The association between ward staffing levels, mortality and hospital readmission in older hospitalised adults, according to presence of cognitive impairment: a retrospective cohort study.

Background: Lower nurse staffing levels are associated with increased hospital mortality. Older patients with cognitive impairments (CI) have higher mortality rates than similar patients without CI and may be additionally vulnerable to low staffing.

Objectives: To explore associations between registered nurse (RN) and nursing assistant (NA) staffing levels, mortality and readmission in older patients admitted to general medical/surgical wards.

Research Design: Retrospective cohort.

Participants: All unscheduled admissions to an English hospital of people aged ≥75 with cognitive screening over 14 months.

Measures: The exposure was defined as deviation in staffing hours from the ward daily mean, averaged across the patient stay. Outcomes were mortality in hospital/within 30 days of discharge and 30-day re-admission. Analyses were stratified by CI.

Results: 12,544 admissions were included. Patients with CI (33.2%) were exposed to similar levels of staffing as those without. An additional 0.5 RN hours per day was associated with 10% reduction in the odds of death overall (odds ratio 0.90 [95% CI 0.84-0.97]): 15% in patients with CI (OR 0.85 [0.74-0.98]) and 7% in patients without (OR 0.93 [0.85-1.02]). An additional 0.5 NA hours per day was associated with a 15% increase in mortality in patients with no impairment. Readmissions decreased by 6% for an additional 0.5 RN hours in patients with CI.

Conclusions: Although exposure to low staffing was similar, the impact on mortality and readmission for patients with CI was greater. Increased mortality with higher NA staffing in patients without CI needs exploration.