Beyond staffing ratios: a multi-country observational study linking nursing team composition with patient and staff outcomes.

This dataset accompanies the study "Beyond staffing ratios — team demographic diversity, workforce stability, and their effects on patient safety and workforce sustainability: A multi-country analysis."

It contains ward-level observations on 48 adult medical and surgical nursing teams from six acute-care hospitals across Belgium, Germany, Luxembourg, and the Netherlands, collected prospectively from 1 October 2022 to 30 September 2023. Each row represents one ward. Columns include team identifiers and structure (team_id, org_id, strata_id, total_size, beds, mean_admissions, mean_length_of_stay); demographic composition measures for age, experience, and employment rate (means, standard deviations, normalised Blau indices, polarisation indices); educational composition (EQF-weighted team and nursing educational levels, bachelor proportion); team stability (Team Stability Index, Team Size Stability Index, Jaccard membership similarity); workforce outcomes (turnover, absenteeism, overtime); and patient-safety outcomes (falls, mortality, CLABSI, HAPI, error rates with z-scores). Team composition predictors were derived from HR records at two measurement points (October 2022 and March 2023) and averaged. Patient outcomes were aggregated annually; workforce outcomes cover the final six months (April–September 2023). All continuous predictors are standardised (mean = 0, SD = 1); outcomes underwent Yeo–Johnson transformation and organisation-specific z-normalisation.

Drawing on the Categorisation-Elaboration Model and organisational familiarity theory, we tested whether demographic diversity, educational heterogeneity, and team stability are associated with nurse-sensitive patient outcomes (falls, mortality) and workforce outcomes (turnover, absenteeism, overtime), using Elastic Net variable selection followed by Bayesian ridge regression. Key findings: greater within-team age variability was protective against falls (β = −0.49 SD; 95% CrI [−0.73, −0.26]); employment rate dispersion and mean employment rate increased falls (β ≈ +0.30 SD each); higher educational level reduced falls and overtime; team stability reduced absenteeism but increased overtime; educational diversity was associated with higher turnover. No predictors survived selection for mortality. Findings are exploratory and cross-sectional; important confounders (temporary staff use, patient acuity) could not be controlled. Users should note that predictors are pre-standardised and outcomes pre-transformed.