Abstract
Objective(s): To investigate the association between the monthly turnover rate of hospital nurses and senior doctors and health outcomes (mortality, unplanned readmissions) for patients admitted to hospital. Design: A retrospective longitudinal panel-data regression analysis, using nine years of monthly observations from rich administrative datasets at worker and patient levels. Associations using linear and unconditional quantile regressions were estimated, including controls for seasonality and hospital organisation. Participants: Yearly records on 236,000 nurses; 41,800 senior doctors (SAS doctors and hospital Consultants); and 8.1 million patients admitted to hospitals. Main outcome measures: Four hospital quality indicators (risk-adjusted by patient age, sex and Charlson index comorbidities) are used and measured at a monthly frequency on a % scale: mortality risk, in or outside the hospital, within 30 days from all-cause/emergency/elective admission to hospital; unplanned hospital emergency readmission risk within 30 days from discharge after elective hospital treatment. Results: A one standard deviation increase in nurse turnover rate is associated with 0.035 (95% CI: 0.024 to 0.045) and a 0.052 (95% CI: 0.037 to 0.067) percentage point increases respectively in all-cause and emergency admissions 30-day mortality risks. A one standard deviation increase in senior doctor turnover rate is associated with 0.014 (0.005 to 0.024) and 0.019 (0.006 to 0.033) percentage point increases respectively in all-cause and emergency admissions 30-day mortality risks. Higher nurse turnover rate is associated with higher 30-day mortality risk in surgery (p<0.01) and general medicine (p<0.01) specialties, as well as mortality for patients admitted to hospitals related to ICD-10 Chapters I (infectious and parasitic diseases; p<0.05) and XIX (injury, poisoning and consequences of external causes; p<0.01). Higher senior doctor turnover rates are associated with higher 30-day mortality risk for patients admitted to hospitals related to ICD-10 Chapters I (infectious and parasitic diseases; p<0.05), V (mental and behavioural disorders; p<0.05) and X (respiratory system; p<0.05). Hospital nurse and senior doctor turnover rates are not significantly associated with risk-adjusted hospital mortality and unplanned emergency readmissions for elective patients.