Beyond staffing ratios - team demographic diversity, workforce stability, and their effects on patient safety and...

Nursing team research has focussed primarily on staffing ratios and skill mix; however, even when staffing levels are comparable, patient and staff outcomes can vary substantially. Team composition factors beyond professional roles may fundamentally shape performance through demographic diversity and team stability mechanisms identified in organisational behaviour theory.

To examine whether ward-level nursing team demographic diversity, stability, and educational composition are associated with patient and team outcomes across multiple healthcare systems.

We conducted a cross-sectional observational study across 48 adult medical and surgical nursing units in six hospitals across Belgium, Germany, Luxembourg, and the Netherlands from October 2022 to September 2023. Team composition variables included demographic diversity (age, tenure, employment rate), team stability (workforce retention) metrics, and educational composition. Patient outcomes were falls with injury and mortality rates; staff outcomes included turnover, overtime hours, and absenteeism. We employed a two-stage analytical approach: Elastic Net regularisation for variable selection followed by Bayesian ridge regression for effect estimation with comprehensive uncertainty quantification.

Teams demonstrated substantial demographic diversity with high age heterogeneity (mean Blau index = 0.92; SD = 0.04) and variable stability (60-100% retention rates). Age dispersion within teams showed the strongest association with patient safety, with one standard deviation increase in age dispersion associated with a 0.50 standard deviation reduction in fall rates (95% equal-tails credible interval (CrI): -0.74 to -0.26). Within-team variation in work percentages (employment rate dispersion) was associated with higher fall rates (0.31; 95% CrI: 0.08 to 0.55). Team stability was associated with lower absenteeism, but also with increased overtime hours by 0.34 standard deviations (95% CrI: 0.05 to 0.65). Higher team educational levels were associated with reduced fall-rate (-0.28; 95% CrI: -0.50 to -0.06) and lower overtime (-0.34; 95% CrI: -0.62 to -0.04). All mortality predictors were eliminated during the variable selection process, indicating insufficient power, or absence of any effects.

Greater age dispersion and higher educational levels within nursing teams were associated with better patient safety outcomes. Employment rate dispersion (differences in full-time versus part-time employment within teams) showed associations with poorer outcomes. Team stability presents a trade-off, reducing absenteeism but increasing overtime. Given the observational design and exploratory nature of this multi-country analysis, these findings generate hypotheses about team composition effects that warrant further investigation through longitudinal and interventional research designs.