Hospitalization may offer a natural opportunity to screen and advise patients on the advantages of quitting smoking due to a variety of reasons, such as the smoke‐free environment, availability of medical personnel, suitability of tailoring information, and the potential to catch a teachable moment.1, 2 Additionally, a recent meta‐analysis suggested that hospital‐based cessation programs and referrals to cardiac rehabilitation result in significantly higher rates of cessation among discharged smokers.3 In 2008, the U.S. Public Health Service Task Force on Clinical Practice Guidelines for Treating Tobacco Use and Dependence in hospitalized smokers recommended listing smoking status on problem lists, evaluating a smoker's preparedness to quit, providing counseling and medications to treat in‐hospital withdrawal symptoms, and arranging discharge follow‐up to help smokers remain abstinent.4 To promote these practices, the Center for Medicaid and Medicare Services (CMS) has made smoking cessation counseling a quality of care indicator for patients hospitalized with congestive heart failure (CHF), acute myocardial infarction (AMI), or pneumonia. This indicator is a critical step in recognizing the importance of smoking cessation counseling in improving mortality and morbidity for these patients.

Despite the importance of promoting smoking cessation among hospitalized patients, few studies have looked at whether or not hospitalized patients are prepared to quit smoking. Ascertaining patients' preparedness to quit smoking is an important first step in understanding a patient's readiness to change their health behaviors because smoking cessation is the culmination of a lengthy process of behavior change.5 Studies of healthy factory workers suggest that smokers who were more prepared to quit smoking had a higher number of previous quit attempts and perceived coworker encouragement.6

Understanding patient preparedness to quit smoking is especially important among African American smokers, who face a disproportionate health burden due to smoking‐related illness. Studies show that African Americans are less likely than other racial groups to engage in formal tobacco cessation interventions and have lower long‐term quit rates, despite a higher desire to quit smoking.5, 79 Understanding preparedness to quit among this particular group of hospitalized patients may be an important first step in identifying those most likely to quit and benefit from tailored, intensive interventions, such as using medications to assist in combination with postdischarge tobacco cessation counseling.

The aim of this study was to characterize the preparedness to quit smoking and to assess quit attempts made, methods used for quitting, and the success of such quit attempts at 1‐month follow‐up in a group comprised of a high proportion of underserved African American hospitalized smokers. In addition, the relationship of hospitalized patients' preparedness to quit and the effect of inpatient advice on the likelihood of subsequent tobacco cessation were examined.

Patients and Methods

The data used for this study were collected for the Cardiology Quality of Care Study, an ongoing prospective study of patients hospitalized on the inpatient cardiology service at the University of Chicago Medical Center. Newly admitted patients were approached by research assistants and consented to the study using a previously described protocol for enrolling hospitalized patients.10 Patients that lacked decisional capacity (score of <17 on the telephone version of the Mini‐Mental Status Exam)11 were excluded. Patients did not receive any scripted intervention during this admission to assist with cessation. The study left cessation counseling and advice to quit up to the discretion of the individual physician caring for the patient in the hospital. The Institutional Review Board at the University of Chicago approved this study.

Results

From February 2006 through July 2007, 86% (2906/3364) of all cardiology inpatients approached were interviewed. Fifteen percent (436/2906) of patients enrolled in the study indicated that they were current smokers. Contemplation Ladder scores were obtained on 95% (415/436) of the current smokers, and 1‐month postdischarge follow‐up telephone surveys were completed in 67% (276/415) of the current smokers. Three attempts were made to contact patients who were lost to follow‐up (Figure 1). The major reasons for inability to contact patients included wrong telephone numbers, disconnected phone lines, or no method to leave a message for the patient (ie, no answering machine). Given that we were only able to complete follow‐up interviews on 276 patients, we conducted our analyses on only this group of patients.

Figure 1

The average age of current smokers in the sample was 55 years (95% confidence interval [CI], 54‐58). Most current smokers were of the African American race (83%; 224/276). More than 65% of smokers had completed high school or higher, and nearly one‐half (46%) had an average household income of $25,000 or less before taxes. The most common admitting diagnoses per chart audit among current inpatient smokers were AMI (31%) and CHF (27%). The vast majority (95%) of hospitalized smokers in this sample were first‐time admissions to the University of Chicago. Table 1 shows the demographic data for current smokers compared to former smokers (those who have quit smoking prior to admission). Current smokers were more likely to be African American, had lower income levels, and were less likely to have completed high school. Additionally, current smokers were more likely to carry a potential diagnosis of AMI or CHF and to be a first‐time admission (Table 1).

Approximately three‐quarters (76%; 210/276) of current smokers were identified as prepared to quit, with a Ladder score 6. There was a wide distribution of Ladder scores, with one‐third (31%; 86/276) of smokers reporting a Ladder score of 8, indicating that they still smoke, but are ready to set a quit date and another 34% (95/276 patients) with Ladder scores of either 6 or 7 also indicating they were planning to quit smoking (Figure 2). A significant portion of smokers (71%; 195/276) reported making a quit attempt after discharge, and 38% of smokers (106/276) self‐reported that their quit attempt was successful (ie, no longer smoking at 1 month post discharge). Note that the quit rate is reduced to 26% (106/415) at 1 month if one conservatively assumes that those who did not take part in follow‐up were relapsers. Among those who did participate in follow‐up, as shown in Figure 3, the most frequently reported (53%; 145/276) method used to quit smoking was cold turkey. Thirteen percent (37/276) of patients reported making a quit attempt using pharmacological therapy (ie, NRT or bupropion) and only 4% (12/276) of patients reported making a quit attempt using the help of a smoking cessation program (Figure 3).

Figure 2

Figure 3

Preparedness was an important predictor of making a quit attempt. Prepared patients (ie, Ladder score 6) were significantly more likely than patients who were less prepared to report making a quit attempt after discharge (163/212 [77%] vs. 32/64 [50%], respectively; P < 0.001). This result remained significant after adjusting for sociodemographic characteristics with a similar effect size (adjusted estimates 76% [95% CI, 75.7‐76.7] prepared vs. 49% [95% CI, 48.5‐49.8]; P < 0.001). These results also remained significant with a similar effect size in analyses using multivariate logistic regression (Table 2). Of those patients who made quit attempts, prepared patients were slightly more likely to report a successful quit attempt (90/163; 55%) than were less‐prepared patients (16/32; 50%), though this was not significant (P = 0.205).

In the follow‐up sample, 17% could not remember if they received advice to quit smoking. Among those who were able to recall receiving advice, the majority (78%; 180/230) reported that they received advice from a nurse or physician during hospitalization, compared to 22% who did not recall ever being advised to quit by any healthcare provider during the admission. Patients who reported receiving advice to quit were more likely to report making a quit attempt postdischarge as compared to those that did not recall receiving advice (70% vs. 46%, respectively; P = 0.002). In a multivariate logistic regression, controlling for demographic factors and admitting diagnosis, both preparedness and receipt of in‐hospital advice were independent predictors of making a future quit attempt (odds ratio [OR] = 4.05; 95% CI, 1.91‐8.60; P < 0.001 for preparedness; OR = 3.96; 95% CI, 1.84‐8.54; P < 0.001 for advice). Additionally, there was no significant interaction or synergistic effect between being prepared to quit smoking and receiving in‐hospital advice to quit (OR = 1.24; 95% CI, 0.17‐9.21; P = 0.836) (Figure 4). When analyzing the effects of preparedness and advice on quit attempts, only preparedness to quit remained a significant predictor of a successful quit attempt (OR = 2.93; 95% CI, 1.13‐7.60; P = 0.027 for preparedness; OR = 2.16; 95% CI, 0.85‐5.49; P = 0.10 for advice to quit). As demonstrated in Table 2, a higher percentage of prepared patients made a quit attempt after discharge (76.9% vs. 50%) and had a successful quit attempt and short‐term abstinence (55% prepared patients vs. 50% less prepared patients).

Figure 4

Discussion

This study demonstrated that in a group of hospitalized underserved, and predominantly African American smokers, the majority of patients reported being prepared to quit smoking at the time of hospitalization. Prepared patients were more likely to report making a quit attempt after discharge and more likely to report being successful in their quit attempt than patients who reported being less prepared to quit during their hospitalization. Nevertheless, approximately one‐half of unprepared patients did make a quit attempt 1 month after discharge, demonstrating a desire to quit smoking after hospitalization among this population. However, short‐term success rates in this group were lower than in patients prepared to quit. In addition, preparedness to quit and receipt of in‐hospital advice to quit smoking were both found to be independent predictors of making a quit attempt, with nearly identical ORs; however, only preparedness remained significant after controlling for advice to quit. Last, although the majority of hospitalized cardiac patients were making quit attempts after discharge, most patients reported using the least effective quit methods (ie, cold turkey) rather than more effective and intensive interventions such as counseling in combination with pharmacotherapy.

These findings have important implications for current quality initiatives targeted at promoting smoking cessation among cardiac patients. First, these results highlight the need for evidence‐based methods to be made available to hospitalized smokers who are prepared to quit. Our results are consistent with other studies reporting rare use (5.2%) of NRT in the hospital setting, despite the proven benefit in treating nicotine withdrawal symptoms.17 This is also consistent with data reporting that among nonhospitalized smokers, quitting cold turkey was the most commonly used and least effective cessation method.18 Second, the rate of recall of in‐hospital advice among patients (78%) was generally consistent with those reported to CMS (most recent quarter 95% for AMI and 88% for CHF).19

In addition to receiving advice, preparedness to quit was associated with higher quit attempts, therefore highlighting the importance of assessing level of preparedness in addition to giving advice. The fact that most quit attempts were made using cold turkey and resulted in low short‐term success rates underscores the need to reevaluate the current CMS quality indicator of advice alone for hospitalized smokers. Furthermore, the recently updated 2008 U.S. Public Health guidelines recently recommend, in addition to advice, that all hospitalized smokers be assessed for readiness to change, be assisted in quitting with pharmacotherapy, and be arranged follow‐up for tobacco cessation postdischarge, highlighting the inadequacy of advice alone.4 While it is important to continue to advise all hospitalized smokers to quit, the study findings demonstrate that assessing preparedness may result in targeting more prepared patients with more intensive interventions. Further policy implications include that less prepared patients may need motivational techniques to increase their level of preparedness to quit during hospitalization.

Several limitations are worth mentioning. First, the study included a relatively small sample size drawn from a single urban medical center. The prevalence of current smokers in our sample was 15%, which is lower than many studies looking at cardiology inpatient smokers.3, 20 This limitation of our study may be attributed to the advanced age of the majority of our patients, as compared with other studies, as well as the possibility of socially desirable response bias that many low‐income African American smokers may experience, leaving them less likely to admit to smoking at the time of hospitalization. Second, there was a low follow‐up rate, with 66% of patients undergoing follow‐up postdischarge. While this may raise the concerns of differences between ladder scores in those patients that participated in follow‐up and those that did not, analyses show no significant difference between level of preparedness in these 2 groups (68% prepared in patients who received follow‐up vs. 63% prepared patients in those who did not participate in follow‐up; P = 0.36). Third, follow‐up of quit attempts and receipt of advice were all assessed using self‐report, and, therefore, were limited by lack of verification and lack of assessment for potential recall bias. Fourth, in this pilot study, the follow‐up period was relatively short at 1 month postdischarge. It is likely that rates of successful quit attempts would be lower with longer‐term follow‐up periods, given previous literature demonstrating the difficulty with long‐term abstinence.21 Last, the study was not able to account for potential effects that hospitalization itself may have on preparedness, as patients may be more likely to report being prepared to quit when in the face of a health shock,22 as well as the fact that some patients may demonstrate a socially desirable response bias influenced by hospitalization.

In conclusion, the majority of underserved smokers with cardiac disease reported being prepared to quit smoking and were more likely to self‐report making a quit attempt after discharge. However, the majority of these quit attempts were made via cold turkey, without the support of available evidence‐based methods to quit. It is possible that by directly providing education, access to pharmacotherapy, and counseling options, the utilization rates for more efficacious treatments would increase in cardiac patients who are prepared to quit. While recall of in‐hospital advice was associated with future quit attempts, prepared patients who recalled receiving advice were more likely to make a quit attempt than prepared patients who did not recall receiving advice, as well as unprepared patients. Together, these findings highlight the need to consider a patient's level of preparedness to quit in understanding the success of in‐hospital advice and the importance of making evidence‐based cessation methods available to hospitalized smokers who are prepared to quit. Additionally, identifying patients not prepared to quit may help in providing them with appropriate motivational therapy, to move them along the stages of change, as well as educational information on how to quit once they have decided to do so.

Hospitalization may offer a natural opportunity to screen and advise patients on the advantages of quitting smoking due to a variety of reasons, such as the smoke‐free environment, availability of medical personnel, suitability of tailoring information, and the potential to catch a teachable moment.1, 2 Additionally, a recent meta‐analysis suggested that hospital‐based cessation programs and referrals to cardiac rehabilitation result in significantly higher rates of cessation among discharged smokers.3 In 2008, the U.S. Public Health Service Task Force on Clinical Practice Guidelines for Treating Tobacco Use and Dependence in hospitalized smokers recommended listing smoking status on problem lists, evaluating a smoker's preparedness to quit, providing counseling and medications to treat in‐hospital withdrawal symptoms, and arranging discharge follow‐up to help smokers remain abstinent.4 To promote these practices, the Center for Medicaid and Medicare Services (CMS) has made smoking cessation counseling a quality of care indicator for patients hospitalized with congestive heart failure (CHF), acute myocardial infarction (AMI), or pneumonia. This indicator is a critical step in recognizing the importance of smoking cessation counseling in improving mortality and morbidity for these patients.

Despite the importance of promoting smoking cessation among hospitalized patients, few studies have looked at whether or not hospitalized patients are prepared to quit smoking. Ascertaining patients' preparedness to quit smoking is an important first step in understanding a patient's readiness to change their health behaviors because smoking cessation is the culmination of a lengthy process of behavior change.5 Studies of healthy factory workers suggest that smokers who were more prepared to quit smoking had a higher number of previous quit attempts and perceived coworker encouragement.6

Understanding patient preparedness to quit smoking is especially important among African American smokers, who face a disproportionate health burden due to smoking‐related illness. Studies show that African Americans are less likely than other racial groups to engage in formal tobacco cessation interventions and have lower long‐term quit rates, despite a higher desire to quit smoking.5, 79 Understanding preparedness to quit among this particular group of hospitalized patients may be an important first step in identifying those most likely to quit and benefit from tailored, intensive interventions, such as using medications to assist in combination with postdischarge tobacco cessation counseling.

The aim of this study was to characterize the preparedness to quit smoking and to assess quit attempts made, methods used for quitting, and the success of such quit attempts at 1‐month follow‐up in a group comprised of a high proportion of underserved African American hospitalized smokers. In addition, the relationship of hospitalized patients' preparedness to quit and the effect of inpatient advice on the likelihood of subsequent tobacco cessation were examined.

Patients and Methods

The data used for this study were collected for the Cardiology Quality of Care Study, an ongoing prospective study of patients hospitalized on the inpatient cardiology service at the University of Chicago Medical Center. Newly admitted patients were approached by research assistants and consented to the study using a previously described protocol for enrolling hospitalized patients.10 Patients that lacked decisional capacity (score of <17 on the telephone version of the Mini‐Mental Status Exam)11 were excluded. Patients did not receive any scripted intervention during this admission to assist with cessation. The study left cessation counseling and advice to quit up to the discretion of the individual physician caring for the patient in the hospital. The Institutional Review Board at the University of Chicago approved this study.

Results

From February 2006 through July 2007, 86% (2906/3364) of all cardiology inpatients approached were interviewed. Fifteen percent (436/2906) of patients enrolled in the study indicated that they were current smokers. Contemplation Ladder scores were obtained on 95% (415/436) of the current smokers, and 1‐month postdischarge follow‐up telephone surveys were completed in 67% (276/415) of the current smokers. Three attempts were made to contact patients who were lost to follow‐up (Figure 1). The major reasons for inability to contact patients included wrong telephone numbers, disconnected phone lines, or no method to leave a message for the patient (ie, no answering machine). Given that we were only able to complete follow‐up interviews on 276 patients, we conducted our analyses on only this group of patients.

Figure 1

The average age of current smokers in the sample was 55 years (95% confidence interval [CI], 54‐58). Most current smokers were of the African American race (83%; 224/276). More than 65% of smokers had completed high school or higher, and nearly one‐half (46%) had an average household income of $25,000 or less before taxes. The most common admitting diagnoses per chart audit among current inpatient smokers were AMI (31%) and CHF (27%). The vast majority (95%) of hospitalized smokers in this sample were first‐time admissions to the University of Chicago. Table 1 shows the demographic data for current smokers compared to former smokers (those who have quit smoking prior to admission). Current smokers were more likely to be African American, had lower income levels, and were less likely to have completed high school. Additionally, current smokers were more likely to carry a potential diagnosis of AMI or CHF and to be a first‐time admission (Table 1).

Approximately three‐quarters (76%; 210/276) of current smokers were identified as prepared to quit, with a Ladder score 6. There was a wide distribution of Ladder scores, with one‐third (31%; 86/276) of smokers reporting a Ladder score of 8, indicating that they still smoke, but are ready to set a quit date and another 34% (95/276 patients) with Ladder scores of either 6 or 7 also indicating they were planning to quit smoking (Figure 2). A significant portion of smokers (71%; 195/276) reported making a quit attempt after discharge, and 38% of smokers (106/276) self‐reported that their quit attempt was successful (ie, no longer smoking at 1 month post discharge). Note that the quit rate is reduced to 26% (106/415) at 1 month if one conservatively assumes that those who did not take part in follow‐up were relapsers. Among those who did participate in follow‐up, as shown in Figure 3, the most frequently reported (53%; 145/276) method used to quit smoking was cold turkey. Thirteen percent (37/276) of patients reported making a quit attempt using pharmacological therapy (ie, NRT or bupropion) and only 4% (12/276) of patients reported making a quit attempt using the help of a smoking cessation program (Figure 3).

