Association of Hospital Mortality With Initiation of Mechanical Ventilation on a Weekend: A Retrospective Cohort Study

Rationale: Increased mortality in patients admitted to hospitals on weekends is a well-described phenomenon labeled the "weekend effect." Studies evaluating the weekend effect in intensive care units (ICUs) have arrived at conflicting results. Identifying a weekend effect for critically-ill patients may inform clinical care pathways and resource allocation.

Objectives: Determine the association of initiation of mechanical ventilation (MV) upon admission on a weekend versus weekday with hospital mortality.

Methods: We conducted a retrospective cohort study of non-surgical adult patients using the California Patient Discharge Database from 2018 to 2019. We identified MV initiated on the day of admission and diagnoses using discharge billing codes. The primary exposure was admission and initiation of MV on a weekend versus weekday and the primary outcome was hospital mortality. Hierarchical logistic regression was used to determine the association between hospital mortality and MV initiation timing, adjusting for case-mix.

Results: Among 90 288 admissions in 2018 and 2019 meeting inclusion criteria, 24 771 (27.5%) had MV initiated on weekends, while 65 517 (72.6%) had MV initiated on weekdays. Patient demographics and comorbidities were similar between groups. Chronic alcohol and substance use disorders, and acute intoxications and traumas were more prevalent among patients with MV initiated on weekends. No difference in hospital mortality was observed with initiation of MV on weekends versus weekdays (23.1% vs 22.8%, ARD = 0.3%, aOR = 1.02, 95% CI 0.98, 1.07).

Conclusions: Contrary to prior studies, no increased mortality was observed among newly admitted patients initiated on MV on weekends compared to weekdays. While weekend effects may exist in other settings, newly admitted patients likely have MV initiated in the emergency department or ICU, which tend to have more consistent staffing levels. Further research is needed to determine if care patterns in these units could be used as a model for units where weekend effects continue to impact outcomes.