Early effects of resident work-hour restrictions on patient safety: a systematic review and plea for improved studies.

BACKGROUND: since the inception of the eighty-hour work week, work hour restrictions have incited considerable debate. Work hour policies were designed to prevent medical errors and to reduce patient morbidity and mortality. It is unclear whether work hour restrictions have been helpful in medicine in general and in orthopaedic surgery specifically. This systematic review of the literature was designed to determine the success of these restrictions in terms of patient mortality, medical errors, and complications.
METHODS: a systematic review of the literature was performed to determine if work hour rules have improved patient and systems-based outcomes and reduced physician errors as measured by mortality, medical errors, and complications. A random effects model was utilized to determine whether patient mortality rates were improved under the new rules.
RESULTS: the odds of patient death before implementation of the work hour rules were 1.12 (95% confidence interval, 1.07 to 1.17) times those after implementation. These differences were consistent across disciplines. The data concerning medical or surgical complications before and after the institution of the work hour rules were mixed. There was little information in these studies concerning direct medical errors. The odds of death in nonteaching cohorts were not significantly different from that in teaching cohorts.
CONCLUSIONS: there appears to be a decrease in mortality following the institution of work hour rules. The difference seen in teaching cohorts is not significantly different from that in nonteaching cohorts. It is unclear whether this difference would have been observed even without work hour restrictions. No study has shown a reduction in mortality for orthopaedic patients in teaching cohorts that was greater than that observed in nonteaching cohorts. Because of methodological concerns and the lack of current literature linking physician fatigue and physician underperformance with patient mortality, it is unclear whether the goals of the work hour reductions have been achieved. Furthermore, because of a lack of a so-called dose-response relationship between work hour reduction and patient mortality, it is uncertain whether further reductions would be beneficial. LEVEL OF EVIDENCE: therapeutic Level III. See Instructions to Authors for a complete description of levels of evidence.