Figure 2

Figure 3

Preparedness was an important predictor of making a quit attempt. Prepared patients (ie, Ladder score 6) were significantly more likely than patients who were less prepared to report making a quit attempt after discharge (163/212 [77%] vs. 32/64 [50%], respectively; P < 0.001). This result remained significant after adjusting for sociodemographic characteristics with a similar effect size (adjusted estimates 76% [95% CI, 75.7‐76.7] prepared vs. 49% [95% CI, 48.5‐49.8]; P < 0.001). These results also remained significant with a similar effect size in analyses using multivariate logistic regression (Table 2). Of those patients who made quit attempts, prepared patients were slightly more likely to report a successful quit attempt (90/163; 55%) than were less‐prepared patients (16/32; 50%), though this was not significant (P = 0.205).

In the follow‐up sample, 17% could not remember if they received advice to quit smoking. Among those who were able to recall receiving advice, the majority (78%; 180/230) reported that they received advice from a nurse or physician during hospitalization, compared to 22% who did not recall ever being advised to quit by any healthcare provider during the admission. Patients who reported receiving advice to quit were more likely to report making a quit attempt postdischarge as compared to those that did not recall receiving advice (70% vs. 46%, respectively; P = 0.002). In a multivariate logistic regression, controlling for demographic factors and admitting diagnosis, both preparedness and receipt of in‐hospital advice were independent predictors of making a future quit attempt (odds ratio [OR] = 4.05; 95% CI, 1.91‐8.60; P < 0.001 for preparedness; OR = 3.96; 95% CI, 1.84‐8.54; P < 0.001 for advice). Additionally, there was no significant interaction or synergistic effect between being prepared to quit smoking and receiving in‐hospital advice to quit (OR = 1.24; 95% CI, 0.17‐9.21; P = 0.836) (Figure 4). When analyzing the effects of preparedness and advice on quit attempts, only preparedness to quit remained a significant predictor of a successful quit attempt (OR = 2.93; 95% CI, 1.13‐7.60; P = 0.027 for preparedness; OR = 2.16; 95% CI, 0.85‐5.49; P = 0.10 for advice to quit). As demonstrated in Table 2, a higher percentage of prepared patients made a quit attempt after discharge (76.9% vs. 50%) and had a successful quit attempt and short‐term abstinence (55% prepared patients vs. 50% less prepared patients).

Figure 4

Discussion

This study demonstrated that in a group of hospitalized underserved, and predominantly African American smokers, the majority of patients reported being prepared to quit smoking at the time of hospitalization. Prepared patients were more likely to report making a quit attempt after discharge and more likely to report being successful in their quit attempt than patients who reported being less prepared to quit during their hospitalization. Nevertheless, approximately one‐half of unprepared patients did make a quit attempt 1 month after discharge, demonstrating a desire to quit smoking after hospitalization among this population. However, short‐term success rates in this group were lower than in patients prepared to quit. In addition, preparedness to quit and receipt of in‐hospital advice to quit smoking were both found to be independent predictors of making a quit attempt, with nearly identical ORs; however, only preparedness remained significant after controlling for advice to quit. Last, although the majority of hospitalized cardiac patients were making quit attempts after discharge, most patients reported using the least effective quit methods (ie, cold turkey) rather than more effective and intensive interventions such as counseling in combination with pharmacotherapy.

These findings have important implications for current quality initiatives targeted at promoting smoking cessation among cardiac patients. First, these results highlight the need for evidence‐based methods to be made available to hospitalized smokers who are prepared to quit. Our results are consistent with other studies reporting rare use (5.2%) of NRT in the hospital setting, despite the proven benefit in treating nicotine withdrawal symptoms.17 This is also consistent with data reporting that among nonhospitalized smokers, quitting cold turkey was the most commonly used and least effective cessation method.18 Second, the rate of recall of in‐hospital advice among patients (78%) was generally consistent with those reported to CMS (most recent quarter 95% for AMI and 88% for CHF).19

In addition to receiving advice, preparedness to quit was associated with higher quit attempts, therefore highlighting the importance of assessing level of preparedness in addition to giving advice. The fact that most quit attempts were made using cold turkey and resulted in low short‐term success rates underscores the need to reevaluate the current CMS quality indicator of advice alone for hospitalized smokers. Furthermore, the recently updated 2008 U.S. Public Health guidelines recently recommend, in addition to advice, that all hospitalized smokers be assessed for readiness to change, be assisted in quitting with pharmacotherapy, and be arranged follow‐up for tobacco cessation postdischarge, highlighting the inadequacy of advice alone.4 While it is important to continue to advise all hospitalized smokers to quit, the study findings demonstrate that assessing preparedness may result in targeting more prepared patients with more intensive interventions. Further policy implications include that less prepared patients may need motivational techniques to increase their level of preparedness to quit during hospitalization.

Several limitations are worth mentioning. First, the study included a relatively small sample size drawn from a single urban medical center. The prevalence of current smokers in our sample was 15%, which is lower than many studies looking at cardiology inpatient smokers.3, 20 This limitation of our study may be attributed to the advanced age of the majority of our patients, as compared with other studies, as well as the possibility of socially desirable response bias that many low‐income African American smokers may experience, leaving them less likely to admit to smoking at the time of hospitalization. Second, there was a low follow‐up rate, with 66% of patients undergoing follow‐up postdischarge. While this may raise the concerns of differences between ladder scores in those patients that participated in follow‐up and those that did not, analyses show no significant difference between level of preparedness in these 2 groups (68% prepared in patients who received follow‐up vs. 63% prepared patients in those who did not participate in follow‐up; P = 0.36). Third, follow‐up of quit attempts and receipt of advice were all assessed using self‐report, and, therefore, were limited by lack of verification and lack of assessment for potential recall bias. Fourth, in this pilot study, the follow‐up period was relatively short at 1 month postdischarge. It is likely that rates of successful quit attempts would be lower with longer‐term follow‐up periods, given previous literature demonstrating the difficulty with long‐term abstinence.21 Last, the study was not able to account for potential effects that hospitalization itself may have on preparedness, as patients may be more likely to report being prepared to quit when in the face of a health shock,22 as well as the fact that some patients may demonstrate a socially desirable response bias influenced by hospitalization.

In conclusion, the majority of underserved smokers with cardiac disease reported being prepared to quit smoking and were more likely to self‐report making a quit attempt after discharge. However, the majority of these quit attempts were made via cold turkey, without the support of available evidence‐based methods to quit. It is possible that by directly providing education, access to pharmacotherapy, and counseling options, the utilization rates for more efficacious treatments would increase in cardiac patients who are prepared to quit. While recall of in‐hospital advice was associated with future quit attempts, prepared patients who recalled receiving advice were more likely to make a quit attempt than prepared patients who did not recall receiving advice, as well as unprepared patients. Together, these findings highlight the need to consider a patient's level of preparedness to quit in understanding the success of in‐hospital advice and the importance of making evidence‐based cessation methods available to hospitalized smokers who are prepared to quit. Additionally, identifying patients not prepared to quit may help in providing them with appropriate motivational therapy, to move them along the stages of change, as well as educational information on how to quit once they have decided to do so.

Hospitalization may offer a natural opportunity to screen and advise patients on the advantages of quitting smoking due to a variety of reasons, such as the smoke‐free environment, availability of medical personnel, suitability of tailoring information, and the potential to catch a teachable moment.1, 2 Additionally, a recent meta‐analysis suggested that hospital‐based cessation programs and referrals to cardiac rehabilitation result in significantly higher rates of cessation among discharged smokers.3 In 2008, the U.S. Public Health Service Task Force on Clinical Practice Guidelines for Treating Tobacco Use and Dependence in hospitalized smokers recommended listing smoking status on problem lists, evaluating a smoker's preparedness to quit, providing counseling and medications to treat in‐hospital withdrawal symptoms, and arranging discharge follow‐up to help smokers remain abstinent.4 To promote these practices, the Center for Medicaid and Medicare Services (CMS) has made smoking cessation counseling a quality of care indicator for patients hospitalized with congestive heart failure (CHF), acute myocardial infarction (AMI), or pneumonia. This indicator is a critical step in recognizing the importance of smoking cessation counseling in improving mortality and morbidity for these patients.

Despite the importance of promoting smoking cessation among hospitalized patients, few studies have looked at whether or not hospitalized patients are prepared to quit smoking. Ascertaining patients' preparedness to quit smoking is an important first step in understanding a patient's readiness to change their health behaviors because smoking cessation is the culmination of a lengthy process of behavior change.5 Studies of healthy factory workers suggest that smokers who were more prepared to quit smoking had a higher number of previous quit attempts and perceived coworker encouragement.6

Understanding patient preparedness to quit smoking is especially important among African American smokers, who face a disproportionate health burden due to smoking‐related illness. Studies show that African Americans are less likely than other racial groups to engage in formal tobacco cessation interventions and have lower long‐term quit rates, despite a higher desire to quit smoking.5, 79 Understanding preparedness to quit among this particular group of hospitalized patients may be an important first step in identifying those most likely to quit and benefit from tailored, intensive interventions, such as using medications to assist in combination with postdischarge tobacco cessation counseling.

The aim of this study was to characterize the preparedness to quit smoking and to assess quit attempts made, methods used for quitting, and the success of such quit attempts at 1‐month follow‐up in a group comprised of a high proportion of underserved African American hospitalized smokers. In addition, the relationship of hospitalized patients' preparedness to quit and the effect of inpatient advice on the likelihood of subsequent tobacco cessation were examined.

Patients and Methods

The data used for this study were collected for the Cardiology Quality of Care Study, an ongoing prospective study of patients hospitalized on the inpatient cardiology service at the University of Chicago Medical Center. Newly admitted patients were approached by research assistants and consented to the study using a previously described protocol for enrolling hospitalized patients.10 Patients that lacked decisional capacity (score of <17 on the telephone version of the Mini‐Mental Status Exam)11 were excluded. Patients did not receive any scripted intervention during this admission to assist with cessation. The study left cessation counseling and advice to quit up to the discretion of the individual physician caring for the patient in the hospital. The Institutional Review Board at the University of Chicago approved this study.

Results

From February 2006 through July 2007, 86% (2906/3364) of all cardiology inpatients approached were interviewed. Fifteen percent (436/2906) of patients enrolled in the study indicated that they were current smokers. Contemplation Ladder scores were obtained on 95% (415/436) of the current smokers, and 1‐month postdischarge follow‐up telephone surveys were completed in 67% (276/415) of the current smokers. Three attempts were made to contact patients who were lost to follow‐up (Figure 1). The major reasons for inability to contact patients included wrong telephone numbers, disconnected phone lines, or no method to leave a message for the patient (ie, no answering machine). Given that we were only able to complete follow‐up interviews on 276 patients, we conducted our analyses on only this group of patients.

Figure 1

The average age of current smokers in the sample was 55 years (95% confidence interval [CI], 54‐58). Most current smokers were of the African American race (83%; 224/276). More than 65% of smokers had completed high school or higher, and nearly one‐half (46%) had an average household income of $25,000 or less before taxes. The most common admitting diagnoses per chart audit among current inpatient smokers were AMI (31%) and CHF (27%). The vast majority (95%) of hospitalized smokers in this sample were first‐time admissions to the University of Chicago. Table 1 shows the demographic data for current smokers compared to former smokers (those who have quit smoking prior to admission). Current smokers were more likely to be African American, had lower income levels, and were less likely to have completed high school. Additionally, current smokers were more likely to carry a potential diagnosis of AMI or CHF and to be a first‐time admission (Table 1).

Approximately three‐quarters (76%; 210/276) of current smokers were identified as prepared to quit, with a Ladder score 6. There was a wide distribution of Ladder scores, with one‐third (31%; 86/276) of smokers reporting a Ladder score of 8, indicating that they still smoke, but are ready to set a quit date and another 34% (95/276 patients) with Ladder scores of either 6 or 7 also indicating they were planning to quit smoking (Figure 2). A significant portion of smokers (71%; 195/276) reported making a quit attempt after discharge, and 38% of smokers (106/276) self‐reported that their quit attempt was successful (ie, no longer smoking at 1 month post discharge). Note that the quit rate is reduced to 26% (106/415) at 1 month if one conservatively assumes that those who did not take part in follow‐up were relapsers. Among those who did participate in follow‐up, as shown in Figure 3, the most frequently reported (53%; 145/276) method used to quit smoking was cold turkey. Thirteen percent (37/276) of patients reported making a quit attempt using pharmacological therapy (ie, NRT or bupropion) and only 4% (12/276) of patients reported making a quit attempt using the help of a smoking cessation program (Figure 3).

Figure 2

Figure 3

Preparedness was an important predictor of making a quit attempt. Prepared patients (ie, Ladder score 6) were significantly more likely than patients who were less prepared to report making a quit attempt after discharge (163/212 [77%] vs. 32/64 [50%], respectively; P < 0.001). This result remained significant after adjusting for sociodemographic characteristics with a similar effect size (adjusted estimates 76% [95% CI, 75.7‐76.7] prepared vs. 49% [95% CI, 48.5‐49.8]; P < 0.001). These results also remained significant with a similar effect size in analyses using multivariate logistic regression (Table 2). Of those patients who made quit attempts, prepared patients were slightly more likely to report a successful quit attempt (90/163; 55%) than were less‐prepared patients (16/32; 50%), though this was not significant (P = 0.205).

In the follow‐up sample, 17% could not remember if they received advice to quit smoking. Among those who were able to recall receiving advice, the majority (78%; 180/230) reported that they received advice from a nurse or physician during hospitalization, compared to 22% who did not recall ever being advised to quit by any healthcare provider during the admission. Patients who reported receiving advice to quit were more likely to report making a quit attempt postdischarge as compared to those that did not recall receiving advice (70% vs. 46%, respectively; P = 0.002). In a multivariate logistic regression, controlling for demographic factors and admitting diagnosis, both preparedness and receipt of in‐hospital advice were independent predictors of making a future quit attempt (odds ratio [OR] = 4.05; 95% CI, 1.91‐8.60; P < 0.001 for preparedness; OR = 3.96; 95% CI, 1.84‐8.54; P < 0.001 for advice). Additionally, there was no significant interaction or synergistic effect between being prepared to quit smoking and receiving in‐hospital advice to quit (OR = 1.24; 95% CI, 0.17‐9.21; P = 0.836) (Figure 4). When analyzing the effects of preparedness and advice on quit attempts, only preparedness to quit remained a significant predictor of a successful quit attempt (OR = 2.93; 95% CI, 1.13‐7.60; P = 0.027 for preparedness; OR = 2.16; 95% CI, 0.85‐5.49; P = 0.10 for advice to quit). As demonstrated in Table 2, a higher percentage of prepared patients made a quit attempt after discharge (76.9% vs. 50%) and had a successful quit attempt and short‐term abstinence (55% prepared patients vs. 50% less prepared patients).

Figure 4

Discussion

This study demonstrated that in a group of hospitalized underserved, and predominantly African American smokers, the majority of patients reported being prepared to quit smoking at the time of hospitalization. Prepared patients were more likely to report making a quit attempt after discharge and more likely to report being successful in their quit attempt than patients who reported being less prepared to quit during their hospitalization. Nevertheless, approximately one‐half of unprepared patients did make a quit attempt 1 month after discharge, demonstrating a desire to quit smoking after hospitalization among this population. However, short‐term success rates in this group were lower than in patients prepared to quit. In addition, preparedness to quit and receipt of in‐hospital advice to quit smoking were both found to be independent predictors of making a quit attempt, with nearly identical ORs; however, only preparedness remained significant after controlling for advice to quit. Last, although the majority of hospitalized cardiac patients were making quit attempts after discharge, most patients reported using the least effective quit methods (ie, cold turkey) rather than more effective and intensive interventions such as counseling in combination with pharmacotherapy.

These findings have important implications for current quality initiatives targeted at promoting smoking cessation among cardiac patients. First, these results highlight the need for evidence‐based methods to be made available to hospitalized smokers who are prepared to quit. Our results are consistent with other studies reporting rare use (5.2%) of NRT in the hospital setting, despite the proven benefit in treating nicotine withdrawal symptoms.17 This is also consistent with data reporting that among nonhospitalized smokers, quitting cold turkey was the most commonly used and least effective cessation method.18 Second, the rate of recall of in‐hospital advice among patients (78%) was generally consistent with those reported to CMS (most recent quarter 95% for AMI and 88% for CHF).19

In addition to receiving advice, preparedness to quit was associated with higher quit attempts, therefore highlighting the importance of assessing level of preparedness in addition to giving advice. The fact that most quit attempts were made using cold turkey and resulted in low short‐term success rates underscores the need to reevaluate the current CMS quality indicator of advice alone for hospitalized smokers. Furthermore, the recently updated 2008 U.S. Public Health guidelines recently recommend, in addition to advice, that all hospitalized smokers be assessed for readiness to change, be assisted in quitting with pharmacotherapy, and be arranged follow‐up for tobacco cessation postdischarge, highlighting the inadequacy of advice alone.4 While it is important to continue to advise all hospitalized smokers to quit, the study findings demonstrate that assessing preparedness may result in targeting more prepared patients with more intensive interventions. Further policy implications include that less prepared patients may need motivational techniques to increase their level of preparedness to quit during hospitalization.

Several limitations are worth mentioning. First, the study included a relatively small sample size drawn from a single urban medical center. The prevalence of current smokers in our sample was 15%, which is lower than many studies looking at cardiology inpatient smokers.3, 20 This limitation of our study may be attributed to the advanced age of the majority of our patients, as compared with other studies, as well as the possibility of socially desirable response bias that many low‐income African American smokers may experience, leaving them less likely to admit to smoking at the time of hospitalization. Second, there was a low follow‐up rate, with 66% of patients undergoing follow‐up postdischarge. While this may raise the concerns of differences between ladder scores in those patients that participated in follow‐up and those that did not, analyses show no significant difference between level of preparedness in these 2 groups (68% prepared in patients who received follow‐up vs. 63% prepared patients in those who did not participate in follow‐up; P = 0.36). Third, follow‐up of quit attempts and receipt of advice were all assessed using self‐report, and, therefore, were limited by lack of verification and lack of assessment for potential recall bias. Fourth, in this pilot study, the follow‐up period was relatively short at 1 month postdischarge. It is likely that rates of successful quit attempts would be lower with longer‐term follow‐up periods, given previous literature demonstrating the difficulty with long‐term abstinence.21 Last, the study was not able to account for potential effects that hospitalization itself may have on preparedness, as patients may be more likely to report being prepared to quit when in the face of a health shock,22 as well as the fact that some patients may demonstrate a socially desirable response bias influenced by hospitalization.

