Reed G Williams1, Steven Verhulst2, John D Mellinger3, Gary L Dunnington4. 1. Department of Surgery, Indiana University School of Medicine, Indianapolis, Indiana. Electronic address: reedwill@iupui.edu. 2. Department of Medical Education, Southern Illinois University School of Medicine, Springfield, Illinois. 3. Department of Surgery, Southern Illinois University School of Medicine, Springfield, Illinois. 4. Department of Surgery, Indiana University School of Medicine, Indianapolis, Indiana.
Abstract
OBJECTIVE: A valid measure of resident operative performance ability requires direct observation and accurate rating of multiple resident performances under the normal range of operating conditions. The challenge is to create an operative performance rating (OPR) system that: is easy to use, encourages completion of many ratings immediately after performances and minimally disrupts supervising surgeons' work days. The purpose of this study was to determine whether a score based on a single-item overall OPR provides a valid and stable appraisal of resident operative performances. DESIGN: A retrospective comparison of a single-item OPR with a gold-standard rating based on multiple procedure-specific and general OPR items. SETTING: Data were collected in the general surgery residency program at Southern Illinois University from 2001 through 2012. PARTICIPANTS: Assessments of 1033 operative performances (3 common procedures, 2 laparoscopic, and 1 open) by general surgery residents were collected. OPRs based on single-item overall performance scale scores were compared with gold-standard ratings for the same performances. RESULTS: Differences in performance scores using the 2 scales averaged 0.02 points (5-point scale). Correlations of the single-item and gold-standard scale scores averaged 0.95. Based on generalizability analyses of laparoscopic cholecystectomy ratings, each instrument required 5 observations to achieve reliabilities of 0.80 and 11 observations to achieve reliabilities of 0.90. Only 4.4% of single-item ratings misclassified the performance when compared with the gold-standard rating and all misclassifications were near misses. For 80% of misclassified ratings, single-item ratings were lower. CONCLUSIONS: Single-item operative performance measures produced ratings that were virtually identical to gold-standard scale ratings. Misclassifications occurred infrequently and were minor in magnitude. Ratings using the single-item scale: take less time to complete, should increase the sample of procedures rated, and encourage attending surgeons to complete ratings immediately after observing performances. Face-to-face and written comments and suggestions should continue to be used to provide the granular feedback residents need to improve subsequent performances.
OBJECTIVE: A valid measure of resident operative performance ability requires direct observation and accurate rating of multiple resident performances under the normal range of operating conditions. The challenge is to create an operative performance rating (OPR) system that: is easy to use, encourages completion of many ratings immediately after performances and minimally disrupts supervising surgeons' work days. The purpose of this study was to determine whether a score based on a single-item overall OPR provides a valid and stable appraisal of resident operative performances. DESIGN: A retrospective comparison of a single-item OPR with a gold-standard rating based on multiple procedure-specific and general OPR items. SETTING: Data were collected in the general surgery residency program at Southern Illinois University from 2001 through 2012. PARTICIPANTS: Assessments of 1033 operative performances (3 common procedures, 2 laparoscopic, and 1 open) by general surgery residents were collected. OPRs based on single-item overall performance scale scores were compared with gold-standard ratings for the same performances. RESULTS: Differences in performance scores using the 2 scales averaged 0.02 points (5-point scale). Correlations of the single-item and gold-standard scale scores averaged 0.95. Based on generalizability analyses of laparoscopic cholecystectomy ratings, each instrument required 5 observations to achieve reliabilities of 0.80 and 11 observations to achieve reliabilities of 0.90. Only 4.4% of single-item ratings misclassified the performance when compared with the gold-standard rating and all misclassifications were near misses. For 80% of misclassified ratings, single-item ratings were lower. CONCLUSIONS: Single-item operative performance measures produced ratings that were virtually identical to gold-standard scale ratings. Misclassifications occurred infrequently and were minor in magnitude. Ratings using the single-item scale: take less time to complete, should increase the sample of procedures rated, and encourage attending surgeons to complete ratings immediately after observing performances. Face-to-face and written comments and suggestions should continue to be used to provide the granular feedback residents need to improve subsequent performances.
Authors: Jay N Nathwani; Carly E Glarner; Katherine E Law; Robert J McDonald; Amy B Zelenski; Jacob A Greenberg; Eugene F Foley Journal: J Surg Educ Date: 2016-11-25 Impact factor: 2.891