- Email: [email protected]
Evaluating the Effect of Emergency Residency Training on Productivity in the Emergency Department
The Journal of Emergency Medicine, Vol. 45, No. 3, pp. 414–418, 2013 Copyright Ó 2013 Elsevier Inc. Printed in the USA. All rights reserved 0736-4679/$ - see front matter
http://dx.doi.org/10.1016/j.jemermed.2013.03.022
Education EVALUATING THE EFFECT OF EMERGENCY RESIDENCY TRAINING ON PRODUCTIVITY IN THE EMERGENCY DEPARTMENT Daniel J. Henning, MD, Daniel C. McGillicuddy, MD, and Leon D. Sanchez, MD Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts Reprint Address: Daniel J. Henning, MD, Department of Emergency Medicine, Beth Israel Deaconess Medical Center, One Deaconess Road, W-CC2, Boston, MA 02215
, Abstract—Background: Resident productivity, defined as patients seen per unit time, is one measure that is used to assess the performance and educational progress of residents in the emergency department (ED). One published study suggested that emergency residency training (EM) does not improve productivity compared with that in other specialties, including internal medicine (IM). Objectives: This study assesses how EM and IM trainees perform in the ED and illustrates how resident productivity changes through the academic year. Methods: A retrospective review of attending physicians and residents working 8-h shifts in the higher acuity zone of a large-volume, tertiary, academic health care center was performed for July 2009, October 2009, January 2010, and April 2010. The total number of patients seen primarily and admitted during each shift was recorded. ED volume was approximated by the number of patients seen by the attending physician, and acuity was approximated by admission rate. A mixed model regression assessed the impact of year and type of residency training (e.g., EM1, EM2, IM1, and IM2), ED volume, and acuity on resident productivity (number of patients per shift). The study was granted waiver of informed consent by our institutional review board. Results: We reviewed 936 shifts. After adjusting for acuity and ED volume, the EM1 group had a significant increase in patients per shift
over the year, from 6.11 in July to 10.3 in April (p < 0.001). No other group increased productivity significantly. Conclusions: The first EM training year leads to a significant change in productivity that separates EM from IM residents. This contradicts the previous assertion that non-EM residents have the same productivity as EM residents in the ED. Ó 2013 Elsevier Inc. , Keywords—education; productivity
INTRODUCTION Productivity in emergency residency (EM) training has been established as a metric for resident clinical performance because it tends to increase with each year of training (1–3). Whether measured as relative value units (RVUs) or patients seen per hour, productivity represents the amount of patient care a physician is able to provide in a set time, and it facilitates the evaluation of an individual’s clinical performance. Beyond year of training, the relationship of other variables to productivity has been studied, including shift length, consecutive shifts, and emergency department (ED) volume (4–7). One previous study suggested that non-EM residents were as productive as EM residents in the ED, countering the anecdote that non-EM residents worked up significantly fewer patients per hour (8). This study compared residents from one institution that did not have an EM residency with previously published
Presented as a poster at the American College of Emergency Physicians Research Forum, Las Vegas, NV, September 28– 29, 2010. Support was provided entirely by the Beth Israel Deaconess Department of Emergency Medicine.
