Resident duty hours and resident and patient outcomes: Systematic review and meta-analysis

The policies regarding resident physician work hours are constantly being evaluated and changed. However, the results of randomised control trials (RCTs) are mixed. This systematic review of RCTs aims to synthesise the evidence associated with resident duty hour restrictions and its impact on resident- and patient-based outcomes.

A comprehensive search of the Cochrane Library, EMBASE and PubMed was conducted from inception until 31 July 2020. Any RCT evaluating the impact of longer resident physician work hours compared to shorter resident physician work hours on resident- and patient-based outcomes was eligible for inclusion. Two reviewers extracted data independently. The primary outcome was the impact of resident duty hour restrictions on emotional exhaustion, depersonalisation and personal accomplishment, as defined by the Maslach Burnout Inventory. The secondary patient-related outcomes were patient hospital length of stay, serious medical errors and preventable adverse events. Data were pooled using a random-effects model.

Of the 873 references, nine RCTs met the inclusion criteria. A shorter shift length compared with longer shift length was associated with significantly less emotional exhaustion (standardised mean difference [SMD] = -0.11, 95% CI = -0.21, -0.00) and less dissatisfaction with overall well-being (OR = 0.61, 95% CI 0.38, 0.99) but not with hospital length of stay (SMD = -0.01, 95% CI = -0.05, 0.02, p = 0.45) and serious medical errors per 1000 patient hours (OR = 1.07, 95% CI = 0.52, 2.21; p = 0.86).

Shorter resident duty hours is possibly associated with improvement in resident-based outcomes, specifically, emotional exhaustion, dissatisfaction with overall well-being, sleep duration and sleepiness. These findings may inform the policy change in support of reduced shift hours resulting in overall well-being for the residents with possible reduction in burnout without adverse impact on patient-based outcomes.