2019 CYIR Finalized
62 MEDICAL EDUCATION Peter H. Lenz, MD, MEd University of Cincinnati Pulmonary, Critical Care, and Sleep Division, Department of Internal Medicine Cincinnati, OH MEDICAL EDUCATION WEDNESDAY, MAY 22 DUTY HOURS Silber JH, Bellini LM, Shea JA, Desai SV, Dinges DF, Basner M, Even-Shoshan O, Hill AS, Hochman LL, Katz JT, Ross RN, Shade DM, Small DS, Sternberg AL, Tonascia J, Volpp KG, Asch DA. Patient Safety Outcomes under Flexible and Standard Resident Duty-Hour Rules. N Engl J of Med 2019; 380: 905-14. Summary This cluster randomized noninferiority trial is the primary outcome and patient safety reporting arm of the iCOM- PARE (Individualized Comparative Effectiveness of Models Optimizing Patient Safety and Resident Education) 2015- 2016 investigation of flexible versus standard duty hours for internal medicine (IM) residents. 31 IM residency programs were randomized to standard duty hour restrictions and compared on mortality and safety outcomes to 32 residen- cy programs allowed to use more flexible duty hours (most notably removing the 16-hour restriction on shift length). Program directors (PDs) in the flexible duty-hour arm were given latitude, but not required, to remove limits on shift length and mandatory time between shifts. Difference-in- difference analysis evaluated the primary outcome of unad- justed 30-day mortality change from pretrial year to trial year for both groups. Secondary safety outcomes were also evaluated. The change in 30-day mortality for patients in the flexible programs (12.5 % in trial year vs. 12.6% in pre- trial year) was noninferior to that in the standard programs (12.2% in trial year vs. 12.7% in pretrial year). The test for noninferiority was significant (p=0.03) essentially conclud- ing that giving PDs the opportunity to use flexible duty hours did not harm patients. Most secondary outcomes were also noninferior. Comments 1. The results of this noninferiority study suggest that giving program directors discretion to make schedules without con- tinuous duty hour limits did not harm patients or lead to worse outcomes. 2. 30-day mortality and several other patient safety outcomes were not adversely affected, including unadjusted 7-day read- missions, Medicare payments, and AHRQ patient safety indica- tors (Agency for Healthcare Research and Quality); whilst 30 day readmissions missed the noninferiority threshold, this may be due to the generous and conservative standard afforded a 1% threshold margin in the context of a 61% baseline of pro- longed stay in the standard groups. 3. Program directors were permitted but not required to use extended shifts, and the PDs did not use all the latitude offered to them as standard duty hours were maintained for at least some rotations at each institution. 4. This study did not evaluate what happens when trainees work extended shifts. 5. Consideration should be made to re-examine well-intended duty hour shift regulations enabled in 2011. ASSESSMENT VALIDATION: MILESTONES Hauer KE, Vandergrift J, Lipner RS, Holmboe ES, Hood S, McDonald FS. National Internal Medicine Milestone Ratings: Validity Evidence From Longitudinal Three-Year Follow-up. Acad Med. 2018 Aug; 93(8): 1189-1204. Summary This retrospective cohort study assessed 35,217 PGY1- PGY3 residents in US IM categorical residency programs in academic years (AY) 2013, 2014, and 2015 in order to further validate milestones via 3 approaches: determination if “not assessable” items decreased over time, reporting of mean longitudinal milestone ratings for individual IM residents over their 3 years of training, and evaluation of the correlation of medical knowledge (MK) milestones in each PGY with certification exam scores to determine pre- dictive validity of milestones to certification exam scores. The results showed a significant trend toward a decreased percentage of residents with any “not assessable” ratings from AY2013 to AY2015 with drops from 22.5% to 16.6% in PGY1s, 12.3% to 8.1% in PGY2s, and 5.1% to 2.6% in PGY3s. PGY1s starting in AY2013 showed an increase in milestones average from PGY1 average of around 3 (range 2.73-3.19) to around 3.5 in PGY2 (range 3.27-3.66) to around 4 in PGY3 (range 4.00-4.22), which was significant (p trend < 0.001). Both MK1 and MK2 milestones correlated to the 2016 certification exam scores in all 3 PGYs and for each increase of 0.5 units in MK ratings the difference in exam scores for PGY3s was about 19. Comments 1. Prior work confirmed validation of the milestones as post graduate year (PGY) increased and correlated end-of-training medical knowledge milestones with certification exam scores, but also revealed that some competencies were “not assess- able” and this study addresses prior gaps in knowledge.
Made with FlippingBook
RkJQdWJsaXNoZXIy MTM1ODMw