0995 Schedule Re-design and Patient Safety: the Randomized Order Safety Trial Evaluating Resident-Physician Schedules (ROSTERS). (12th April 2019)

• Record Type: Journal Article
• Authors: Landrigan, Christopher P; Rahman, Shadab A; Sullivan, Jason P; Blackwell, Terri; Vittinghoff, Eric; Barger, Laura K; Sanderson, Amy P; Wright, Kenneth P; O'Brien, Conor S; Poynter, Sue E; Lockley, Steven W; Stone, Katie; Czeisler, Charles A
• Is Part Of: Sleep. Volume 42(2019)Supplement 1
• Page Start: A400
• Page End: A401
• Publication Date: 2019-04-12
• Subjects: Sleep -- Physiological aspects -- Periodicals; Sleep disorders -- Periodicals; Sommeil -- Aspect physiologique -- Périodiques; Sommeil, Troubles du -- Périodiques; Sleep disorders; Sleep -- Physiological aspects; Sleep -- physiological aspects; Sleep Wake Disorders; Psychophysiology; Electronic journals; Periodicals; 616.8498
• DOI: 10.1093/sleep/zsz067.992
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Resident-physicians' extended duration work shifts impair their safety and performance, but the effects on patient safety of eliminating them remain unclear. Methods: We carried out a 6-center cluster-randomized trial comparing rates of serious medical errors (SMEs) when resident-physicians worked on an extended duration work roster (EDWR) that included extended shifts of 24 hours or more, compared with a rapidly cycling work roster (RCWR) that eliminated extended shifts. A well-established, intensive systemic surveillance methodology that included direct continuous observation and chart review was used to measure rates of SMEs. All final classifications of suspected incidents were made by two independent physician reviewers blinded to schedule and site; discrepancies were resolved by consensus (pre-consensus kappa 0.52-0.67). Results: On the RCWR, resident-physicians made significantly more SMEs (97.0 vs. 79.1 per 1000 patient-days [RR 1.53 (95% CI 1.37-1.72)], p<0.0005). Rates of SMEs across study units (including errors involving and not involving resident-physicians) were likewise higher (181.3 vs. 131.5 per 1000 patient-days [RR 1.56 (95% CI 1.43-1.71)], p<0.001). The relative effectiveness of the intervention across sites varied widely. At one site, resident-physician-related SMEs decreased on the RCWR [RR 0.24 (95% CI 0.17-0.34)]; at two sites, there was no significant change; and at three sites, rates of SMEs increased. To explore reasons for the site level variation, we conducted a series of post-hoc analyses. We found that as resident-physician workload increased, the intervention went from being protective to being harmful. Specifically, GEE analyses demonstrated that introduction of the RCWR intervention led to an improvement in patient safety (RR 0.53, p=0.0018) when IPRP was≤ 7; once IPRP increased above 7, however, the intervention led to decrements in patient safety (RR 1.42, p<0.0001). Conclusion: An intervention that eliminated resident-physicians' extended work shifts but simultaneously increased workload increased SMEs. Managing workload is essential as schedules are introduced to reduce resident-physician work hours. Support (If Any): NHLBI U01-HL-111478 and U01-HL-111691