In conclusion, the majority of underserved smokers with cardiac disease reported being prepared to quit smoking and were more likely to self‐report making a quit attempt after discharge. However, the majority of these quit attempts were made via cold turkey, without the support of available evidence‐based methods to quit. It is possible that by directly providing education, access to pharmacotherapy, and counseling options, the utilization rates for more efficacious treatments would increase in cardiac patients who are prepared to quit. While recall of in‐hospital advice was associated with future quit attempts, prepared patients who recalled receiving advice were more likely to make a quit attempt than prepared patients who did not recall receiving advice, as well as unprepared patients. Together, these findings highlight the need to consider a patient's level of preparedness to quit in understanding the success of in‐hospital advice and the importance of making evidence‐based cessation methods available to hospitalized smokers who are prepared to quit. Additionally, identifying patients not prepared to quit may help in providing them with appropriate motivational therapy, to move them along the stages of change, as well as educational information on how to quit once they have decided to do so.

Matching the severity of illness to the appropriate intensity of care is important for the effective delivery of medical care. Overtriage to critical care units results in unnecessary resource consumption. Undertriage to the wards may result in worsening of physiologic parameters1, 2 that often go unnoticed or unaddressed for more than 24 hours.3 Therefore, it is important for emergency department (ED) admission decisions to be accurate with respect to the level of care. Because of the importance of this decision, objective criteria to aid in this decision process, if accurate, would improve medical care delivery.

Physiologic measurements and procedural interventions appear to predict the need for a higher level of care among inpatients.2, 4, 5 This knowledge has led to the development of tools meant to identify inpatients on general wards who are at risk for deterioration. Such tools for identification of inpatients at risk generally use single threshold models triggered by a single abnormal physiologic value, or models that combine multiple parameters into a summative score.6, 7 The performance of previously described risk stratification tools has generally been to exhibit high sensitivity at the sacrifice of low specificity and discriminatory value.8

The value of these models as they apply in the emergency department is less well characterized. Because derangements in physiologic parameters are common among ED patients, one might expect that single‐threshold systems would exhibit high sensitivity at the expense of specificity when applied to this population. In contrast, a summative risk score may be better suited for the complexities of illness in undifferentiated ED patients and offer better discriminatory value in this population. Summative scoring systems have been shown to retain a higher specificity as the score increases compared to single‐threshold systems.8

The Modified Early Warning Score (MEWS)9 is a predictive tool for higher level of care that has been tested in the ED setting. This tool produces a summative score using temperature, respiratory rate, heart rate, level of consciousness, and systolic blood pressure. In a single‐site study from the United Kingdom, MEWS, when calculated at the time of ED presentation, did not improve decision making over a commonly used triage system, exhibiting inadequate sensitivity in identifying patients who would be admitted to the intensive care unit (ICU).10 However, as a result of the care delivered in the ED, patients' conditions can change significantly throughout their stay. Therefore we postulate that the MEWS calculated at a single time in the ED (eg, at the time of admission) is not the most accurate predictor of care intensity requirements.

The primary objective of this research was to add to the literature provided by Subbe et al.10 by describing the performance characteristics and discriminatory ability of the most abnormal MEWS (MEWS Max) score during the entire ED stay in predicting the need for higher levels of care among ED patients presenting to a tertiary care facility in North America.

Patients and Methods

Results

Complete chart abstraction was performed for 299 patient encounters. After abstraction, 19 charts were excluded from final analysis due to missing outcome data (n = 6) or implausible and/or missing crucial data values (n = 13). Pairwise kappa values for abstraction of the MEWS Max score demonstrated agreement ranging from good to very good (0.67‐0.88). Of the 280 analyzed encounters, 76 (27%) met the primary composite outcome of death (n = 1) or need for higher care (n = 76). Of these 76 patients, 69 were admitted from the ED to a high level of care, and 7 were initially admitted to a lower level of care and required transfer to a higher level of care within 24 hours. Thirty‐seven patients requiring a higher level of care were admitted to an ICU (ICU = 31; BMT, CCU, and burn unit with 2 patients each), 9 to intermediate care, and 30 to an acute care bed.

Demographics and presenting characteristics from the study participants can be seen in Table 2. The mean age of participants was 56 years and was similar for the 2 groups. Approximately one‐half of the study participants were female (49%) and there was no statistical association between experiencing the composite outcome and gender (P = 0.28). The majority (64%) of participants were Caucasian, followed by African American (33%) and Hispanic or other (2%). Similar distributions were seen when stratified by outcome. Vital signs of the participants in total and stratified by outcome fell within normal parameters. ED length of stay was similar among those meeting and not meeting the composite outcome (5.5 hours vs. 5.8 hours, P = 0.15). Patients who met the composite outcome were more likely to have arrived by ambulance (63% vs. 43%, P = 0.004).

The distribution of scores and the proportion of participants with each score that met the composite outcome are shown in Figure 1. The MEWS Max was significantly associated with the primary composite outcome (P < 0.001, Cochran‐Armitage trend test). The scoring system demonstrates an increase in the proportion of participants meeting the composite endpoint as the score increases, and all participants with a MEWS Max score 9 met the composite outcome.

Figure 1

ROC are shown in Figure 2. The optimum threshold for MEWS Max based on the sum of sensitivity and specificity is 4, associated with a sensitivity of 62% and a specificity of 79% (Table 3) The predictive ability of the MEWS Max was moderate (C statistic MEWS Max 0.73; 95% CI, 0.66‐0.79), with each 1‐point increase in the MEWS Max score associated with a 60% increase in the odds of meeting the composite endpoint (odds ratio [OR], 1.6; 95% CI, 1.3‐1.8).

Figure 2

Table 4 shows calibration of the model using different subgroups of the patient population. Grouping patients by age or gender did not reveal a higher event rate in any particular group. Using the Hosmer and Lemeshow18 goodness‐of‐fit test to stratify by risk category, no evidence for lack of fit was found (P = 0.06).

In the exploratory analysis, 267 subjects had complete data for all candidate variables. Simple logistic regression revealed that the most predictive MEWS measurement was the MEWS Max (C statistic MEWS Max 0.725, MEWS Initial 0.668, MEWS Admit 0.653). Stepwise selection, forward selection, and backward elimination produced the same model containing method of arrival (P = 0.03), MEWS Max (P < 0.001), IV antibiotics in the ED (P = 0.17), length of stay (P = 0.05), and gender (P = 0.12). In the subset of subjects with these complete data elements (n = 268), the inclusion of the additional measures increased the C statistic to 0.76 (95% CI, 0.69‐0.82), a 0.04 increase over the model that only included MEWS Max in the same subset of subjects.

MEWS Max resulted in no patients being classified as low‐risk, with the majority (81.7%) classified as intermediate‐risk, 15.7% classified as high‐risk, and 2.6% classified as very high risk (Table 5). In all categories the actual event rate fell within the predicted event rate interval. The addition of variables included in MEWS Plus resulted in 14.6% of patients being classified as low‐risk, 64.0% as intermediate risk, 17.2% with high‐risk, and 4.1% as very‐high‐risk. In 58 cases (21.7%), using MEWS Plus would have placed patients in a more appropriate risk category than that assigned by MEWS Max; ie, a lower risk category for those who did not have events, and a higher risk category for those experiencing events. The majority of this correct reclassification was seen in the intermediate risk group by MEWS Max, where 17.6% were appropriately reclassified. Alternatively, 5.6% of cases would have resulted in inappropriate reclassification. Again, the actual event rate fell within the boundaries of predicted risk in all cases.

Discussion

Matching the initial level of care to the patient's severity of illness can be expected to improve the efficiency of health care delivery. The MEWS is a simple prediction instrument that can be calculated at the bedside and would be ideal for this purpose. The MEWS has good predictive ability among patients on the wards or awaiting admission,9, 10 and in this investigation a variation of MEWS appears to have potential to discriminate among high‐risk and low‐risk ED patients.

Examination of the ROC curve for the MEWS Max score demonstrates a fair performance (C statistic = 0.73). In this analysis, we created low‐risk, intermediate‐risk, high‐risk, and very‐high‐risk groups. The strength of the MEWS Max rests in its ability to classify patients as high‐risk or very‐high‐risk. Approximately 16% of patients are classified by MEWS Max as high‐risk, and 3% as very‐high‐risk, making the practitioner more confident in the decision to admit to a high level of care. However, MEWS Max classifies no patients as low risk and approximately 80% of patients are classified as intermediate‐risk. The majority of patients being classified into this gray zone and the inability to classify patients as low‐risk significantly limits the utility of MEWS Max.

In exploratory analysis, these data propose a model using additional readily available parameters that when added to the MEWS Max can improve patient classification. Of particular interest is the ability of the MEWS Plus model to more accurately identify patients at low risk of requiring a higher level of care. When compared to MEWS Max, approximately 22% of patients were correctly reclassified by MEWS Plus, with only 5% incorrectly reclassified. Importantly, MEWS Plus is able to reduce the size of the intermediate‐risk group, predominantly by reclassifying patients as low risk. Forty‐seven (17.6%) of the patients previously categorized as intermediate risk with MEWS Max were reclassified, with 39 of them becoming low risk, 2 (5.1%) of whom had events. However, the major limitation of the MEWS Plus is that it is currently not able to be calculated at the bedside as many of the included variables are time dependent. More analysis is needed to validate precisely which variables are most important, determine how they add to the calculation, and understand when or how often during the ED visit risk should be calculated. Further exploration and validation of this model is necessary.

The results of this investigation add in important ways to a previous study of the MEWS in ED patient triage.10 Subbe et al.10 examined the ability of the MEWS to improve admission decisions beyond those recommended by the Manchester Triage System. Their investigation was conducted among 153 ED patients who belonged to 1 of 3 cohorts being admitted from the ED in the United Kingdom. They concluded that the MEWS was unable to significantly improve admission level of care decisions over the Manchester Triage System. Our investigation differs from that reported by Subbe et al.10 in several important ways. Methodologically, we chose to include a broad population of ED patients rather than selecting 3 cohorts for comparison, and excluded trauma and cardiology patients due to suspected differences in admission patterns in these patients. Further, we conducted our analysis using the maximum MEWS score obtained during a patient's encounter. We felt that using the maximum MEWS score takes full of advantage of all clinical data obtained during the patient's ED visit rather than relying on their severity of illness when the patient first arrives. Additionally, we selected an outcome measure that was determined at 24 hours because we feel events occurring within 24 hours of admission are more likely to reflect a progression of a disease process present at the time of the ED evaluation. Subbe et al.10 analyzed ICU admissions after any duration of hospitalization on the wards. However, ICU admission after several days of ward care may neither be avoidable, nor predictable, while the patient is in the ED.

Subbe et al.10 concluded that the MEWS score did not significantly add to triage decisions aided by the Manchester Triage System. However, in their results, a MEWS score >2 would have classified 7 additional patients as high risk out of 50 who required a transfer to a higher level of care when compared to the Manchester Triage System. Our findings explore the discriminatory value of the maximum MEWS score for a patient throughout the ED visit. This approach, combined with our methodologic differences, have led to more encouraging findings about the utility of the MEWS Max score, especially when combined with a few simple and reliably abstracted variables, to predict the required level of care within 24 hours.

Limitations to our results mainly relate to the study design. We chose a nonconcurrent cohort design using an explicit chart review. Chart reviews have inherent limitations that can include inaccuracy of abstracted data elements, missing data, systematic bias imposed by the abstraction process, and unmeasured confounding. To minimize avoidable biases and maintain accuracy while conducting this chart review, we followed well‐described methods.19 However, because we were relying on retrospective data, some data elements were incomplete. For instance, not all participants had multiple sets of vital signs recorded, which could have affected the predictive accuracy of the risk scores. Anticipating this difficulty, we had algorithms established to handle missing data, which we feel minimized this effect. However, despite this effort, 13 patients had to be excluded due to incomplete data. During review, it was noted that some patients admitted due to a traumatic mechanism were included in the final data analysis despite our intent to exclude them. We expect that this was a very small number, and should have had a minimal effect on risk score calculation. In addition, we modified the original MEWS model in that the GCS was used in substitution for the AVPU score. The conversion of the AVPU score to the GCS is well‐described and is unlikely to have affected the accuracy of the MEWS. We did not adjust for ED length of stay in our primary MEWS model. It is possible that more severely ill patients were in the ED longer and therefore had more opportunity to have abnormal vital signs recorded. ED length of stay was incorporated into the MEWS Plus model. Another limitation relates to our reference standard. We chose a composite of the need for a higher level of care or death within 24 hours. The need for higher care is a subjective endpoint. However, we felt this reflection of actual decision making is more informing than comparisons to other objective, unvalidated scoring systems. As more robust scoring systems are developed, researchers will need to consider developing a reference standard employing blinded adjudicators. Pediatric, cardiology, and trauma patients were excluded from this analysis and therefore our results cannot be extrapolated to these populations. MEWS model calibration was performed using the same data set as that on which the model was tested. This may have resulted in overfit of the model to the data, possibly leading to an overstatement of the model's predictive ability. Additionally, the MEWS Plus model requires validation in another study population. A final limitation is that in performing the study at 1 institution, the results may not be generalizable to other settings.

Building on previous work in defining and testing risk scores to predict poor outcomes, we have shown that the MEWS Max is a potentially useful tool to categorize patients as high‐risk or very‐high‐risk for requiring a higher level of care. MEWS Max suffers from the creation of a large intermediate risk group and the inability to classify patients as low‐risk. Adding further variables to MEWS Max creates a model with improved performance (MEWS Plus). This model may allow for 15% of admissions to be classified as low‐risk and shows promise as a tool to be used in ED triage of patients who are being admitted. Further work should attempt to further refine and validate the MEWS Plus model and examine the effect of implementation of these models on admission decision making and clinical outcomes.

Matching the severity of illness to the appropriate intensity of care is important for the effective delivery of medical care. Overtriage to critical care units results in unnecessary resource consumption. Undertriage to the wards may result in worsening of physiologic parameters1, 2 that often go unnoticed or unaddressed for more than 24 hours.3 Therefore, it is important for emergency department (ED) admission decisions to be accurate with respect to the level of care. Because of the importance of this decision, objective criteria to aid in this decision process, if accurate, would improve medical care delivery.

Physiologic measurements and procedural interventions appear to predict the need for a higher level of care among inpatients.2, 4, 5 This knowledge has led to the development of tools meant to identify inpatients on general wards who are at risk for deterioration. Such tools for identification of inpatients at risk generally use single threshold models triggered by a single abnormal physiologic value, or models that combine multiple parameters into a summative score.6, 7 The performance of previously described risk stratification tools has generally been to exhibit high sensitivity at the sacrifice of low specificity and discriminatory value.8

The value of these models as they apply in the emergency department is less well characterized. Because derangements in physiologic parameters are common among ED patients, one might expect that single‐threshold systems would exhibit high sensitivity at the expense of specificity when applied to this population. In contrast, a summative risk score may be better suited for the complexities of illness in undifferentiated ED patients and offer better discriminatory value in this population. Summative scoring systems have been shown to retain a higher specificity as the score increases compared to single‐threshold systems.8

The Modified Early Warning Score (MEWS)9 is a predictive tool for higher level of care that has been tested in the ED setting. This tool produces a summative score using temperature, respiratory rate, heart rate, level of consciousness, and systolic blood pressure. In a single‐site study from the United Kingdom, MEWS, when calculated at the time of ED presentation, did not improve decision making over a commonly used triage system, exhibiting inadequate sensitivity in identifying patients who would be admitted to the intensive care unit (ICU).10 However, as a result of the care delivered in the ED, patients' conditions can change significantly throughout their stay. Therefore we postulate that the MEWS calculated at a single time in the ED (eg, at the time of admission) is not the most accurate predictor of care intensity requirements.

The primary objective of this research was to add to the literature provided by Subbe et al.10 by describing the performance characteristics and discriminatory ability of the most abnormal MEWS (MEWS Max) score during the entire ED stay in predicting the need for higher levels of care among ED patients presenting to a tertiary care facility in North America.

Patients and Methods

Results

Complete chart abstraction was performed for 299 patient encounters. After abstraction, 19 charts were excluded from final analysis due to missing outcome data (n = 6) or implausible and/or missing crucial data values (n = 13). Pairwise kappa values for abstraction of the MEWS Max score demonstrated agreement ranging from good to very good (0.67‐0.88). Of the 280 analyzed encounters, 76 (27%) met the primary composite outcome of death (n = 1) or need for higher care (n = 76). Of these 76 patients, 69 were admitted from the ED to a high level of care, and 7 were initially admitted to a lower level of care and required transfer to a higher level of care within 24 hours. Thirty‐seven patients requiring a higher level of care were admitted to an ICU (ICU = 31; BMT, CCU, and burn unit with 2 patients each), 9 to intermediate care, and 30 to an acute care bed.

Demographics and presenting characteristics from the study participants can be seen in Table 2. The mean age of participants was 56 years and was similar for the 2 groups. Approximately one‐half of the study participants were female (49%) and there was no statistical association between experiencing the composite outcome and gender (P = 0.28). The majority (64%) of participants were Caucasian, followed by African American (33%) and Hispanic or other (2%). Similar distributions were seen when stratified by outcome. Vital signs of the participants in total and stratified by outcome fell within normal parameters. ED length of stay was similar among those meeting and not meeting the composite outcome (5.5 hours vs. 5.8 hours, P = 0.15). Patients who met the composite outcome were more likely to have arrived by ambulance (63% vs. 43%, P = 0.004).

The distribution of scores and the proportion of participants with each score that met the composite outcome are shown in Figure 1. The MEWS Max was significantly associated with the primary composite outcome (P < 0.001, Cochran‐Armitage trend test). The scoring system demonstrates an increase in the proportion of participants meeting the composite endpoint as the score increases, and all participants with a MEWS Max score 9 met the composite outcome.