RECEIVED: 5 May 2012; FINAL SUBMISSION RECEIVED: 11 October 2012; ACCEPTED: 15 March 2013 414
Emergency Training and Productivity
productivity results, so it could not account for differences in systems and patient characteristics (9). Since the aforementioned study, there has not been a re-evaluation of the suggestion that EM residency does not confer a significant benefit in the productivity of residents in the ED. Expecting non-EM residents to have the same productivity of EM residents confuses the educational goals and staffing expectations for nonEM residents in the ED. Because many academic EDs are staffed by both EM and non-EM residents, including internal medicine (IM) residents, evaluating the productivity of IM and EM residents will create more current expectations for IM residents in the ED. At the same time, evaluating how productivity changes through the year will improve resident staffing and inform how educators evaluate educational growth. Goals of This Investigation Our study assesses the assertion that IM resident productivity is similar to the productivity of EM residents in the ED of a single institution. Secondarily, this study evaluates how productivity changes throughout the academic year. METHODS We performed a retrospective review of resident productivity during the 2009–2010 academic year. The study design was approved by our institutional review board with waiver of informed consent. This study was performed at an urban, academic, tertiary care ED with 55,000 annual visits. The study ED is staffed 24 h a day with residents from a 3-year EM residency, along with first- and second-year IM residents. The ED is separated into higher and lower acuity zones, which patients are triaged into after a nursing assessment of illness severity and resources that will be required for patient care. The higher acuity zone is made up of 18 beds plus 6 trauma/resuscitation beds, and it is staffed 24 h a day with a second-year EM resident and at least one other resident. First-year EM and first- and second-year IM residents work varying shifts in the higher acuity area. Residents are supervised by a third-year EM resident who does not typically see patients primarily. Shifts are all 8 h in duration: 7 AM–3 PM, 3 PM–11 PM, or 11 PM– 7 AM. IM residents spend 2 weeks per year in the ED. During the study period, there were no changes to the staffing patterns for residents in the ED. This study included all shifts worked by first- and second-year EM and IM residents in the ED that started during the study months. Shift data from July 2009, October 2009, January 2010, and April 2010 were abstracted for analysis. Because IM residents staff only
415
the higher acuity area, we excluded all shifts from the lower acuity area. Computer-generated shift lists from our local patient database are available for residents working in the ED. These lists, which included information related to number of patients seen primarily, inherited at shift-change, admitted, and discharged, were manually matched to the electronically published shift schedule. These data were collected by the principal investigator and entered into an Excel spreadsheet. A data-checking step was included to ensure accuracy between computer records and actual patients seen. Our primary outcome was patients seen per shift, which was recorded for every high-acuity shift by resident. Other shift-based data collected were admissions per shift and number of patients seen by the high acuity attending or attending physicians for each shift. The number of patients seen by attending physicians was used as a proxy for daily ED volume. Basic demographic information, including specialty and year of training, were also collected for each resident. Data analysis was performed by a consulting statistician from Technomics Research, LLC (Minneapolis, MN). A mixed model was used to evaluate productivity, with the dependent variable being productivity measured as the number of patients seen per 8-h shift and the independent variables being month, shift, program (EM or IM), year of residency (1st or 2nd), acuity measured as the percentage of patients admitted, and total ED volume over the entire shift. A compound symmetrical covariance structure was used to model the correlation expected among the multiple observations collected for each resident. A stepwise backward elimination technique was used to determine the final model. First, the main effects (i.e., month, shift, program, year of training, acuity, and ED shift volume) and all possible interaction terms were included in the model. The interaction terms were reviewed for statistical significance (p < 0.05) from the largest to the smallest number of terms. First, the fiveand six-way interaction terms were reviewed for statistical significance. Those that were statistically significant remained in the model and the model was refit. Then the four-way interaction terms were reviewed and retained as appropriate. This process was repeated for the three- and two-way interaction terms. None of the main effects was removed from the model. The resultant model was established as the final model and is discussed in the manuscript. Because fit statistics are not used in this process of determining the final model, an R2 value is not reported. We used admission rate as a proxy for ED acuity. To evaluate acuity, only the main effects (e.g., month, shift, program, and year) and two intuitive interaction terms—month*shift and program*year—were included.
416
D. J. Henning et al.
Similar to the process previously described to create a model for productivity, a mixed model was used with a compound symmetrical covariance structure to model the correlation expected among repeated measures. The familywise type I error rate for the data presented is 5%. When multiple pairwise comparisons were performed, the Tukey–Kramer adjustment was used to preserve the familywise type I error at 5% across all pairwise comparisons performed for a given main effect or interaction term. RESULTS The total number of shifts for each group and the breakdown of the shifts worked were not equal among the study groups. EM2 residents worked the greatest number of shifts (380), followed by the EM1 (205), IM1 (179), and IM2 (172) groups. The breakdown of shift times was likewise unequal, reflecting an administrative mandate that an EM2 be scheduled for every shift and the preference for either an EM1 or IM2 resident (but not an IM1 resident) to work overnight shifts. EM2 residents primarily evaluated 4114 patients during the study months–more patients than any other group–followed by the EM1, IM2, and IM1 groups (Table 1). Over the course of the academic year, EM1 residents experienced a significant increase in the number of patients seen per shift. In July, they saw a mean of 6.3 patients per shift, but by June this had increased to 10.3 patients per shift (p < 0.0001). None of the other groups had a significant increase in productivity as the year progressed. EM1 and IM2 groups begin the academic year with similar productivity rates (EM1 6.3 patients/shift; IM2 6.38 patients/shift) but separate through the year. IM residents start at 3.9 patients per shift and show a nonsignificant trend toward increased productivity as the year progresses (Figure 1). Overall, EM2 residents admitted 60% of their patients, compared with 50% of patients seen by IM2, 49% for EM1, and 47% for IM1 residents. Overall, there was no significant difference in the admission rates for any study month for the entire ED. A significant difference in admission rates was not present among EM1, IM1, and Table 1. Shift Distribution Program and Year EM1 EM2 IM1 IM2
Total Shift Total No. of No. of Shifts 7 AM–3 PM 3 PM–11 PM 11 PM–7 AM Patients 205 380 179 172
61 121 98 71
109 137 77 38
35 122 2 63
EM = emergency medicine; IM = internal medicine.
