion Hospitals) was added to the model, the ORs for factors associated with readmission did not change appreciably (0.01).