Figure 1

ROC are shown in Figure 2. The optimum threshold for MEWS Max based on the sum of sensitivity and specificity is 4, associated with a sensitivity of 62% and a specificity of 79% (Table 3) The predictive ability of the MEWS Max was moderate (C statistic MEWS Max 0.73; 95% CI, 0.66‐0.79), with each 1‐point increase in the MEWS Max score associated with a 60% increase in the odds of meeting the composite endpoint (odds ratio [OR], 1.6; 95% CI, 1.3‐1.8).

Figure 2

Table 4 shows calibration of the model using different subgroups of the patient population. Grouping patients by age or gender did not reveal a higher event rate in any particular group. Using the Hosmer and Lemeshow18 goodness‐of‐fit test to stratify by risk category, no evidence for lack of fit was found (P = 0.06).

In the exploratory analysis, 267 subjects had complete data for all candidate variables. Simple logistic regression revealed that the most predictive MEWS measurement was the MEWS Max (C statistic MEWS Max 0.725, MEWS Initial 0.668, MEWS Admit 0.653). Stepwise selection, forward selection, and backward elimination produced the same model containing method of arrival (P = 0.03), MEWS Max (P < 0.001), IV antibiotics in the ED (P = 0.17), length of stay (P = 0.05), and gender (P = 0.12). In the subset of subjects with these complete data elements (n = 268), the inclusion of the additional measures increased the C statistic to 0.76 (95% CI, 0.69‐0.82), a 0.04 increase over the model that only included MEWS Max in the same subset of subjects.

MEWS Max resulted in no patients being classified as low‐risk, with the majority (81.7%) classified as intermediate‐risk, 15.7% classified as high‐risk, and 2.6% classified as very high risk (Table 5). In all categories the actual event rate fell within the predicted event rate interval. The addition of variables included in MEWS Plus resulted in 14.6% of patients being classified as low‐risk, 64.0% as intermediate risk, 17.2% with high‐risk, and 4.1% as very‐high‐risk. In 58 cases (21.7%), using MEWS Plus would have placed patients in a more appropriate risk category than that assigned by MEWS Max; ie, a lower risk category for those who did not have events, and a higher risk category for those experiencing events. The majority of this correct reclassification was seen in the intermediate risk group by MEWS Max, where 17.6% were appropriately reclassified. Alternatively, 5.6% of cases would have resulted in inappropriate reclassification. Again, the actual event rate fell within the boundaries of predicted risk in all cases.

Discussion

Matching the initial level of care to the patient's severity of illness can be expected to improve the efficiency of health care delivery. The MEWS is a simple prediction instrument that can be calculated at the bedside and would be ideal for this purpose. The MEWS has good predictive ability among patients on the wards or awaiting admission,9, 10 and in this investigation a variation of MEWS appears to have potential to discriminate among high‐risk and low‐risk ED patients.

Examination of the ROC curve for the MEWS Max score demonstrates a fair performance (C statistic = 0.73). In this analysis, we created low‐risk, intermediate‐risk, high‐risk, and very‐high‐risk groups. The strength of the MEWS Max rests in its ability to classify patients as high‐risk or very‐high‐risk. Approximately 16% of patients are classified by MEWS Max as high‐risk, and 3% as very‐high‐risk, making the practitioner more confident in the decision to admit to a high level of care. However, MEWS Max classifies no patients as low risk and approximately 80% of patients are classified as intermediate‐risk. The majority of patients being classified into this gray zone and the inability to classify patients as low‐risk significantly limits the utility of MEWS Max.

In exploratory analysis, these data propose a model using additional readily available parameters that when added to the MEWS Max can improve patient classification. Of particular interest is the ability of the MEWS Plus model to more accurately identify patients at low risk of requiring a higher level of care. When compared to MEWS Max, approximately 22% of patients were correctly reclassified by MEWS Plus, with only 5% incorrectly reclassified. Importantly, MEWS Plus is able to reduce the size of the intermediate‐risk group, predominantly by reclassifying patients as low risk. Forty‐seven (17.6%) of the patients previously categorized as intermediate risk with MEWS Max were reclassified, with 39 of them becoming low risk, 2 (5.1%) of whom had events. However, the major limitation of the MEWS Plus is that it is currently not able to be calculated at the bedside as many of the included variables are time dependent. More analysis is needed to validate precisely which variables are most important, determine how they add to the calculation, and understand when or how often during the ED visit risk should be calculated. Further exploration and validation of this model is necessary.

The results of this investigation add in important ways to a previous study of the MEWS in ED patient triage.10 Subbe et al.10 examined the ability of the MEWS to improve admission decisions beyond those recommended by the Manchester Triage System. Their investigation was conducted among 153 ED patients who belonged to 1 of 3 cohorts being admitted from the ED in the United Kingdom. They concluded that the MEWS was unable to significantly improve admission level of care decisions over the Manchester Triage System. Our investigation differs from that reported by Subbe et al.10 in several important ways. Methodologically, we chose to include a broad population of ED patients rather than selecting 3 cohorts for comparison, and excluded trauma and cardiology patients due to suspected differences in admission patterns in these patients. Further, we conducted our analysis using the maximum MEWS score obtained during a patient's encounter. We felt that using the maximum MEWS score takes full of advantage of all clinical data obtained during the patient's ED visit rather than relying on their severity of illness when the patient first arrives. Additionally, we selected an outcome measure that was determined at 24 hours because we feel events occurring within 24 hours of admission are more likely to reflect a progression of a disease process present at the time of the ED evaluation. Subbe et al.10 analyzed ICU admissions after any duration of hospitalization on the wards. However, ICU admission after several days of ward care may neither be avoidable, nor predictable, while the patient is in the ED.

Subbe et al.10 concluded that the MEWS score did not significantly add to triage decisions aided by the Manchester Triage System. However, in their results, a MEWS score >2 would have classified 7 additional patients as high risk out of 50 who required a transfer to a higher level of care when compared to the Manchester Triage System. Our findings explore the discriminatory value of the maximum MEWS score for a patient throughout the ED visit. This approach, combined with our methodologic differences, have led to more encouraging findings about the utility of the MEWS Max score, especially when combined with a few simple and reliably abstracted variables, to predict the required level of care within 24 hours.

Limitations to our results mainly relate to the study design. We chose a nonconcurrent cohort design using an explicit chart review. Chart reviews have inherent limitations that can include inaccuracy of abstracted data elements, missing data, systematic bias imposed by the abstraction process, and unmeasured confounding. To minimize avoidable biases and maintain accuracy while conducting this chart review, we followed well‐described methods.19 However, because we were relying on retrospective data, some data elements were incomplete. For instance, not all participants had multiple sets of vital signs recorded, which could have affected the predictive accuracy of the risk scores. Anticipating this difficulty, we had algorithms established to handle missing data, which we feel minimized this effect. However, despite this effort, 13 patients had to be excluded due to incomplete data. During review, it was noted that some patients admitted due to a traumatic mechanism were included in the final data analysis despite our intent to exclude them. We expect that this was a very small number, and should have had a minimal effect on risk score calculation. In addition, we modified the original MEWS model in that the GCS was used in substitution for the AVPU score. The conversion of the AVPU score to the GCS is well‐described and is unlikely to have affected the accuracy of the MEWS. We did not adjust for ED length of stay in our primary MEWS model. It is possible that more severely ill patients were in the ED longer and therefore had more opportunity to have abnormal vital signs recorded. ED length of stay was incorporated into the MEWS Plus model. Another limitation relates to our reference standard. We chose a composite of the need for a higher level of care or death within 24 hours. The need for higher care is a subjective endpoint. However, we felt this reflection of actual decision making is more informing than comparisons to other objective, unvalidated scoring systems. As more robust scoring systems are developed, researchers will need to consider developing a reference standard employing blinded adjudicators. Pediatric, cardiology, and trauma patients were excluded from this analysis and therefore our results cannot be extrapolated to these populations. MEWS model calibration was performed using the same data set as that on which the model was tested. This may have resulted in overfit of the model to the data, possibly leading to an overstatement of the model's predictive ability. Additionally, the MEWS Plus model requires validation in another study population. A final limitation is that in performing the study at 1 institution, the results may not be generalizable to other settings.

Building on previous work in defining and testing risk scores to predict poor outcomes, we have shown that the MEWS Max is a potentially useful tool to categorize patients as high‐risk or very‐high‐risk for requiring a higher level of care. MEWS Max suffers from the creation of a large intermediate risk group and the inability to classify patients as low‐risk. Adding further variables to MEWS Max creates a model with improved performance (MEWS Plus). This model may allow for 15% of admissions to be classified as low‐risk and shows promise as a tool to be used in ED triage of patients who are being admitted. Further work should attempt to further refine and validate the MEWS Plus model and examine the effect of implementation of these models on admission decision making and clinical outcomes.

Matching the severity of illness to the appropriate intensity of care is important for the effective delivery of medical care. Overtriage to critical care units results in unnecessary resource consumption. Undertriage to the wards may result in worsening of physiologic parameters1, 2 that often go unnoticed or unaddressed for more than 24 hours.3 Therefore, it is important for emergency department (ED) admission decisions to be accurate with respect to the level of care. Because of the importance of this decision, objective criteria to aid in this decision process, if accurate, would improve medical care delivery.

Physiologic measurements and procedural interventions appear to predict the need for a higher level of care among inpatients.2, 4, 5 This knowledge has led to the development of tools meant to identify inpatients on general wards who are at risk for deterioration. Such tools for identification of inpatients at risk generally use single threshold models triggered by a single abnormal physiologic value, or models that combine multiple parameters into a summative score.6, 7 The performance of previously described risk stratification tools has generally been to exhibit high sensitivity at the sacrifice of low specificity and discriminatory value.8

The value of these models as they apply in the emergency department is less well characterized. Because derangements in physiologic parameters are common among ED patients, one might expect that single‐threshold systems would exhibit high sensitivity at the expense of specificity when applied to this population. In contrast, a summative risk score may be better suited for the complexities of illness in undifferentiated ED patients and offer better discriminatory value in this population. Summative scoring systems have been shown to retain a higher specificity as the score increases compared to single‐threshold systems.8

The Modified Early Warning Score (MEWS)9 is a predictive tool for higher level of care that has been tested in the ED setting. This tool produces a summative score using temperature, respiratory rate, heart rate, level of consciousness, and systolic blood pressure. In a single‐site study from the United Kingdom, MEWS, when calculated at the time of ED presentation, did not improve decision making over a commonly used triage system, exhibiting inadequate sensitivity in identifying patients who would be admitted to the intensive care unit (ICU).10 However, as a result of the care delivered in the ED, patients' conditions can change significantly throughout their stay. Therefore we postulate that the MEWS calculated at a single time in the ED (eg, at the time of admission) is not the most accurate predictor of care intensity requirements.

The primary objective of this research was to add to the literature provided by Subbe et al.10 by describing the performance characteristics and discriminatory ability of the most abnormal MEWS (MEWS Max) score during the entire ED stay in predicting the need for higher levels of care among ED patients presenting to a tertiary care facility in North America.

Patients and Methods

Results

Complete chart abstraction was performed for 299 patient encounters. After abstraction, 19 charts were excluded from final analysis due to missing outcome data (n = 6) or implausible and/or missing crucial data values (n = 13). Pairwise kappa values for abstraction of the MEWS Max score demonstrated agreement ranging from good to very good (0.67‐0.88). Of the 280 analyzed encounters, 76 (27%) met the primary composite outcome of death (n = 1) or need for higher care (n = 76). Of these 76 patients, 69 were admitted from the ED to a high level of care, and 7 were initially admitted to a lower level of care and required transfer to a higher level of care within 24 hours. Thirty‐seven patients requiring a higher level of care were admitted to an ICU (ICU = 31; BMT, CCU, and burn unit with 2 patients each), 9 to intermediate care, and 30 to an acute care bed.

Demographics and presenting characteristics from the study participants can be seen in Table 2. The mean age of participants was 56 years and was similar for the 2 groups. Approximately one‐half of the study participants were female (49%) and there was no statistical association between experiencing the composite outcome and gender (P = 0.28). The majority (64%) of participants were Caucasian, followed by African American (33%) and Hispanic or other (2%). Similar distributions were seen when stratified by outcome. Vital signs of the participants in total and stratified by outcome fell within normal parameters. ED length of stay was similar among those meeting and not meeting the composite outcome (5.5 hours vs. 5.8 hours, P = 0.15). Patients who met the composite outcome were more likely to have arrived by ambulance (63% vs. 43%, P = 0.004).

The distribution of scores and the proportion of participants with each score that met the composite outcome are shown in Figure 1. The MEWS Max was significantly associated with the primary composite outcome (P < 0.001, Cochran‐Armitage trend test). The scoring system demonstrates an increase in the proportion of participants meeting the composite endpoint as the score increases, and all participants with a MEWS Max score 9 met the composite outcome.

Figure 1

ROC are shown in Figure 2. The optimum threshold for MEWS Max based on the sum of sensitivity and specificity is 4, associated with a sensitivity of 62% and a specificity of 79% (Table 3) The predictive ability of the MEWS Max was moderate (C statistic MEWS Max 0.73; 95% CI, 0.66‐0.79), with each 1‐point increase in the MEWS Max score associated with a 60% increase in the odds of meeting the composite endpoint (odds ratio [OR], 1.6; 95% CI, 1.3‐1.8).

Figure 2

Table 4 shows calibration of the model using different subgroups of the patient population. Grouping patients by age or gender did not reveal a higher event rate in any particular group. Using the Hosmer and Lemeshow18 goodness‐of‐fit test to stratify by risk category, no evidence for lack of fit was found (P = 0.06).

In the exploratory analysis, 267 subjects had complete data for all candidate variables. Simple logistic regression revealed that the most predictive MEWS measurement was the MEWS Max (C statistic MEWS Max 0.725, MEWS Initial 0.668, MEWS Admit 0.653). Stepwise selection, forward selection, and backward elimination produced the same model containing method of arrival (P = 0.03), MEWS Max (P < 0.001), IV antibiotics in the ED (P = 0.17), length of stay (P = 0.05), and gender (P = 0.12). In the subset of subjects with these complete data elements (n = 268), the inclusion of the additional measures increased the C statistic to 0.76 (95% CI, 0.69‐0.82), a 0.04 increase over the model that only included MEWS Max in the same subset of subjects.

MEWS Max resulted in no patients being classified as low‐risk, with the majority (81.7%) classified as intermediate‐risk, 15.7% classified as high‐risk, and 2.6% classified as very high risk (Table 5). In all categories the actual event rate fell within the predicted event rate interval. The addition of variables included in MEWS Plus resulted in 14.6% of patients being classified as low‐risk, 64.0% as intermediate risk, 17.2% with high‐risk, and 4.1% as very‐high‐risk. In 58 cases (21.7%), using MEWS Plus would have placed patients in a more appropriate risk category than that assigned by MEWS Max; ie, a lower risk category for those who did not have events, and a higher risk category for those experiencing events. The majority of this correct reclassification was seen in the intermediate risk group by MEWS Max, where 17.6% were appropriately reclassified. Alternatively, 5.6% of cases would have resulted in inappropriate reclassification. Again, the actual event rate fell within the boundaries of predicted risk in all cases.

Discussion

Matching the initial level of care to the patient's severity of illness can be expected to improve the efficiency of health care delivery. The MEWS is a simple prediction instrument that can be calculated at the bedside and would be ideal for this purpose. The MEWS has good predictive ability among patients on the wards or awaiting admission,9, 10 and in this investigation a variation of MEWS appears to have potential to discriminate among high‐risk and low‐risk ED patients.

Examination of the ROC curve for the MEWS Max score demonstrates a fair performance (C statistic = 0.73). In this analysis, we created low‐risk, intermediate‐risk, high‐risk, and very‐high‐risk groups. The strength of the MEWS Max rests in its ability to classify patients as high‐risk or very‐high‐risk. Approximately 16% of patients are classified by MEWS Max as high‐risk, and 3% as very‐high‐risk, making the practitioner more confident in the decision to admit to a high level of care. However, MEWS Max classifies no patients as low risk and approximately 80% of patients are classified as intermediate‐risk. The majority of patients being classified into this gray zone and the inability to classify patients as low‐risk significantly limits the utility of MEWS Max.

In exploratory analysis, these data propose a model using additional readily available parameters that when added to the MEWS Max can improve patient classification. Of particular interest is the ability of the MEWS Plus model to more accurately identify patients at low risk of requiring a higher level of care. When compared to MEWS Max, approximately 22% of patients were correctly reclassified by MEWS Plus, with only 5% incorrectly reclassified. Importantly, MEWS Plus is able to reduce the size of the intermediate‐risk group, predominantly by reclassifying patients as low risk. Forty‐seven (17.6%) of the patients previously categorized as intermediate risk with MEWS Max were reclassified, with 39 of them becoming low risk, 2 (5.1%) of whom had events. However, the major limitation of the MEWS Plus is that it is currently not able to be calculated at the bedside as many of the included variables are time dependent. More analysis is needed to validate precisely which variables are most important, determine how they add to the calculation, and understand when or how often during the ED visit risk should be calculated. Further exploration and validation of this model is necessary.

The results of this investigation add in important ways to a previous study of the MEWS in ED patient triage.10 Subbe et al.10 examined the ability of the MEWS to improve admission decisions beyond those recommended by the Manchester Triage System. Their investigation was conducted among 153 ED patients who belonged to 1 of 3 cohorts being admitted from the ED in the United Kingdom. They concluded that the MEWS was unable to significantly improve admission level of care decisions over the Manchester Triage System. Our investigation differs from that reported by Subbe et al.10 in several important ways. Methodologically, we chose to include a broad population of ED patients rather than selecting 3 cohorts for comparison, and excluded trauma and cardiology patients due to suspected differences in admission patterns in these patients. Further, we conducted our analysis using the maximum MEWS score obtained during a patient's encounter. We felt that using the maximum MEWS score takes full of advantage of all clinical data obtained during the patient's ED visit rather than relying on their severity of illness when the patient first arrives. Additionally, we selected an outcome measure that was determined at 24 hours because we feel events occurring within 24 hours of admission are more likely to reflect a progression of a disease process present at the time of the ED evaluation. Subbe et al.10 analyzed ICU admissions after any duration of hospitalization on the wards. However, ICU admission after several days of ward care may neither be avoidable, nor predictable, while the patient is in the ED.