1595 4114 882 1231
Figure 1. Resident productivity per month, based on the number of patients per shift between July 2009 and April 2010. EM1 (p < 0.0001), EM2 (p = 0.7366), IM1 (p = 0.4918), and IM2 (p = 1.0000). EM = emergency medicine; IM = internal medicine.
IM2 groups. A significant difference was present in the admission rates among shifts, with the 3 PM–11 PM shift having the highest admission rate (59%) and the 11 PM– 7 AM shift having the lowest admission rate (43%; Table 2). DISCUSSION EM interns show a significant increase in resident productivity over the course of the first year of EM residency, which was not shown by any of the other study groups. EM1 and IM2 residents begin the academic year with similar productivity rates and separate through
Table 2. Comparison of Admission Rate to Specialty, Year of Training, Month, and Shift* Admission Rate (6 SD) Program and Year EM1 EM2 IM1 IM2 Month July October January April Shift 11 PM–7 AM 7 AM–3 PM 3 PM–11 PM
49 6 2 60 6 2 47 6 2 50 6 2 52 6 1 51 6 1 52 6 1 51 6 2 43 6 2 52 6 1 59 6 1
EM = emergency medicine; IM = internal medicine; SD = standard deviation. * The admission rate was significantly higher for EM2 residents compared with that in the other groups, but there was no significant different among EM1, IM1, and IM2 admission rates. There was no difference in admission rates through the year, but a significant difference existed between shifts (p < 0.0001).
Emergency Training and Productivity
the year, which suggests that professional growth during the first year of EM training provides the skillset to see more patients during a set amount of time in the ED— even compared with that of a more advanced IM resident. These changes in productivity occurred despite similar levels of acuity among IM1, IM2, and EM1 residents and as indicated by the admission rates throughout the year. This study establishes an objective difference in performance between residents trained in EM and IM programs using metrics that affect ED patient care and flow. The difference between EM and IM productivity revealed here contradicts the previous suggestion of similar productivity between EM and non-EM residents (8). Certainly, the far greater time that EM residents spend in the ED affects this conclusion. In fact, Figure 1 suggests that it takes 6 months for EM interns to show a significant difference in productivity–far more time than IM residents spend in the ED. From an educational standpoint, this study shows that a significant amount of time is required for EM residents to increase their productivity; for IM residents, overemphasizing speed during the brief time spent in the ED could detract from the benefit of rotating in the ED. Administratively, these differences should be taken into account when scheduling resident ED staffing. Future analyses will expand the present analysis to include surgical specialties. Also, additional study needs to incorporate lower acuity patient care, because less complex patient encounters may affect productivity differences between specialties. Limitations When evaluating EM and IM resident productivity, we only evaluated high-acuity shifts, which are almost exclusively staffed by EM and IM residents in our ED. Although the largest resident group studied by Stone et al. was composed primarily of IM residents, that study also included residents in obstetrics and gynecology and surgery, among others (8). In our ED, surgery, obstetrics and gynecology and podiatry residents almost exclusively staff the lower acuity zone, because they are all interns and most minor surgical procedures are triaged to
417
this zone. This may bias our results if IM residents are significantly less productive than residents from surgical specialties, although Stone et al. also found that there was no significant productivity difference among any non-EM specialties (8). This study does not account for the effect of patients who are signed out between shifts. All residents have patients signed out to them as their shifts start, but how many each resident takes and how these signed-out patients affect the residents’ ability to pick up new patients was not part of our study. CONCLUSIONS Emergency residency training affects resident productivity in the ED. When compared with IM residents, EM interns show a greater increase in productivity through the year. These changes in productivity should be taken into account when planning resident staffing of the ED. Acknowledgment—Statistical analysis was provided by Teresa Nelson of Technomics Research, LLC (Minneapolis, MN).
