Nurse staffing levels and patient outcomes: a systematic review of longitudinal studies

ABSTRACT Background The contribution of registered nurses (RN) towards safe patient care has been demonstrated by several studies. However, most of the evidence is cross-sectional, hence the inability to demonstrate that staffing levels precede patient outcomes. No reviews have summarised longitudinal studies considering nurse staffing and patient outcomes. Objectives To synthesise longitudinal studies focusing on associations between nurse staffing levels and patient outcomes. Methods Systematic review. We conducted our search in 2020 and updated it in July 2021. We searched Medline, CINAHL, Embase and the Cochrane Library. We used the ROBINS-I tool for assessing risk of bias. We reported results narratively grouped by outcome. Results 28 papers were included. Most studies were either at serious (n=12) or critical (n=6) risk of bias, with 3 studies at low risk of bias. Studies were conducted in a variety of settings and populations. Notwithstanding the limitations, findings are consistent with an overall picture of a beneficial effect from higher RN staffing on preventing patient death. Studies with the greatest risk of bias were judged as most likely to underestimate the effect of higher RN staffing. The evidence is less clear for other patient outcomes, but estimates, though at moderate or serious risk of bias, indicate that higher RN staffing is likely to lead to better patient outcomes. Evidence about the contribution of other nursing staff groups and skill mix of the team is unclear. Conclusion There is a likely causal relationship between low RN staffing and harm to patients, although uncertainties remain regarding the magnitude of effect. To address these uncertainties, future studies should be conducted in more than one hospital and using standardised measures when reporting staffing levels.