Subbe et al.10 concluded that the MEWS score did not significantly add to triage decisions aided by the Manchester Triage System. However, in their results, a MEWS score >2 would have classified 7 additional patients as high risk out of 50 who required a transfer to a higher level of care when compared to the Manchester Triage System. Our findings explore the discriminatory value of the maximum MEWS score for a patient throughout the ED visit. This approach, combined with our methodologic differences, have led to more encouraging findings about the utility of the MEWS Max score, especially when combined with a few simple and reliably abstracted variables, to predict the required level of care within 24 hours.

Limitations to our results mainly relate to the study design. We chose a nonconcurrent cohort design using an explicit chart review. Chart reviews have inherent limitations that can include inaccuracy of abstracted data elements, missing data, systematic bias imposed by the abstraction process, and unmeasured confounding. To minimize avoidable biases and maintain accuracy while conducting this chart review, we followed well‐described methods.19 However, because we were relying on retrospective data, some data elements were incomplete. For instance, not all participants had multiple sets of vital signs recorded, which could have affected the predictive accuracy of the risk scores. Anticipating this difficulty, we had algorithms established to handle missing data, which we feel minimized this effect. However, despite this effort, 13 patients had to be excluded due to incomplete data. During review, it was noted that some patients admitted due to a traumatic mechanism were included in the final data analysis despite our intent to exclude them. We expect that this was a very small number, and should have had a minimal effect on risk score calculation. In addition, we modified the original MEWS model in that the GCS was used in substitution for the AVPU score. The conversion of the AVPU score to the GCS is well‐described and is unlikely to have affected the accuracy of the MEWS. We did not adjust for ED length of stay in our primary MEWS model. It is possible that more severely ill patients were in the ED longer and therefore had more opportunity to have abnormal vital signs recorded. ED length of stay was incorporated into the MEWS Plus model. Another limitation relates to our reference standard. We chose a composite of the need for a higher level of care or death within 24 hours. The need for higher care is a subjective endpoint. However, we felt this reflection of actual decision making is more informing than comparisons to other objective, unvalidated scoring systems. As more robust scoring systems are developed, researchers will need to consider developing a reference standard employing blinded adjudicators. Pediatric, cardiology, and trauma patients were excluded from this analysis and therefore our results cannot be extrapolated to these populations. MEWS model calibration was performed using the same data set as that on which the model was tested. This may have resulted in overfit of the model to the data, possibly leading to an overstatement of the model's predictive ability. Additionally, the MEWS Plus model requires validation in another study population. A final limitation is that in performing the study at 1 institution, the results may not be generalizable to other settings.

Building on previous work in defining and testing risk scores to predict poor outcomes, we have shown that the MEWS Max is a potentially useful tool to categorize patients as high‐risk or very‐high‐risk for requiring a higher level of care. MEWS Max suffers from the creation of a large intermediate risk group and the inability to classify patients as low‐risk. Adding further variables to MEWS Max creates a model with improved performance (MEWS Plus). This model may allow for 15% of admissions to be classified as low‐risk and shows promise as a tool to be used in ED triage of patients who are being admitted. Further work should attempt to further refine and validate the MEWS Plus model and examine the effect of implementation of these models on admission decision making and clinical outcomes.

Even though imposition of smoke‐free policies and workplaces comprise one of the most effective antismoking strategies,1 hospital administrators hesitate to implement a smoke‐free medical campus policy.2 They fear losing patients who smoke because these patients will opt for other facilities that permit smoking.

Apart from studies evaluating Joint Commission on Accreditation of Healthcare Organizations (JCAHO)‐required indoor smoking bans in hospitals in 1992,3, 4 there are few published studies or formal evaluations of the impact of medical campuses going smoke‐free. One study of the implementation of a smoke‐free medical campus policy at a university hospital in Little Rock, AR, showed that the policy had no impact on employee retention, bed occupancy, or mean daily census; however, inpatient smoking status was not ascertained.5 Most (83%) employees were supportive of the policy. More importantly, employees at 2 university medical centers reported reduced cigarette consumption and increased attempts to quit after implementation of a smoke‐free medical campus policy.6, 7

Our hospital is 180‐bed, acute care inpatient teaching facility in upstate New York. Prior to the implementation of the smoke‐free medical campus policy, it was common to see employees, visitors, and patients lined up outdoors around the main hospital entrances and smoking just beyond the no smoking signage. Inpatients could look out their windows at the main entrance or into the courtyard and see hospital staff, other patients, and visitors smoking.

This study prospectively evaluates the impact of implementing the smoke‐free medical campus policy and starting an inpatient smoking cessation service. It addresses the following questions that have also been raised by the Task Force for Community Preventive Services.8 Does the institution of hospital smoking bans reduce the percentage of inpatients who smoke or increase the percentage who sign out against medical advice? What are the extended effects (beyond 1 year after implementation) of medical campus smoking bans on employee smoking rates?

Materials and Methods

Results

Inpatient Outcomes

An average of 959 patients were admitted per month in the 18‐month period pre‐ban (January 2005 to June 2006) vs. 988 per month in the 23‐month period post‐ban (July 2006 to September 2008). A monthly average of 89% of inpatients were screened for tobacco use when admitted. The monthly average for the percentage of inpatients who currently smoke has been approximately 21.6% following the implementation of the smoke‐free hospital plan. There has been little variation (Figure 1) in the percentage of inpatients who smoke pre‐ban and post‐ban except for the startup period in 2006 and the onset of the 2007 respiratory illness season.

Figure 1

Among all inpatients who currently smoke, 69.8% received a brief nursing intervention at the time of admission and 25% received an inpatient visit from our part‐time smoking cessation specialist.

The percentage of inpatients who signed out against medical advice (AMA) with the reason of having to smoke was 13.8% (4/29) 6 months pre‐ban, and 13.6% (3/22) 6 months post‐ban. In 2007, there were no inpatients who signed out AMA stating that they needed to smoke. Because the reason for signing out AMA may be underreported, we also examined the rate of smoking among all inpatients who sign out AMA. Six months pre‐ban, this percentage was 48.3% (14/29), but increased 6 months post‐ban to 59% (13/22). In 2007, the percentage of smokers among inpatients who sign out AMA leveled off at 50.8% (29/57).

Review of computerized inpatient prescription orders shows that orders for NRT nearly tripled after the inpatient smoking cessation service started April 1, 2006 (3 months prior to the ban) (Figure 2). Inpatient orders for these medications increased from 832 in a 2‐year period before the ban (April 1, 2004 to March 31, 2006) to 2475 in the 2 years following the initiation of the inpatient smoking service (April 1, 2006 to March 31, 2008). The Chow test is highly significant for a break point in June 2006 (P = 0.008), 1 month prior to the ban.

Figure 2

Discussion

Following implementation of a smoke‐free medical campus, no adverse effects were observed on inpatient volume at our hospital. The percentage of inpatients who smoke and the percentage of inpatients signing out AMA have remained stable after the smoke‐free policy went into effect. In addition, self‐reported employee smoking rates decreased significantly. Fears about losing inpatients (who smoke) following the implementation of a smoke‐free hospital plan were unfounded.

This study employs the electronic medical record to not only monitor trends in the proportion of inpatients who smoke pre‐ban and post‐ban, but also to notify our inpatient smoking cessation specialist, on the day of admission, to consult on patients who currently smoke. Unfortunately, our cessation specialist, who is part‐time, was unable to see all inpatients who smoke on account of the inpatient's acuity, pain, hospice status, weekend or night admission, or not being available due to testing, surgery, or other procedures. Nevertheless, use of NRT increased sharply following the initiation of this program. As shown in Figure 2, a linear rise in NRT orders was already underway starting April 2005, probably in anticipation of the ban and coinciding with the start of the inpatient smoking cessation program. However, the Chow test is highly significant for a breakpoint in June 2006 (P = 0.008), 1 month prior to the ban, meaning that the slope was climbing even more steeply after that point.

As hospitalized smokers may be more motivated to stop smoking, the updated 2008 clinical practice guidelines for Treating Tobacco Use and Dependence now recommend that all patients in the hospital be given medications, advised, counseled, and receive follow‐up after discharge.13 Although our inpatient cessation program was started before these clinical practice guidelines were available, we are currently evaluating the efficacy of our inpatient program by assessing self‐reported quit rates 6‐months posthospitalization (data collection in process). Provision of inpatient smoking cessation has been shown to be an effective smoking cessation intervention if combined with outpatient follow‐up.14 Our current program will be expanded to include outpatient follow‐up, if the inpatient's primary care provider is unable to provide it or if the inpatient refuses faxed referral to the New York State quit line program.

This study evaluates the impact of simultaneously introduced interventions such as medical campus smoking ban, inpatient smoking cessation program, hospital staff education, and other elements of the University of Michigan Smoke‐Free Hospital Implementation Plan. The role of individual components of the plan cannot be evaluated in this study as they were intentionally implemented simultaneously in order to achieve a synergistic effect.

Another limitation of this study is that smoking status is self‐reported and not validated biochemically. Although validated smoking status measures such as salivary cotinine testing would be more scientifically valid, it was not feasible to validate the smoking status of inpatients, nor that of employees. Thus smoking status, as ascertained in this study, is subject to underreporting. Social desirability bias has been recognized as potential limitation of self‐reported smoking status in other evaluations of smoke‐free policies.3, 4, 15

In the 1990s, the employee benefits of instituting indoor smoking bans in hospitals were theorized to include reduced employee sick time, break time, and tobacco use, as well as increased motivation for smoking cessation and reduced legitimacy of tobacco use.16, 17 Peer pressure, workplace socialization, and being forced to stay away from cigarettes for the length of entire workdays have been credited with helping hospital workers to quit.4, 7 In our study, extending the ban to the outdoor areas of our medical campus as well as provision of employee smoking cessation services may augment these mechanisms. This study extends findings of older studies that showed hospital smoking bans (primarily indoor) decreased hospital employee smoking rates. Currently, our reduced employee smoking rate approaches the Healthy People 2010 goal of 12%.18

In conclusion, implementing a smoke‐free medical campus does not adversely affect inpatient volume (even among smokers), does not increase inpatient signing out AMA and can significantly increase inpatient NRT use, which in turn can increase the success of a quit attempt.19 In addition, implementing an outdoor smoking ban further reduces hospital employee smoking rates.

Even though imposition of smoke‐free policies and workplaces comprise one of the most effective antismoking strategies,1 hospital administrators hesitate to implement a smoke‐free medical campus policy.2 They fear losing patients who smoke because these patients will opt for other facilities that permit smoking.

Apart from studies evaluating Joint Commission on Accreditation of Healthcare Organizations (JCAHO)‐required indoor smoking bans in hospitals in 1992,3, 4 there are few published studies or formal evaluations of the impact of medical campuses going smoke‐free. One study of the implementation of a smoke‐free medical campus policy at a university hospital in Little Rock, AR, showed that the policy had no impact on employee retention, bed occupancy, or mean daily census; however, inpatient smoking status was not ascertained.5 Most (83%) employees were supportive of the policy. More importantly, employees at 2 university medical centers reported reduced cigarette consumption and increased attempts to quit after implementation of a smoke‐free medical campus policy.6, 7

Our hospital is 180‐bed, acute care inpatient teaching facility in upstate New York. Prior to the implementation of the smoke‐free medical campus policy, it was common to see employees, visitors, and patients lined up outdoors around the main hospital entrances and smoking just beyond the no smoking signage. Inpatients could look out their windows at the main entrance or into the courtyard and see hospital staff, other patients, and visitors smoking.

This study prospectively evaluates the impact of implementing the smoke‐free medical campus policy and starting an inpatient smoking cessation service. It addresses the following questions that have also been raised by the Task Force for Community Preventive Services.8 Does the institution of hospital smoking bans reduce the percentage of inpatients who smoke or increase the percentage who sign out against medical advice? What are the extended effects (beyond 1 year after implementation) of medical campus smoking bans on employee smoking rates?

Materials and Methods

Results

Inpatient Outcomes

An average of 959 patients were admitted per month in the 18‐month period pre‐ban (January 2005 to June 2006) vs. 988 per month in the 23‐month period post‐ban (July 2006 to September 2008). A monthly average of 89% of inpatients were screened for tobacco use when admitted. The monthly average for the percentage of inpatients who currently smoke has been approximately 21.6% following the implementation of the smoke‐free hospital plan. There has been little variation (Figure 1) in the percentage of inpatients who smoke pre‐ban and post‐ban except for the startup period in 2006 and the onset of the 2007 respiratory illness season.

Figure 1

Among all inpatients who currently smoke, 69.8% received a brief nursing intervention at the time of admission and 25% received an inpatient visit from our part‐time smoking cessation specialist.

The percentage of inpatients who signed out against medical advice (AMA) with the reason of having to smoke was 13.8% (4/29) 6 months pre‐ban, and 13.6% (3/22) 6 months post‐ban. In 2007, there were no inpatients who signed out AMA stating that they needed to smoke. Because the reason for signing out AMA may be underreported, we also examined the rate of smoking among all inpatients who sign out AMA. Six months pre‐ban, this percentage was 48.3% (14/29), but increased 6 months post‐ban to 59% (13/22). In 2007, the percentage of smokers among inpatients who sign out AMA leveled off at 50.8% (29/57).

Review of computerized inpatient prescription orders shows that orders for NRT nearly tripled after the inpatient smoking cessation service started April 1, 2006 (3 months prior to the ban) (Figure 2). Inpatient orders for these medications increased from 832 in a 2‐year period before the ban (April 1, 2004 to March 31, 2006) to 2475 in the 2 years following the initiation of the inpatient smoking service (April 1, 2006 to March 31, 2008). The Chow test is highly significant for a break point in June 2006 (P = 0.008), 1 month prior to the ban.

Figure 2

Discussion

Following implementation of a smoke‐free medical campus, no adverse effects were observed on inpatient volume at our hospital. The percentage of inpatients who smoke and the percentage of inpatients signing out AMA have remained stable after the smoke‐free policy went into effect. In addition, self‐reported employee smoking rates decreased significantly. Fears about losing inpatients (who smoke) following the implementation of a smoke‐free hospital plan were unfounded.

This study employs the electronic medical record to not only monitor trends in the proportion of inpatients who smoke pre‐ban and post‐ban, but also to notify our inpatient smoking cessation specialist, on the day of admission, to consult on patients who currently smoke. Unfortunately, our cessation specialist, who is part‐time, was unable to see all inpatients who smoke on account of the inpatient's acuity, pain, hospice status, weekend or night admission, or not being available due to testing, surgery, or other procedures. Nevertheless, use of NRT increased sharply following the initiation of this program. As shown in Figure 2, a linear rise in NRT orders was already underway starting April 2005, probably in anticipation of the ban and coinciding with the start of the inpatient smoking cessation program. However, the Chow test is highly significant for a breakpoint in June 2006 (P = 0.008), 1 month prior to the ban, meaning that the slope was climbing even more steeply after that point.

As hospitalized smokers may be more motivated to stop smoking, the updated 2008 clinical practice guidelines for Treating Tobacco Use and Dependence now recommend that all patients in the hospital be given medications, advised, counseled, and receive follow‐up after discharge.13 Although our inpatient cessation program was started before these clinical practice guidelines were available, we are currently evaluating the efficacy of our inpatient program by assessing self‐reported quit rates 6‐months posthospitalization (data collection in process). Provision of inpatient smoking cessation has been shown to be an effective smoking cessation intervention if combined with outpatient follow‐up.14 Our current program will be expanded to include outpatient follow‐up, if the inpatient's primary care provider is unable to provide it or if the inpatient refuses faxed referral to the New York State quit line program.

This study evaluates the impact of simultaneously introduced interventions such as medical campus smoking ban, inpatient smoking cessation program, hospital staff education, and other elements of the University of Michigan Smoke‐Free Hospital Implementation Plan. The role of individual components of the plan cannot be evaluated in this study as they were intentionally implemented simultaneously in order to achieve a synergistic effect.

Another limitation of this study is that smoking status is self‐reported and not validated biochemically. Although validated smoking status measures such as salivary cotinine testing would be more scientifically valid, it was not feasible to validate the smoking status of inpatients, nor that of employees. Thus smoking status, as ascertained in this study, is subject to underreporting. Social desirability bias has been recognized as potential limitation of self‐reported smoking status in other evaluations of smoke‐free policies.3, 4, 15

In the 1990s, the employee benefits of instituting indoor smoking bans in hospitals were theorized to include reduced employee sick time, break time, and tobacco use, as well as increased motivation for smoking cessation and reduced legitimacy of tobacco use.16, 17 Peer pressure, workplace socialization, and being forced to stay away from cigarettes for the length of entire workdays have been credited with helping hospital workers to quit.4, 7 In our study, extending the ban to the outdoor areas of our medical campus as well as provision of employee smoking cessation services may augment these mechanisms. This study extends findings of older studies that showed hospital smoking bans (primarily indoor) decreased hospital employee smoking rates. Currently, our reduced employee smoking rate approaches the Healthy People 2010 goal of 12%.18

In conclusion, implementing a smoke‐free medical campus does not adversely affect inpatient volume (even among smokers), does not increase inpatient signing out AMA and can significantly increase inpatient NRT use, which in turn can increase the success of a quit attempt.19 In addition, implementing an outdoor smoking ban further reduces hospital employee smoking rates.

Even though imposition of smoke‐free policies and workplaces comprise one of the most effective antismoking strategies,1 hospital administrators hesitate to implement a smoke‐free medical campus policy.2 They fear losing patients who smoke because these patients will opt for other facilities that permit smoking.