REFERENCES 1. Brennan DF, Silvestri S, Sun JY, Papa L. Progression of emergency medicine resident productivity. Acad Emerg Med 2007;14:790–4. 2. DeBehnke D, O’Brien S, Leschke R. Emergency medicine resident work productivity in an academic emergency department. Acad Emerg Med 2000;7:90–2. 3. Shiber J, Fontane E. Progression of emergency medicine resident productivity. Acad Emerg Med 2008;15:107. 4. Jeanmonod R, Damewood S, Brook C. Resident productivity: trends over consecutive shifts. Int J Emerg Med 2009;2:107–10. 5. Jeanmonod R, Jeanmonod D, Ngiam R. Resident productivity: does shift length matter? Am J Emerg Med 2008;26:789–91. 6. Jeanmonod R, Brook C, Winther M, Pathak S, Boyd M. Resident productivity as a function of emergency department volume, shift time of day, and cumulative time in the emergency department. Am J Emerg Med 2009;27:313–9. 7. Ledrick D, Fisher S, Thompson J, Sniadanko M. An assessment of emergency medicine residents’ ability to perform in a multitasking environment. Acad Med 2009;84:1289–94. 8. Stone CK, Stapczynski JS, Thomas SH, Koury SI. Rate of patient workups by non-emergency medicine residents in an academic emergency department. Acad Emerg Med 1996;3:153–6. 9. Keim SM. Can resident physicians be profiled? Acad Emerg Med 2000;7:72–4.
418
D. J. Henning et al.
ARTICLE SUMMARY 1. Why is this topic important? Resident productivity is an objective measure of resident performance, and can be used to follow educational progress in the emergency department (ED). 2. What does this study attempt to show? Emergency medicine training uniquely prepares physicians to work in the ED compared with internal medicine training. 3. What are the key findings? The productivity of emergency medicine residents increases significantly through the first year, but a similar increase is not appreciated for internal medicine residents. 4. How is patient care impacted? These results will help optimize resident staffing in the ED, and they reinforce the trend of staffing EDs with physicians trained in emergency medicine.
http://dx.doi.org/10.1016/j.jemermed.2013.03.022
Education EVALUATING THE EFFECT OF EMERGENCY RESIDENCY TRAINING ON PRODUCTIVITY IN THE EMERGENCY DEPARTMENT Daniel J. Henning, MD, Daniel C. McGillicuddy, MD, and Leon D. Sanchez, MD Department of Emergency Medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts Reprint Address: Daniel J. Henning, MD, Department of Emergency Medicine, Beth Israel Deaconess Medical Center, One Deaconess Road, W-CC2, Boston, MA 02215
, Abstract—Background: Resident productivity, defined as patients seen per unit time, is one measure that is used to assess the performance and educational progress of residents in the emergency department (ED). One published study suggested that emergency residency training (EM) does not improve productivity compared with that in other specialties, including internal medicine (IM). Objectives: This study assesses how EM and IM trainees perform in the ED and illustrates how resident productivity changes through the academic year. Methods: A retrospective review of attending physicians and residents working 8-h shifts in the higher acuity zone of a large-volume, tertiary, academic health care center was performed for July 2009, October 2009, January 2010, and April 2010. The total number of patients seen primarily and admitted during each shift was recorded. ED volume was approximated by the number of patients seen by the attending physician, and acuity was approximated by admission rate. A mixed model regression assessed the impact of year and type of residency training (e.g., EM1, EM2, IM1, and IM2), ED volume, and acuity on resident productivity (number of patients per shift). The study was granted waiver of informed consent by our institutional review board. Results: We reviewed 936 shifts. After adjusting for acuity and ED volume, the EM1 group had a significant increase in patients per shift
over the year, from 6.11 in July to 10.3 in April (p < 0.001). No other group increased productivity significantly. Conclusions: The first EM training year leads to a significant change in productivity that separates EM from IM residents. This contradicts the previous assertion that non-EM residents have the same productivity as EM residents in the ED. Ó 2013 Elsevier Inc. , Keywords—education; productivity
INTRODUCTION Productivity in emergency residency (EM) training has been established as a metric for resident clinical performance because it tends to increase with each year of training (1–3). Whether measured as relative value units (RVUs) or patients seen per hour, productivity represents the amount of patient care a physician is able to provide in a set time, and it facilitates the evaluation of an individual’s clinical performance. Beyond year of training, the relationship of other variables to productivity has been studied, including shift length, consecutive shifts, and emergency department (ED) volume (4–7). One previous study suggested that non-EM residents were as productive as EM residents in the ED, countering the anecdote that non-EM residents worked up significantly fewer patients per hour (8). This study compared residents from one institution that did not have an EM residency with previously published
Presented as a poster at the American College of Emergency Physicians Research Forum, Las Vegas, NV, September 28– 29, 2010. Support was provided entirely by the Beth Israel Deaconess Department of Emergency Medicine.