Apart from studies evaluating Joint Commission on Accreditation of Healthcare Organizations (JCAHO)‐required indoor smoking bans in hospitals in 1992,3, 4 there are few published studies or formal evaluations of the impact of medical campuses going smoke‐free. One study of the implementation of a smoke‐free medical campus policy at a university hospital in Little Rock, AR, showed that the policy had no impact on employee retention, bed occupancy, or mean daily census; however, inpatient smoking status was not ascertained.5 Most (83%) employees were supportive of the policy. More importantly, employees at 2 university medical centers reported reduced cigarette consumption and increased attempts to quit after implementation of a smoke‐free medical campus policy.6, 7

Our hospital is 180‐bed, acute care inpatient teaching facility in upstate New York. Prior to the implementation of the smoke‐free medical campus policy, it was common to see employees, visitors, and patients lined up outdoors around the main hospital entrances and smoking just beyond the no smoking signage. Inpatients could look out their windows at the main entrance or into the courtyard and see hospital staff, other patients, and visitors smoking.

This study prospectively evaluates the impact of implementing the smoke‐free medical campus policy and starting an inpatient smoking cessation service. It addresses the following questions that have also been raised by the Task Force for Community Preventive Services.8 Does the institution of hospital smoking bans reduce the percentage of inpatients who smoke or increase the percentage who sign out against medical advice? What are the extended effects (beyond 1 year after implementation) of medical campus smoking bans on employee smoking rates?

Materials and Methods

Results

Inpatient Outcomes

An average of 959 patients were admitted per month in the 18‐month period pre‐ban (January 2005 to June 2006) vs. 988 per month in the 23‐month period post‐ban (July 2006 to September 2008). A monthly average of 89% of inpatients were screened for tobacco use when admitted. The monthly average for the percentage of inpatients who currently smoke has been approximately 21.6% following the implementation of the smoke‐free hospital plan. There has been little variation (Figure 1) in the percentage of inpatients who smoke pre‐ban and post‐ban except for the startup period in 2006 and the onset of the 2007 respiratory illness season.

Figure 1

Among all inpatients who currently smoke, 69.8% received a brief nursing intervention at the time of admission and 25% received an inpatient visit from our part‐time smoking cessation specialist.

The percentage of inpatients who signed out against medical advice (AMA) with the reason of having to smoke was 13.8% (4/29) 6 months pre‐ban, and 13.6% (3/22) 6 months post‐ban. In 2007, there were no inpatients who signed out AMA stating that they needed to smoke. Because the reason for signing out AMA may be underreported, we also examined the rate of smoking among all inpatients who sign out AMA. Six months pre‐ban, this percentage was 48.3% (14/29), but increased 6 months post‐ban to 59% (13/22). In 2007, the percentage of smokers among inpatients who sign out AMA leveled off at 50.8% (29/57).

Review of computerized inpatient prescription orders shows that orders for NRT nearly tripled after the inpatient smoking cessation service started April 1, 2006 (3 months prior to the ban) (Figure 2). Inpatient orders for these medications increased from 832 in a 2‐year period before the ban (April 1, 2004 to March 31, 2006) to 2475 in the 2 years following the initiation of the inpatient smoking service (April 1, 2006 to March 31, 2008). The Chow test is highly significant for a break point in June 2006 (P = 0.008), 1 month prior to the ban.

Figure 2

Discussion

Following implementation of a smoke‐free medical campus, no adverse effects were observed on inpatient volume at our hospital. The percentage of inpatients who smoke and the percentage of inpatients signing out AMA have remained stable after the smoke‐free policy went into effect. In addition, self‐reported employee smoking rates decreased significantly. Fears about losing inpatients (who smoke) following the implementation of a smoke‐free hospital plan were unfounded.

This study employs the electronic medical record to not only monitor trends in the proportion of inpatients who smoke pre‐ban and post‐ban, but also to notify our inpatient smoking cessation specialist, on the day of admission, to consult on patients who currently smoke. Unfortunately, our cessation specialist, who is part‐time, was unable to see all inpatients who smoke on account of the inpatient's acuity, pain, hospice status, weekend or night admission, or not being available due to testing, surgery, or other procedures. Nevertheless, use of NRT increased sharply following the initiation of this program. As shown in Figure 2, a linear rise in NRT orders was already underway starting April 2005, probably in anticipation of the ban and coinciding with the start of the inpatient smoking cessation program. However, the Chow test is highly significant for a breakpoint in June 2006 (P = 0.008), 1 month prior to the ban, meaning that the slope was climbing even more steeply after that point.

As hospitalized smokers may be more motivated to stop smoking, the updated 2008 clinical practice guidelines for Treating Tobacco Use and Dependence now recommend that all patients in the hospital be given medications, advised, counseled, and receive follow‐up after discharge.13 Although our inpatient cessation program was started before these clinical practice guidelines were available, we are currently evaluating the efficacy of our inpatient program by assessing self‐reported quit rates 6‐months posthospitalization (data collection in process). Provision of inpatient smoking cessation has been shown to be an effective smoking cessation intervention if combined with outpatient follow‐up.14 Our current program will be expanded to include outpatient follow‐up, if the inpatient's primary care provider is unable to provide it or if the inpatient refuses faxed referral to the New York State quit line program.

This study evaluates the impact of simultaneously introduced interventions such as medical campus smoking ban, inpatient smoking cessation program, hospital staff education, and other elements of the University of Michigan Smoke‐Free Hospital Implementation Plan. The role of individual components of the plan cannot be evaluated in this study as they were intentionally implemented simultaneously in order to achieve a synergistic effect.

Another limitation of this study is that smoking status is self‐reported and not validated biochemically. Although validated smoking status measures such as salivary cotinine testing would be more scientifically valid, it was not feasible to validate the smoking status of inpatients, nor that of employees. Thus smoking status, as ascertained in this study, is subject to underreporting. Social desirability bias has been recognized as potential limitation of self‐reported smoking status in other evaluations of smoke‐free policies.3, 4, 15

In the 1990s, the employee benefits of instituting indoor smoking bans in hospitals were theorized to include reduced employee sick time, break time, and tobacco use, as well as increased motivation for smoking cessation and reduced legitimacy of tobacco use.16, 17 Peer pressure, workplace socialization, and being forced to stay away from cigarettes for the length of entire workdays have been credited with helping hospital workers to quit.4, 7 In our study, extending the ban to the outdoor areas of our medical campus as well as provision of employee smoking cessation services may augment these mechanisms. This study extends findings of older studies that showed hospital smoking bans (primarily indoor) decreased hospital employee smoking rates. Currently, our reduced employee smoking rate approaches the Healthy People 2010 goal of 12%.18

In conclusion, implementing a smoke‐free medical campus does not adversely affect inpatient volume (even among smokers), does not increase inpatient signing out AMA and can significantly increase inpatient NRT use, which in turn can increase the success of a quit attempt.19 In addition, implementing an outdoor smoking ban further reduces hospital employee smoking rates.

In an ideal situation, a child could be cared for by 1 physician from childhood through adolescence. This physician could care for the child from the first days in the nursery, through multiple well‐child and sick visits, and during any hospitalizations. In 1992, the American Academy of Pediatrics (AAP) introduced the concept of the medical home to provide accessible, continuous, comprehensive, family‐centered, coordinated, and compassionate care for children.1 These ideals were reaffirmed in a 2002 AAP policy statement.2 The demands for outpatient care have become more intense and patient safety issues have become a public focus. Combined with the necessity for increased efficiency, the ideal medical home has become difficult to achieve simultaneously in both an outpatient and inpatient setting. This has led to the growth of the hospital specialist or hospitalist, a term first coined by Drs. Wachter and Goldman in August 1996.3 A hospitalist has been defined as a physician whose primary professional focus is the general medical care of hospitalized patients . . . teaching, research, and leadership related to hospital care.4 Despite early concerns voiced after the publication of this landmark article, there has been tremendous growth in the number of hospitalists nationwide.5 Initially, this growth was seen in adult medicine, but pediatric hospitalists have become increasingly more common. There are an estimated 10,000 to 20,000 hospitalists in the profession, with more than 30,000 expected by the end of the decade, with just over 10% being pediatric hospitalists.4 Several studies have shown the benefits of pediatric hospitalist programs with decreased length of stay, hospital charges, and utilization of unnecessary tests and therapies, and increased satisfaction among the physicians, students, and patients.611

History of the Diagnostic Referral Service

Despite the coining of the term hospitalist in 1996, specialization in inpatient care existed in various forms long before then.12 The Diagnostic Referral Service (DRS) at Children's Hospital of Pittsburgh (CHP) initially began under the guidance of Dr. Edmund R. McCluskey, chairman of the Department of Pediatrics at CHP, and Dr. Paul C. Gaffney, a beloved and revered clinician and educator. In 1951, Dr. Gaffney joined the full‐time CHP faculty after residency, chief residency, and a year of fellowship, providing his expertise in hematology and oncology. Over time, Dr. Gaffney's role expanded to that of a master physician, and pediatricians and family practitioners in the community began sending their most diagnostically challenging patients to be seen in his clinic. His activities further extended to providing inpatient care and consultations for complex patients.

With these growing responsibilities, Dr. Gaffney formed the DRS as a separate division within the Department of Pediatrics in the mid‐1970s and began developing the division. The group, then and now, is comprised of general pediatricians who provide multidisciplinary care for hospitalized children as well as for ambulatory consultations. The DRS was initially described in the literature 20 years ago; the roles of the 5 full‐time physicians at that time included a variety of clinical, teaching, and scholarly activities, as both inpatient and outpatient consultative physicians.13 Though much growth has occurred within the division since then, Dr. Gaffney's initial goals of providing excellent patient care and education in an academic setting still remain at the heart of each group member.

Growth of the Division

In March 2002, there were 4 full‐time physicians within the DRS. A remarkable increase in the group size has occurred since then, and currently there are 16 physicians. Each member of the division is assigned a specific activity, either outpatient or inpatient, for at least a 5‐day block. This allows the division to provide continuity in the care of complex patients in both settings and help maintain a medical home for these challenging patients. The primary care physician (PCP) remains responsible for primary care of the patient while the DRS can help manage the patient's complexities in the outpatient and inpatient setting. In essence, there is joint patient ownership between the DRS and the PCP, with relegation of different skill sets to provide a complete medical home for the patient.

The current activities of the group are summarized as follows.

Inpatient Activities

Inpatient Care

Corresponding to the growing pediatric hospitalist movement in the past decade, several area PCPs began requesting that their office patients be followed by the DRS when admitted to CHP. In 2003, the DRS physician referral list had 150 physicians, and it currently has over 325 physicians who refer the inpatient care of their patients to the DRS. These practices are located in a variety of locations in Western Pennsylvania, Ohio, and West Virginia. There are only 7 private practices that continue to maintain their admitting privileges to CHP, and these practices account for 0.5% of general pediatric admissions. The remaining admissions (those PCPs not on the DRS referral list or with admitting privileges) are covered by a rotating attending physician; 85% of the time this is a DRS physician. Therefore, while 0.5% of general pediatric admissions are cared for by private pediatricians, >95% of all general pediatric admissions are cared for by DRS, with the remaining patients cared for by a small number of pediatric subspecialists who occasionally serve as rotating attendings.

Associated with the increase in referrals has been a marked increase in inpatient activity (Figures 0, 0, 0, 13). Nearly 1 of every 4 CHP discharges and 1 of every 3 observation patients are cared for by the DRS.

Figure 1

Figure 2

Figure 3

The DRS division has seen an increase in complex inpatient admissions as well as a much larger number of routine pediatric admissions from the community PCP referrals, as described above. Statistically, this has resulted in an overall stable to slightly increased inpatient complexity for the DRS group. During this same time period, there was a steep decrease in DRS inpatient length of stay followed by maintenance at the shorter length of stay thereafter. Inpatient complexity has increased throughout CHP, yet the same decreases in length of stay have not been seen universally in all the divisions (Figure 4). The advantage that a hospitalist group can bring in decreasing length of stay (and, thereby, hospital costs) has been seen in hospitalist programs around the country.6, 7

Figure 4

Each member of the group attends on the general pediatric ward for 9 to 10 months per year as compared to 2 months per year in 1986. In order to prevent job‐related fatigue, 6 to 8 members of the group typically attend on the ward at the same time so that individual patient volume is more manageable. On average, each individual physician is responsible for about 6 to 8 patients per day. This census allows for daily education of residents and medical students as well as faculty participation in a variety of administrative activities. Despite emphasis on careful documentation and billing for both inpatient and ambulatory activities, the division, like most other pediatric hospitalist divisions, depends upon financial support from CHP and the Department of Pediatrics.

There is a wide variety of diagnoses that are seen in the inpatient setting, with notable similarities and differences when compared to 2 decades ago (Table 1). For example, asthma, gastroenteritis, and bronchiolitis continue to be frequent diagnoses, with bronchiolitis admissions becoming more frequent, following national trends.14

Table 1.Most Common Inpatient Diagnoses

FY 1986		FY 2007
Rank	Diagnosis	Rank	Diagnosis
Abbreviations: FY, fiscal year; NOS, not otherwise specified.
1.	Asthma	1.	Bronchiolitis
2.	Gastroenteritis	2.	Asthma
3.	Failure to thrive	3.	Pneumonia
4.	Seizures	4.	Dehydration
5.	Bronchiolitis	5.	Viral enteritis
6.	Pneumonia	6.	Viral infection, NOS
7.	Suspected sepsis	7.	Esophageal reflux
8.	Apneic episodes	8.	Fever
9.	Meningitis	9.	Cellulitis
10.	Otitis media	10.	Convulsions

In general, a DRS faculty member becomes the primary resource for families with medically complex children. The same faculty member tends to follow the patient in the consultative ambulatory clinic, be available for phone calls, and, if possible, follow them as the attending physician when the patient is admitted. This provides a degree of continuity generally not seen in many other hospitalist programs and has the potential to increase patient and physician satisfaction as well as patient safety.

Outpatient Activities

Education

Teaching is a major role for the DRS, and the division is closely involved in leadership in medical education. Two members are directors of the pediatric physical exam course for first‐year and second‐year medical students, 2 members are the third‐year medical school Pediatric clerkship directors, 1 member is co‐director for the fourth‐year acting internship, 1 member is co‐director of the advanced pediatric interviewing program, 1 member is director of the pediatric medical education program, and 1 member is associate residency program director.

The entire group is involved with teaching at all of these levels. The majority of the group is involved with formal mentoring and advising of residents and medical students. The only general pediatric aspect of student and resident medical education in which the DRS is no longer involved is ambulatory pediatric medicine. The full‐time ambulatory faculty is responsible for the primary ambulatory care experience. However, many residents choose to complete an elective in DRS, including the outpatient clinic, to become exposed to the different diagnostic dilemmas and coordination of care visits that they may not see in their primary care continuity clinics.

The division always welcomes new teaching challenges and incorporates new methods of teaching as opportunities arise. For example, the recent family‐centered rounds initiative allowed for new teaching methods that were not previously possible. The team, comprised of a senior resident, 2 interns, 2 students, and an attending physician rounds at the bedside with permission from the parent. The patient's nurse, when available, and a pharmacist are often a part of the team. The case is presented by the student or intern (directed to the parent), and the case is discussed and clarified for the family. A plan for the day is presented and discussed with the family for approval. Through family‐centered rounds, the DRS attending provides patient specific teaching and role modeling during rounds that would not otherwise have been possible with classical didactic teaching. This method of daily rounding also allows for the patients, families, nurses, nursing students, medical students, and residents to be taught by the attending physician simultaneously. Additionally, it affords the nurses the opportunity to participate in medical decision‐making, and the house staff have perceived fewer pages by the nurses to clarify clinical issues.

The EH program also provided new teaching opportunities. Through the EH, the house staff and students are exposed to direct attending teaching in the evenings that otherwise would not occur, such as direct observation of student and resident histories and physical examinations. Based on resident evaluations and comments to the residency program directors, this teaching experience is deemed to be valuable and effective. In fact, since the EH program's inception 3 years ago, 2 EHs have been selected by the residents as Teacher of the Year.

Patient Safety

Patient safety and reduction of medical errors is a major focus of the entire group. One DRS member serves within the hospital administration as Medical Director for Clinical Excellence and Service to enhance patient safety hospital‐wide. One DRS member orchestrates a monthly house staff meeting entitled To Err is Human which provides a nonthreatening environment for residents to discuss medical errors or difficult situations that they have encountered. Two DRS members are part of the Physician Advisory Committee, which serves as a bridge between the information technology group and clinicians. This committee has aided in achieving a smooth transition to a completely electronic medical record (EMR) and works together to use the tools of an EMR to enhance patient safety. This successful EMR implementation was recognized by the Health Information Management Systems Society in October 2008. Additionally, stemming from several successful patient safety initiatives, CHP was 1 of only 7 children's hospitals recognized for patient safety in 2008 by Leapfrog, the nation's premier patient safety evaluation group.

Scholarly Activity

Previously, the DRS was responsible for the general medical care of liver transplant recipients, and many prior publications from the division focused on these patients.13 The division no longer provides that service, and the current focus of scholarly activity is publication of case reports, book chapters, and review articles. There were 13 publications from the division members in the past 3 years. The group also serves as a major resource within the department by referring patients that fulfill the clinical criteria for ongoing clinical studies in other divisions.

One member of the group continues to serve as senior editor of the Atlas of Pediatric Physical Diagnosis, which is currently in its fifth edition. Several members of the group contribute chapters to this well‐known text. One member is an editor and another is a specialty reviewer for FirstConsult.com, a website for physicians. Another member has served as an associate editor for the Journal of Pediatrics Grand Round Section.

At the University of Pittsburgh, both tenured and nontenured faculty promotions carry the same title without a prefix. Academic promotions for clinician‐educators center around clinical excellence and innovation in education. The 3 senior members of the group have been promoted to professor (1 tenured, 2 nontenured), and 2 other members are currently in consideration for promotion to associate professor. Two members of the group have been elected to the School of Medicine Academy of Master Educators, which recognizes and rewards excellence in education.

Future Goals

Future goals include expansion and refinement of the division's current inpatient and ambulatory activities. The group increased in size to 16 physicians in July 2008 due to the increased inpatient volume and growing demand for outpatient referrals. The family‐centered rounds initiative will continue to be refined to provide the best possible service to the patients and their families. The members of the group will have increased activity and involvement at the regional and national level with the growing pediatric hospitalist movement.

A pediatric hospitalist fellowship program certainly would be feasible in the current environment. At this writing, there are only 8 pediatric hospitalist fellowship programs nationwide. The outpatient/emnpatient environment that the DRS provides the community would certainly provide a unique training environment for a hospitalist fellowship. The diversity in hospitalist divisions nationwide and the standardization of fellowship training is an important task for the future.23

Discussion

Pediatrics has undergone major changes since the original description of the DRS 20 years ago.7 These changes have revolutionized the practice of pediatrics in both the ambulatory and inpatient settings. The DRS role has changed significantly along with the national trends in pediatric hospitalist growth over the past decade. Currently, 90% of the division's clinical activity is inpatient care. In essence, this is an extension of the original consultant role, but the model has been extended to provide inpatient multidisciplinary care for pediatric patients.

Despite the remarkable growth in inpatient activity, 1 unique advantage to the DRS model is maintenance of an active outpatient consultation clinic focused on providing a multidisciplinary medical home for chronically ill patients. DRS faculty are able to coordinate the care of these complex patients while not usurping the primary care responsibility of the community physician. The same faculty are able to extend the continuity of care to the inpatient setting should the patient require admission.

There have been several innovations that the DRS has implemented over the past decade. The LSU was designed to provide effective and efficient care. The EH program extends attending in‐house coverage without the disadvantages of 24‐hour, 7‐day‐per‐week coverage. Expanding services to include The Children's Home allows for easier transition to home for technology‐dependent patients and families. At the same time, DRS continues to strive for innovative clinical leadership as well as creative and effective student and resident education.

Conclusion

Despite the remarkable growth and increased clinical activity of the DRS since its inception, Dr. Gaffney's ideals continue to serve as the lifeline for the division. The DRS still maintains the consultative pediatric role that he originated, but the inpatient activity has grown with the pediatric hospitalist movement at the same time. The division also maintains an active outpatient clinic. This dual function allows the DRS to continue to serve the community in a unique manner.

In an ideal situation, a child could be cared for by 1 physician from childhood through adolescence. This physician could care for the child from the first days in the nursery, through multiple well‐child and sick visits, and during any hospitalizations. In 1992, the American Academy of Pediatrics (AAP) introduced the concept of the medical home to provide accessible, continuous, comprehensive, family‐centered, coordinated, and compassionate care for children.1 These ideals were reaffirmed in a 2002 AAP policy statement.2 The demands for outpatient care have become more intense and patient safety issues have become a public focus. Combined with the necessity for increased efficiency, the ideal medical home has become difficult to achieve simultaneously in both an outpatient and inpatient setting. This has led to the growth of the hospital specialist or hospitalist, a term first coined by Drs. Wachter and Goldman in August 1996.3 A hospitalist has been defined as a physician whose primary professional focus is the general medical care of hospitalized patients . . . teaching, research, and leadership related to hospital care.4 Despite early concerns voiced after the publication of this landmark article, there has been tremendous growth in the number of hospitalists nationwide.5 Initially, this growth was seen in adult medicine, but pediatric hospitalists have become increasingly more common. There are an estimated 10,000 to 20,000 hospitalists in the profession, with more than 30,000 expected by the end of the decade, with just over 10% being pediatric hospitalists.4 Several studies have shown the benefits of pediatric hospitalist programs with decreased length of stay, hospital charges, and utilization of unnecessary tests and therapies, and increased satisfaction among the physicians, students, and patients.611

History of the Diagnostic Referral Service

Despite the coining of the term hospitalist in 1996, specialization in inpatient care existed in various forms long before then.12 The Diagnostic Referral Service (DRS) at Children's Hospital of Pittsburgh (CHP) initially began under the guidance of Dr. Edmund R. McCluskey, chairman of the Department of Pediatrics at CHP, and Dr. Paul C. Gaffney, a beloved and revered clinician and educator. In 1951, Dr. Gaffney joined the full‐time CHP faculty after residency, chief residency, and a year of fellowship, providing his expertise in hematology and oncology. Over time, Dr. Gaffney's role expanded to that of a master physician, and pediatricians and family practitioners in the community began sending their most diagnostically challenging patients to be seen in his clinic. His activities further extended to providing inpatient care and consultations for complex patients.

With these growing responsibilities, Dr. Gaffney formed the DRS as a separate division within the Department of Pediatrics in the mid‐1970s and began developing the division. The group, then and now, is comprised of general pediatricians who provide multidisciplinary care for hospitalized children as well as for ambulatory consultations. The DRS was initially described in the literature 20 years ago; the roles of the 5 full‐time physicians at that time included a variety of clinical, teaching, and scholarly activities, as both inpatient and outpatient consultative physicians.13 Though much growth has occurred within the division since then, Dr. Gaffney's initial goals of providing excellent patient care and education in an academic setting still remain at the heart of each group member.

Growth of the Division

In March 2002, there were 4 full‐time physicians within the DRS. A remarkable increase in the group size has occurred since then, and currently there are 16 physicians. Each member of the division is assigned a specific activity, either outpatient or inpatient, for at least a 5‐day block. This allows the division to provide continuity in the care of complex patients in both settings and help maintain a medical home for these challenging patients. The primary care physician (PCP) remains responsible for primary care of the patient while the DRS can help manage the patient's complexities in the outpatient and inpatient setting. In essence, there is joint patient ownership between the DRS and the PCP, with relegation of different skill sets to provide a complete medical home for the patient.

The current activities of the group are summarized as follows.

Inpatient Activities

Inpatient Care

Corresponding to the growing pediatric hospitalist movement in the past decade, several area PCPs began requesting that their office patients be followed by the DRS when admitted to CHP. In 2003, the DRS physician referral list had 150 physicians, and it currently has over 325 physicians who refer the inpatient care of their patients to the DRS. These practices are located in a variety of locations in Western Pennsylvania, Ohio, and West Virginia. There are only 7 private practices that continue to maintain their admitting privileges to CHP, and these practices account for 0.5% of general pediatric admissions. The remaining admissions (those PCPs not on the DRS referral list or with admitting privileges) are covered by a rotating attending physician; 85% of the time this is a DRS physician. Therefore, while 0.5% of general pediatric admissions are cared for by private pediatricians, >95% of all general pediatric admissions are cared for by DRS, with the remaining patients cared for by a small number of pediatric subspecialists who occasionally serve as rotating attendings.

Associated with the increase in referrals has been a marked increase in inpatient activity (Figures 0, 0, 0, 13). Nearly 1 of every 4 CHP discharges and 1 of every 3 observation patients are cared for by the DRS.

Figure 1

Figure 2

Figure 3

The DRS division has seen an increase in complex inpatient admissions as well as a much larger number of routine pediatric admissions from the community PCP referrals, as described above. Statistically, this has resulted in an overall stable to slightly increased inpatient complexity for the DRS group. During this same time period, there was a steep decrease in DRS inpatient length of stay followed by maintenance at the shorter length of stay thereafter. Inpatient complexity has increased throughout CHP, yet the same decreases in length of stay have not been seen universally in all the divisions (Figure 4). The advantage that a hospitalist group can bring in decreasing length of stay (and, thereby, hospital costs) has been seen in hospitalist programs around the country.6, 7

Figure 4

Each member of the group attends on the general pediatric ward for 9 to 10 months per year as compared to 2 months per year in 1986. In order to prevent job‐related fatigue, 6 to 8 members of the group typically attend on the ward at the same time so that individual patient volume is more manageable. On average, each individual physician is responsible for about 6 to 8 patients per day. This census allows for daily education of residents and medical students as well as faculty participation in a variety of administrative activities. Despite emphasis on careful documentation and billing for both inpatient and ambulatory activities, the division, like most other pediatric hospitalist divisions, depends upon financial support from CHP and the Department of Pediatrics.

There is a wide variety of diagnoses that are seen in the inpatient setting, with notable similarities and differences when compared to 2 decades ago (Table 1). For example, asthma, gastroenteritis, and bronchiolitis continue to be frequent diagnoses, with bronchiolitis admissions becoming more frequent, following national trends.14

Table 1.Most Common Inpatient Diagnoses

FY 1986		FY 2007
Rank	Diagnosis	Rank	Diagnosis
Abbreviations: FY, fiscal year; NOS, not otherwise specified.
1.	Asthma	1.	Bronchiolitis
2.	Gastroenteritis	2.	Asthma
3.	Failure to thrive	3.	Pneumonia
4.	Seizures	4.	Dehydration
5.	Bronchiolitis	5.	Viral enteritis
6.	Pneumonia	6.	Viral infection, NOS
7.	Suspected sepsis	7.	Esophageal reflux
8.	Apneic episodes	8.	Fever
9.	Meningitis	9.	Cellulitis
10.	Otitis media	10.	Convulsions

In general, a DRS faculty member becomes the primary resource for families with medically complex children. The same faculty member tends to follow the patient in the consultative ambulatory clinic, be available for phone calls, and, if possible, follow them as the attending physician when the patient is admitted. This provides a degree of continuity generally not seen in many other hospitalist programs and has the potential to increase patient and physician satisfaction as well as patient safety.

Outpatient Activities

Education

Teaching is a major role for the DRS, and the division is closely involved in leadership in medical education. Two members are directors of the pediatric physical exam course for first‐year and second‐year medical students, 2 members are the third‐year medical school Pediatric clerkship directors, 1 member is co‐director for the fourth‐year acting internship, 1 member is co‐director of the advanced pediatric interviewing program, 1 member is director of the pediatric medical education program, and 1 member is associate residency program director.

The entire group is involved with teaching at all of these levels. The majority of the group is involved with formal mentoring and advising of residents and medical students. The only general pediatric aspect of student and resident medical education in which the DRS is no longer involved is ambulatory pediatric medicine. The full‐time ambulatory faculty is responsible for the primary ambulatory care experience. However, many residents choose to complete an elective in DRS, including the outpatient clinic, to become exposed to the different diagnostic dilemmas and coordination of care visits that they may not see in their primary care continuity clinics.

The division always welcomes new teaching challenges and incorporates new methods of teaching as opportunities arise. For example, the recent family‐centered rounds initiative allowed for new teaching methods that were not previously possible. The team, comprised of a senior resident, 2 interns, 2 students, and an attending physician rounds at the bedside with permission from the parent. The patient's nurse, when available, and a pharmacist are often a part of the team. The case is presented by the student or intern (directed to the parent), and the case is discussed and clarified for the family. A plan for the day is presented and discussed with the family for approval. Through family‐centered rounds, the DRS attending provides patient specific teaching and role modeling during rounds that would not otherwise have been possible with classical didactic teaching. This method of daily rounding also allows for the patients, families, nurses, nursing students, medical students, and residents to be taught by the attending physician simultaneously. Additionally, it affords the nurses the opportunity to participate in medical decision‐making, and the house staff have perceived fewer pages by the nurses to clarify clinical issues.

The EH program also provided new teaching opportunities. Through the EH, the house staff and students are exposed to direct attending teaching in the evenings that otherwise would not occur, such as direct observation of student and resident histories and physical examinations. Based on resident evaluations and comments to the residency program directors, this teaching experience is deemed to be valuable and effective. In fact, since the EH program's inception 3 years ago, 2 EHs have been selected by the residents as Teacher of the Year.

Patient Safety

Patient safety and reduction of medical errors is a major focus of the entire group. One DRS member serves within the hospital administration as Medical Director for Clinical Excellence and Service to enhance patient safety hospital‐wide. One DRS member orchestrates a monthly house staff meeting entitled To Err is Human which provides a nonthreatening environment for residents to discuss medical errors or difficult situations that they have encountered. Two DRS members are part of the Physician Advisory Committee, which serves as a bridge between the information technology group and clinicians. This committee has aided in achieving a smooth transition to a completely electronic medical record (EMR) and works together to use the tools of an EMR to enhance patient safety. This successful EMR implementation was recognized by the Health Information Management Systems Society in October 2008. Additionally, stemming from several successful patient safety initiatives, CHP was 1 of only 7 children's hospitals recognized for patient safety in 2008 by Leapfrog, the nation's premier patient safety evaluation group.

Scholarly Activity

Previously, the DRS was responsible for the general medical care of liver transplant recipients, and many prior publications from the division focused on these patients.13 The division no longer provides that service, and the current focus of scholarly activity is publication of case reports, book chapters, and review articles. There were 13 publications from the division members in the past 3 years. The group also serves as a major resource within the department by referring patients that fulfill the clinical criteria for ongoing clinical studies in other divisions.

One member of the group continues to serve as senior editor of the Atlas of Pediatric Physical Diagnosis, which is currently in its fifth edition. Several members of the group contribute chapters to this well‐known text. One member is an editor and another is a specialty reviewer for FirstConsult.com, a website for physicians. Another member has served as an associate editor for the Journal of Pediatrics Grand Round Section.

At the University of Pittsburgh, both tenured and nontenured faculty promotions carry the same title without a prefix. Academic promotions for clinician‐educators center around clinical excellence and innovation in education. The 3 senior members of the group have been promoted to professor (1 tenured, 2 nontenured), and 2 other members are currently in consideration for promotion to associate professor. Two members of the group have been elected to the School of Medicine Academy of Master Educators, which recognizes and rewards excellence in education.

Future Goals

Future goals include expansion and refinement of the division's current inpatient and ambulatory activities. The group increased in size to 16 physicians in July 2008 due to the increased inpatient volume and growing demand for outpatient referrals. The family‐centered rounds initiative will continue to be refined to provide the best possible service to the patients and their families. The members of the group will have increased activity and involvement at the regional and national level with the growing pediatric hospitalist movement.

A pediatric hospitalist fellowship program certainly would be feasible in the current environment. At this writing, there are only 8 pediatric hospitalist fellowship programs nationwide. The outpatient/emnpatient environment that the DRS provides the community would certainly provide a unique training environment for a hospitalist fellowship. The diversity in hospitalist divisions nationwide and the standardization of fellowship training is an important task for the future.23

Discussion

Pediatrics has undergone major changes since the original description of the DRS 20 years ago.7 These changes have revolutionized the practice of pediatrics in both the ambulatory and inpatient settings. The DRS role has changed significantly along with the national trends in pediatric hospitalist growth over the past decade. Currently, 90% of the division's clinical activity is inpatient care. In essence, this is an extension of the original consultant role, but the model has been extended to provide inpatient multidisciplinary care for pediatric patients.

Despite the remarkable growth in inpatient activity, 1 unique advantage to the DRS model is maintenance of an active outpatient consultation clinic focused on providing a multidisciplinary medical home for chronically ill patients. DRS faculty are able to coordinate the care of these complex patients while not usurping the primary care responsibility of the community physician. The same faculty are able to extend the continuity of care to the inpatient setting should the patient require admission.

There have been several innovations that the DRS has implemented over the past decade. The LSU was designed to provide effective and efficient care. The EH program extends attending in‐house coverage without the disadvantages of 24‐hour, 7‐day‐per‐week coverage. Expanding services to include The Children's Home allows for easier transition to home for technology‐dependent patients and families. At the same time, DRS continues to strive for innovative clinical leadership as well as creative and effective student and resident education.

Conclusion

Despite the remarkable growth and increased clinical activity of the DRS since its inception, Dr. Gaffney's ideals continue to serve as the lifeline for the division. The DRS still maintains the consultative pediatric role that he originated, but the inpatient activity has grown with the pediatric hospitalist movement at the same time. The division also maintains an active outpatient clinic. This dual function allows the DRS to continue to serve the community in a unique manner.

In an ideal situation, a child could be cared for by 1 physician from childhood through adolescence. This physician could care for the child from the first days in the nursery, through multiple well‐child and sick visits, and during any hospitalizations. In 1992, the American Academy of Pediatrics (AAP) introduced the concept of the medical home to provide accessible, continuous, comprehensive, family‐centered, coordinated, and compassionate care for children.1 These ideals were reaffirmed in a 2002 AAP policy statement.2 The demands for outpatient care have become more intense and patient safety issues have become a public focus. Combined with the necessity for increased efficiency, the ideal medical home has become difficult to achieve simultaneously in both an outpatient and inpatient setting. This has led to the growth of the hospital specialist or hospitalist, a term first coined by Drs. Wachter and Goldman in August 1996.3 A hospitalist has been defined as a physician whose primary professional focus is the general medical care of hospitalized patients . . . teaching, research, and leadership related to hospital care.4 Despite early concerns voiced after the publication of this landmark article, there has been tremendous growth in the number of hospitalists nationwide.5 Initially, this growth was seen in adult medicine, but pediatric hospitalists have become increasingly more common. There are an estimated 10,000 to 20,000 hospitalists in the profession, with more than 30,000 expected by the end of the decade, with just over 10% being pediatric hospitalists.4 Several studies have shown the benefits of pediatric hospitalist programs with decreased length of stay, hospital charges, and utilization of unnecessary tests and therapies, and increased satisfaction among the physicians, students, and patients.611

History of the Diagnostic Referral Service

Despite the coining of the term hospitalist in 1996, specialization in inpatient care existed in various forms long before then.12 The Diagnostic Referral Service (DRS) at Children's Hospital of Pittsburgh (CHP) initially began under the guidance of Dr. Edmund R. McCluskey, chairman of the Department of Pediatrics at CHP, and Dr. Paul C. Gaffney, a beloved and revered clinician and educator. In 1951, Dr. Gaffney joined the full‐time CHP faculty after residency, chief residency, and a year of fellowship, providing his expertise in hematology and oncology. Over time, Dr. Gaffney's role expanded to that of a master physician, and pediatricians and family practitioners in the community began sending their most diagnostically challenging patients to be seen in his clinic. His activities further extended to providing inpatient care and consultations for complex patients.

With these growing responsibilities, Dr. Gaffney formed the DRS as a separate division within the Department of Pediatrics in the mid‐1970s and began developing the division. The group, then and now, is comprised of general pediatricians who provide multidisciplinary care for hospitalized children as well as for ambulatory consultations. The DRS was initially described in the literature 20 years ago; the roles of the 5 full‐time physicians at that time included a variety of clinical, teaching, and scholarly activities, as both inpatient and outpatient consultative physicians.13 Though much growth has occurred within the division since then, Dr. Gaffney's initial goals of providing excellent patient care and education in an academic setting still remain at the heart of each group member.