RECEIVED: 5 May 2012; FINAL SUBMISSION RECEIVED: 11 October 2012; ACCEPTED: 15 March 2013 414
Emergency Training and Productivity
productivity results, so it could not account for differences in systems and patient characteristics (9). Since the aforementioned study, there has not been a re-evaluation of the suggestion that EM residency does not confer a significant benefit in the productivity of residents in the ED. Expecting non-EM residents to have the same productivity of EM residents confuses the educational goals and staffing expectations for nonEM residents in the ED. Because many academic EDs are staffed by both EM and non-EM residents, including internal medicine (IM) residents, evaluating the productivity of IM and EM residents will create more current expectations for IM residents in the ED. At the same time, evaluating how productivity changes through the year will improve resident staffing and inform how educators evaluate educational growth. Goals of This Investigation Our study assesses the assertion that IM resident productivity is similar to the productivity of EM residents in the ED of a single institution. Secondarily, this study evaluates how productivity changes throughout the academic year. METHODS We performed a retrospective review of resident productivity during the 2009–2010 academic year. The study design was approved by our institutional review board with waiver of informed consent. This study was performed at an urban, academic, tertiary care ED with 55,000 annual visits. The study ED is staffed 24 h a day with residents from a 3-year EM residency, along with first- and second-year IM residents. The ED is separated into higher and lower acuity zones, which patients are triaged into after a nursing assessment of illness severity and resources that will be required for patient care. The higher acuity zone is made up of 18 beds plus 6 trauma/resuscitation beds, and it is staffed 24 h a day with a second-year EM resident and at least one other resident. First-year EM and first- and second-year IM residents work varying shifts in the higher acuity area. Residents are supervised by a third-year EM resident who does not typically see patients primarily. Shifts are all 8 h in duration: 7 AM–3 PM, 3 PM–11 PM, or 11 PM– 7 AM. IM residents spend 2 weeks per year in the ED. During the study period, there were no changes to the staffing patterns for residents in the ED. This study included all shifts worked by first- and second-year EM and IM residents in the ED that started during the study months. Shift data from July 2009, October 2009, January 2010, and April 2010 were abstracted for analysis. Because IM residents staff only
415
the higher acuity area, we excluded all shifts from the lower acuity area. Computer-generated shift lists from our local patient database are available for residents working in the ED. These lists, which included information related to number of patients seen primarily, inherited at shift-change, admitted, and discharged, were manually matched to the electronically published shift schedule. These data were collected by the principal investigator and entered into an Excel spreadsheet. A data-checking step was included to ensure accuracy between computer records and actual patients seen. Our primary outcome was patients seen per shift, which was recorded for every high-acuity shift by resident. Other shift-based data collected were admissions per shift and number of patients seen by the high acuity attending or attending physicians for each shift. The number of patients seen by attending physicians was used as a proxy for daily ED volume. Basic demographic information, including specialty and year of training, were also collected for each resident. Data analysis was performed by a consulting statistician from Technomics Research, LLC (Minneapolis, MN). A mixed model was used to evaluate productivity, with the dependent variable being productivity measured as the number of patients seen per 8-h shift and the independent variables being month, shift, program (EM or IM), year of residency (1st or 2nd), acuity measured as the percentage of patients admitted, and total ED volume over the entire shift. A compound symmetrical covariance structure was used to model the correlation expected among the multiple observations collected for each resident. A stepwise backward elimination technique was used to determine the final model. First, the main effects (i.e., month, shift, program, year of training, acuity, and ED shift volume) and all possible interaction terms were included in the model. The interaction terms were reviewed for statistical significance (p < 0.05) from the largest to the smallest number of terms. First, the fiveand six-way interaction terms were reviewed for statistical significance. Those that were statistically significant remained in the model and the model was refit. Then the four-way interaction terms were reviewed and retained as appropriate. This process was repeated for the three- and two-way interaction terms. None of the main effects was removed from the model. The resultant model was established as the final model and is discussed in the manuscript. Because fit statistics are not used in this process of determining the final model, an R2 value is not reported. We used admission rate as a proxy for ED acuity. To evaluate acuity, only the main effects (e.g., month, shift, program, and year) and two intuitive interaction terms—month*shift and program*year—were included.