Growth of the Division

In March 2002, there were 4 full‐time physicians within the DRS. A remarkable increase in the group size has occurred since then, and currently there are 16 physicians. Each member of the division is assigned a specific activity, either outpatient or inpatient, for at least a 5‐day block. This allows the division to provide continuity in the care of complex patients in both settings and help maintain a medical home for these challenging patients. The primary care physician (PCP) remains responsible for primary care of the patient while the DRS can help manage the patient's complexities in the outpatient and inpatient setting. In essence, there is joint patient ownership between the DRS and the PCP, with relegation of different skill sets to provide a complete medical home for the patient.

The current activities of the group are summarized as follows.

Inpatient Activities

Inpatient Care

Corresponding to the growing pediatric hospitalist movement in the past decade, several area PCPs began requesting that their office patients be followed by the DRS when admitted to CHP. In 2003, the DRS physician referral list had 150 physicians, and it currently has over 325 physicians who refer the inpatient care of their patients to the DRS. These practices are located in a variety of locations in Western Pennsylvania, Ohio, and West Virginia. There are only 7 private practices that continue to maintain their admitting privileges to CHP, and these practices account for 0.5% of general pediatric admissions. The remaining admissions (those PCPs not on the DRS referral list or with admitting privileges) are covered by a rotating attending physician; 85% of the time this is a DRS physician. Therefore, while 0.5% of general pediatric admissions are cared for by private pediatricians, >95% of all general pediatric admissions are cared for by DRS, with the remaining patients cared for by a small number of pediatric subspecialists who occasionally serve as rotating attendings.

Associated with the increase in referrals has been a marked increase in inpatient activity (Figures 0, 0, 0, 13). Nearly 1 of every 4 CHP discharges and 1 of every 3 observation patients are cared for by the DRS.

Figure 1

Figure 2

Figure 3

The DRS division has seen an increase in complex inpatient admissions as well as a much larger number of routine pediatric admissions from the community PCP referrals, as described above. Statistically, this has resulted in an overall stable to slightly increased inpatient complexity for the DRS group. During this same time period, there was a steep decrease in DRS inpatient length of stay followed by maintenance at the shorter length of stay thereafter. Inpatient complexity has increased throughout CHP, yet the same decreases in length of stay have not been seen universally in all the divisions (Figure 4). The advantage that a hospitalist group can bring in decreasing length of stay (and, thereby, hospital costs) has been seen in hospitalist programs around the country.6, 7

Figure 4

Each member of the group attends on the general pediatric ward for 9 to 10 months per year as compared to 2 months per year in 1986. In order to prevent job‐related fatigue, 6 to 8 members of the group typically attend on the ward at the same time so that individual patient volume is more manageable. On average, each individual physician is responsible for about 6 to 8 patients per day. This census allows for daily education of residents and medical students as well as faculty participation in a variety of administrative activities. Despite emphasis on careful documentation and billing for both inpatient and ambulatory activities, the division, like most other pediatric hospitalist divisions, depends upon financial support from CHP and the Department of Pediatrics.

There is a wide variety of diagnoses that are seen in the inpatient setting, with notable similarities and differences when compared to 2 decades ago (Table 1). For example, asthma, gastroenteritis, and bronchiolitis continue to be frequent diagnoses, with bronchiolitis admissions becoming more frequent, following national trends.14

Table 1.Most Common Inpatient Diagnoses

FY 1986		FY 2007
Rank	Diagnosis	Rank	Diagnosis
Abbreviations: FY, fiscal year; NOS, not otherwise specified.
1.	Asthma	1.	Bronchiolitis
2.	Gastroenteritis	2.	Asthma
3.	Failure to thrive	3.	Pneumonia
4.	Seizures	4.	Dehydration
5.	Bronchiolitis	5.	Viral enteritis
6.	Pneumonia	6.	Viral infection, NOS
7.	Suspected sepsis	7.	Esophageal reflux
8.	Apneic episodes	8.	Fever
9.	Meningitis	9.	Cellulitis
10.	Otitis media	10.	Convulsions

In general, a DRS faculty member becomes the primary resource for families with medically complex children. The same faculty member tends to follow the patient in the consultative ambulatory clinic, be available for phone calls, and, if possible, follow them as the attending physician when the patient is admitted. This provides a degree of continuity generally not seen in many other hospitalist programs and has the potential to increase patient and physician satisfaction as well as patient safety.

Outpatient Activities

Education

Teaching is a major role for the DRS, and the division is closely involved in leadership in medical education. Two members are directors of the pediatric physical exam course for first‐year and second‐year medical students, 2 members are the third‐year medical school Pediatric clerkship directors, 1 member is co‐director for the fourth‐year acting internship, 1 member is co‐director of the advanced pediatric interviewing program, 1 member is director of the pediatric medical education program, and 1 member is associate residency program director.

The entire group is involved with teaching at all of these levels. The majority of the group is involved with formal mentoring and advising of residents and medical students. The only general pediatric aspect of student and resident medical education in which the DRS is no longer involved is ambulatory pediatric medicine. The full‐time ambulatory faculty is responsible for the primary ambulatory care experience. However, many residents choose to complete an elective in DRS, including the outpatient clinic, to become exposed to the different diagnostic dilemmas and coordination of care visits that they may not see in their primary care continuity clinics.

The division always welcomes new teaching challenges and incorporates new methods of teaching as opportunities arise. For example, the recent family‐centered rounds initiative allowed for new teaching methods that were not previously possible. The team, comprised of a senior resident, 2 interns, 2 students, and an attending physician rounds at the bedside with permission from the parent. The patient's nurse, when available, and a pharmacist are often a part of the team. The case is presented by the student or intern (directed to the parent), and the case is discussed and clarified for the family. A plan for the day is presented and discussed with the family for approval. Through family‐centered rounds, the DRS attending provides patient specific teaching and role modeling during rounds that would not otherwise have been possible with classical didactic teaching. This method of daily rounding also allows for the patients, families, nurses, nursing students, medical students, and residents to be taught by the attending physician simultaneously. Additionally, it affords the nurses the opportunity to participate in medical decision‐making, and the house staff have perceived fewer pages by the nurses to clarify clinical issues.

The EH program also provided new teaching opportunities. Through the EH, the house staff and students are exposed to direct attending teaching in the evenings that otherwise would not occur, such as direct observation of student and resident histories and physical examinations. Based on resident evaluations and comments to the residency program directors, this teaching experience is deemed to be valuable and effective. In fact, since the EH program's inception 3 years ago, 2 EHs have been selected by the residents as Teacher of the Year.

Patient Safety

Patient safety and reduction of medical errors is a major focus of the entire group. One DRS member serves within the hospital administration as Medical Director for Clinical Excellence and Service to enhance patient safety hospital‐wide. One DRS member orchestrates a monthly house staff meeting entitled To Err is Human which provides a nonthreatening environment for residents to discuss medical errors or difficult situations that they have encountered. Two DRS members are part of the Physician Advisory Committee, which serves as a bridge between the information technology group and clinicians. This committee has aided in achieving a smooth transition to a completely electronic medical record (EMR) and works together to use the tools of an EMR to enhance patient safety. This successful EMR implementation was recognized by the Health Information Management Systems Society in October 2008. Additionally, stemming from several successful patient safety initiatives, CHP was 1 of only 7 children's hospitals recognized for patient safety in 2008 by Leapfrog, the nation's premier patient safety evaluation group.

Scholarly Activity

Previously, the DRS was responsible for the general medical care of liver transplant recipients, and many prior publications from the division focused on these patients.13 The division no longer provides that service, and the current focus of scholarly activity is publication of case reports, book chapters, and review articles. There were 13 publications from the division members in the past 3 years. The group also serves as a major resource within the department by referring patients that fulfill the clinical criteria for ongoing clinical studies in other divisions.

One member of the group continues to serve as senior editor of the Atlas of Pediatric Physical Diagnosis, which is currently in its fifth edition. Several members of the group contribute chapters to this well‐known text. One member is an editor and another is a specialty reviewer for FirstConsult.com, a website for physicians. Another member has served as an associate editor for the Journal of Pediatrics Grand Round Section.

At the University of Pittsburgh, both tenured and nontenured faculty promotions carry the same title without a prefix. Academic promotions for clinician‐educators center around clinical excellence and innovation in education. The 3 senior members of the group have been promoted to professor (1 tenured, 2 nontenured), and 2 other members are currently in consideration for promotion to associate professor. Two members of the group have been elected to the School of Medicine Academy of Master Educators, which recognizes and rewards excellence in education.

Future Goals

Future goals include expansion and refinement of the division's current inpatient and ambulatory activities. The group increased in size to 16 physicians in July 2008 due to the increased inpatient volume and growing demand for outpatient referrals. The family‐centered rounds initiative will continue to be refined to provide the best possible service to the patients and their families. The members of the group will have increased activity and involvement at the regional and national level with the growing pediatric hospitalist movement.

A pediatric hospitalist fellowship program certainly would be feasible in the current environment. At this writing, there are only 8 pediatric hospitalist fellowship programs nationwide. The outpatient/emnpatient environment that the DRS provides the community would certainly provide a unique training environment for a hospitalist fellowship. The diversity in hospitalist divisions nationwide and the standardization of fellowship training is an important task for the future.23

Discussion

Pediatrics has undergone major changes since the original description of the DRS 20 years ago.7 These changes have revolutionized the practice of pediatrics in both the ambulatory and inpatient settings. The DRS role has changed significantly along with the national trends in pediatric hospitalist growth over the past decade. Currently, 90% of the division's clinical activity is inpatient care. In essence, this is an extension of the original consultant role, but the model has been extended to provide inpatient multidisciplinary care for pediatric patients.

Despite the remarkable growth in inpatient activity, 1 unique advantage to the DRS model is maintenance of an active outpatient consultation clinic focused on providing a multidisciplinary medical home for chronically ill patients. DRS faculty are able to coordinate the care of these complex patients while not usurping the primary care responsibility of the community physician. The same faculty are able to extend the continuity of care to the inpatient setting should the patient require admission.

There have been several innovations that the DRS has implemented over the past decade. The LSU was designed to provide effective and efficient care. The EH program extends attending in‐house coverage without the disadvantages of 24‐hour, 7‐day‐per‐week coverage. Expanding services to include The Children's Home allows for easier transition to home for technology‐dependent patients and families. At the same time, DRS continues to strive for innovative clinical leadership as well as creative and effective student and resident education.

Conclusion

Despite the remarkable growth and increased clinical activity of the DRS since its inception, Dr. Gaffney's ideals continue to serve as the lifeline for the division. The DRS still maintains the consultative pediatric role that he originated, but the inpatient activity has grown with the pediatric hospitalist movement at the same time. The division also maintains an active outpatient clinic. This dual function allows the DRS to continue to serve the community in a unique manner.

Pulmonary embolism (PE) is the most common preventable cause of death in hospitals,1 accounting for approximately 10% of hospital deaths. Most cases of PE result from dislodged lower extremity thrombi, so that deep vein thrombosis (DVT) and PE are manifestations of the same disorder, venous thromboembolism (VTE). Even though the majority of hospitalized patients are at increased risk for VTE and proven preventive measures have long been available, most patients do not receive appropriate care.2

Recent surgery is a well‐recognized risk factor for VTE, and surgeons have prescribed prophylactic therapies more consistently than other specialists.3 At the same time, prevention of VTE among hospitalized medical patients has been neglected.4 The American College of Chest Physicians recommends pharmacologic VTE prevention for most acutely ill medical patients, and advises prevention using mechanical devices when pharmacologic intervention is contraindicated.1

In Department of Veterans Affairs (VA) hospitals, compliance with preventive guidelines in surgical patients has been high. During October through December 2007, according to the Office of Quality and Performance, Veterans Health Administration, the national average for administration of VTE prophylaxis within 24 hours of surgery in VA hospitals was 92%. No comparable systemwide performance measure has been applied for medical patients, but an assessment involving intensive care unit patients has been completed and plans are underway for an evaluation of anticoagulant use among all inpatients considered to be at increased risk.

We sought to determine the extent to which hospitalized VA medical patients receive VTE preventive care in accordance with evidence‐based recommendations. Because quality of care may vary in hospitals based on teaching status,5 a secondary goal was to ascertain whether teaching and nonteaching facilities differ with respect to the delivery of care for VTE prevention.

Patient Populations Methods

We examined compliance with accepted VTE clinical practice guidelines in 2 patient populations. First, the care of patients at risk for developing VTE was evaluated for evidence of appropriate preventive measures. Second, the care of patients who developed PE while hospitalized was evaluated for evidence of preventive therapy prior to the event.

We identified patients discharged from VA acute care hospitals during the period April 1, 2006 to March 31, 2007, excluding patients hospitalized for less than 48 hours. We also excluded hospitalizations of VA patients at military and private hospitals because oversight of the quality of care provided at those facilities is beyond the purview of the Inspector General.

We then defined 2 distinct populations:

• 1

  Medical patients at increased risk for VTE. These patients were identified by: (1) age 75 years at the time of admission; and (2) hospitalization with a principal discharge diagnosis of heart failure (International Classification of Diseases, 9th edition [ICD‐9] code 428). Elderly heart failure patients were chosen because advanced age and heart failure are recognized VTE risk factors, medical inpatients have been identified as being neglected in hospital VTE prevention efforts, and the VA was conducting no systemwide assessment of this aspect of care.

• 2

  Patients with established PE. These patients had any discharge diagnosis pulmonary embolism and infarction (ICD‐9 codes 415.1 or 415.19), but those with the diagnostic code personal history of venous thrombosis and embolism (V12.51) were excluded.

Within each population, the discharge date defined an index hospitalization for evaluation. For patients discharged more than once with a qualifying diagnosis during the study period, we analyzed only the most recent hospitalization.

Results

Discussion

Based on a random sample of 4963 elderly heart failure patients admitted to VA hospitals during a 1‐year period, we estimated that 63% received recommended interventions aimed at preventing VTE. Although differences in methodology limit comparisons with published reports, this rate is similar to those observed at individual hospitals,1113 in large multicenter registries of patients with DVT or at risk for VTE,14, 15 and in a recent multinational cross‐sectional study.16 Notably, chronic outpatient anticoagulation that was continued during hospitalization accounted for nearly one‐half of patients receiving preventive care. Compliance did not differ between teaching and nonteaching hospitals.

In a complementary approach to examining the extent of preventive care, we identified 1448 patients discharged with a diagnosis of PE. Most of these patients were excluded because they did not have a new event while hospitalized. Eleven (17%) of the 64 patients with confirmed in‐hospital PE received no preventive care before the event. An additional 18 (28%) received suboptimal heparin regimens or mechanical prophylaxis in the absence of contraindications to anticoagulation. As with the patients at risk for VTE, patients with established PE at teaching and nonteaching hospitals received similar rates of preventive care. Contrary to our expectation, the observed difference in rates between types of hospitals favored nonteaching hospitals. However, the sample size for this comparison was small and the difference did not reach statistical significance.

This study's population‐based approach permits conclusions about the performance of the VA's entire system of acute care hospitals. The results indicate that proven preventive therapies are often neglected at VA hospitals, but overall performance is probably comparable to other settings. VA employs an extensively implemented electronic medical record (EMR) and superior performance might have been expected. However, these results suggest limitations in the EMR as it is currently deployed. Successful efforts probably require a multifaceted approach incorporating decision support and institutional standardization.17

Several additional findings warrant comment. Patients with malignancies accounted for 9 of 10 patients who had PE after receiving no prior anticoagulation. Recent surgery was also a contributing factor for 4 of these cancer patients. Although both cancer and surgery are well‐known risk VTE factors, clinicians may not appreciate the extremely high risk associated with the combination.18 Particular effort may be warranted to ensure prophylaxis in this group, and more intensive measures may be necessary.

These results reveal several barriers to the accurate retrospective measurement of preventable inpatient PE. First, the use of discharge diagnoses to monitor the occurrence of inpatient PE is fraught with hazard. In this study, even after excluding patients with a discharge diagnostic code indicating a past history of PE, very few identified patients in fact had an acute or recent event. In addition, many patients were clearly admitted after having the onset of symptoms as outpatients. Further, reliance on discharge diagnoses alone can lead to the inclusion of patients with a presumptive diagnosis made without the advantage of imaging studies or postmortem examination. Although we overcame these barriers through careful record review and strict diagnostic criteria, our results suggest that efficient performance improvement efforts may require ongoing concurrent review.

There are several limitations of this study. First, we excluded PE patients whose diagnoses were made before the third hospital day. Some of these patients may have had events attributable to recent prior hospitalizations and should have received VTE prophylaxis. Second, we considered preventive measures applied at least 24 hours prior to PE to be acceptable evidence of prevention, potentially neglecting prior periods without treatment that might confer increased risk. Bias due to either of these limitations could exaggerate the compliance rates we report. Finally, the retrospective design of this study did not allow for consistent assessments of whether patients had the risk factor of immobility. Nevertheless, immobility was obvious for the 10 patients with PE who had no prior anticoagulation, all of whom had 2 or more risk factors.

Despite an acknowledged need for improvements in clinical practice, past efforts have had mixed results. For instance, in 1 study at a hospital with a well‐established EMR, computer alerts led to substantial improvement in the use of preventive measures and in VTE outcomes, but overall compliance remained low.19 On the other hand, a multidisciplinary approach can achieve marked reductions in preventable VTE events.20 Key elements of such an approach are a simplified risk assessment tool and concurrent monitoring of patient treatments and outcomes. The Agency for Healthcare Research and Quality has recently published a guide that outlines strategies for achieving breakthrough levels of improvement in the prevention of VTE.21

In conclusion, this population‐based study of hospitalized veterans with PE or at risk for VTE found compliance comparable to rates in published reports. Missed opportunities for prevention included inappropriate and inadequate interventions. Using discharge diagnoses to monitor the occurrence of inpatient PE is of limited value, and efficient performance improvement efforts may require ongoing concurrent review.