416
D. J. Henning et al.
Similar to the process previously described to create a model for productivity, a mixed model was used with a compound symmetrical covariance structure to model the correlation expected among repeated measures. The familywise type I error rate for the data presented is 5%. When multiple pairwise comparisons were performed, the Tukey–Kramer adjustment was used to preserve the familywise type I error at 5% across all pairwise comparisons performed for a given main effect or interaction term. RESULTS The total number of shifts for each group and the breakdown of the shifts worked were not equal among the study groups. EM2 residents worked the greatest number of shifts (380), followed by the EM1 (205), IM1 (179), and IM2 (172) groups. The breakdown of shift times was likewise unequal, reflecting an administrative mandate that an EM2 be scheduled for every shift and the preference for either an EM1 or IM2 resident (but not an IM1 resident) to work overnight shifts. EM2 residents primarily evaluated 4114 patients during the study months–more patients than any other group–followed by the EM1, IM2, and IM1 groups (Table 1). Over the course of the academic year, EM1 residents experienced a significant increase in the number of patients seen per shift. In July, they saw a mean of 6.3 patients per shift, but by June this had increased to 10.3 patients per shift (p < 0.0001). None of the other groups had a significant increase in productivity as the year progressed. EM1 and IM2 groups begin the academic year with similar productivity rates (EM1 6.3 patients/shift; IM2 6.38 patients/shift) but separate through the year. IM residents start at 3.9 patients per shift and show a nonsignificant trend toward increased productivity as the year progresses (Figure 1). Overall, EM2 residents admitted 60% of their patients, compared with 50% of patients seen by IM2, 49% for EM1, and 47% for IM1 residents. Overall, there was no significant difference in the admission rates for any study month for the entire ED. A significant difference in admission rates was not present among EM1, IM1, and Table 1. Shift Distribution Program and Year EM1 EM2 IM1 IM2
Total Shift Total No. of No. of Shifts 7 AM–3 PM 3 PM–11 PM 11 PM–7 AM Patients 205 380 179 172
61 121 98 71
109 137 77 38
35 122 2 63
EM = emergency medicine; IM = internal medicine.
1595 4114 882 1231
Figure 1. Resident productivity per month, based on the number of patients per shift between July 2009 and April 2010. EM1 (p < 0.0001), EM2 (p = 0.7366), IM1 (p = 0.4918), and IM2 (p = 1.0000). EM = emergency medicine; IM = internal medicine.
IM2 groups. A significant difference was present in the admission rates among shifts, with the 3 PM–11 PM shift having the highest admission rate (59%) and the 11 PM– 7 AM shift having the lowest admission rate (43%; Table 2). DISCUSSION EM interns show a significant increase in resident productivity over the course of the first year of EM residency, which was not shown by any of the other study groups. EM1 and IM2 residents begin the academic year with similar productivity rates and separate through
Table 2. Comparison of Admission Rate to Specialty, Year of Training, Month, and Shift* Admission Rate (6 SD) Program and Year EM1 EM2 IM1 IM2 Month July October January April Shift 11 PM–7 AM 7 AM–3 PM 3 PM–11 PM
49 6 2 60 6 2 47 6 2 50 6 2 52 6 1 51 6 1 52 6 1 51 6 2 43 6 2 52 6 1 59 6 1
EM = emergency medicine; IM = internal medicine; SD = standard deviation. * The admission rate was significantly higher for EM2 residents compared with that in the other groups, but there was no significant different among EM1, IM1, and IM2 admission rates. There was no difference in admission rates through the year, but a significant difference existed between shifts (p < 0.0001).
Emergency Training and Productivity
the year, which suggests that professional growth during the first year of EM training provides the skillset to see more patients during a set amount of time in the ED— even compared with that of a more advanced IM resident. These changes in productivity occurred despite similar levels of acuity among IM1, IM2, and EM1 residents and as indicated by the admission rates throughout the year. This study establishes an objective difference in performance between residents trained in EM and IM programs using metrics that affect ED patient care and flow. The difference between EM and IM productivity revealed here contradicts the previous suggestion of similar productivity between EM and non-EM residents (8). Certainly, the far greater time that EM residents spend in the ED affects this conclusion. In fact, Figure 1 suggests that it takes 6 months for EM interns to show a significant difference in productivity–far more time than IM residents spend in the ED. From an educational standpoint, this study shows that a significant amount of time is required for EM residents to increase their productivity; for IM residents, overemphasizing speed during the brief time spent in the ED could detract from the benefit of rotating in the ED. Administratively, these differences should be taken into account when scheduling resident ED staffing. Future analyses will expand the present analysis to include surgical specialties. Also, additional study needs to incorporate lower acuity patient care, because less complex patient encounters may affect productivity differences between specialties. Limitations When evaluating EM and IM resident productivity, we only evaluated high-acuity shifts, which are almost exclusively staffed by EM and IM residents in our ED. Although the largest resident group studied by Stone et al. was composed primarily of IM residents, that study also included residents in obstetrics and gynecology and surgery, among others (8). In our ED, surgery, obstetrics and gynecology and podiatry residents almost exclusively staff the lower acuity zone, because they are all interns and most minor surgical procedures are triaged to
417
this zone. This may bias our results if IM residents are significantly less productive than residents from surgical specialties, although Stone et al. also found that there was no significant productivity difference among any non-EM specialties (8). This study does not account for the effect of patients who are signed out between shifts. All residents have patients signed out to them as their shifts start, but how many each resident takes and how these signed-out patients affect the residents’ ability to pick up new patients was not part of our study. CONCLUSIONS Emergency residency training affects resident productivity in the ED. When compared with IM residents, EM interns show a greater increase in productivity through the year. These changes in productivity should be taken into account when planning resident staffing of the ED. Acknowledgment—Statistical analysis was provided by Teresa Nelson of Technomics Research, LLC (Minneapolis, MN).
REFERENCES 1. Brennan DF, Silvestri S, Sun JY, Papa L. Progression of emergency medicine resident productivity. Acad Emerg Med 2007;14:790–4. 2. DeBehnke D, O’Brien S, Leschke R. Emergency medicine resident work productivity in an academic emergency department. Acad Emerg Med 2000;7:90–2. 3. Shiber J, Fontane E. Progression of emergency medicine resident productivity. Acad Emerg Med 2008;15:107. 4. Jeanmonod R, Damewood S, Brook C. Resident productivity: trends over consecutive shifts. Int J Emerg Med 2009;2:107–10. 5. Jeanmonod R, Jeanmonod D, Ngiam R. Resident productivity: does shift length matter? Am J Emerg Med 2008;26:789–91. 6. Jeanmonod R, Brook C, Winther M, Pathak S, Boyd M. Resident productivity as a function of emergency department volume, shift time of day, and cumulative time in the emergency department. Am J Emerg Med 2009;27:313–9. 7. Ledrick D, Fisher S, Thompson J, Sniadanko M. An assessment of emergency medicine residents’ ability to perform in a multitasking environment. Acad Med 2009;84:1289–94. 8. Stone CK, Stapczynski JS, Thomas SH, Koury SI. Rate of patient workups by non-emergency medicine residents in an academic emergency department. Acad Emerg Med 1996;3:153–6. 9. Keim SM. Can resident physicians be profiled? Acad Emerg Med 2000;7:72–4.
418
D. J. Henning et al.
ARTICLE SUMMARY 1. Why is this topic important? Resident productivity is an objective measure of resident performance, and can be used to follow educational progress in the emergency department (ED). 2. What does this study attempt to show? Emergency medicine training uniquely prepares physicians to work in the ED compared with internal medicine training. 3. What are the key findings? The productivity of emergency medicine residents increases significantly through the first year, but a similar increase is not appreciated for internal medicine residents. 4. How is patient care impacted? These results will help optimize resident staffing in the ED, and they reinforce the trend of staffing EDs with physicians trained in emergency medicine.