Objectives To examine associations between mortality and registered nurse (RN) staffing in English hospital trusts taking account of medical and healthcare support worker (HCSW) staffing.

Setting Secondary care provided in acute hospital National Health Service (NHS) trusts in England.

Participants Two data sets are examined: Administrative data from 137 NHS acute hospital trusts (staffing measured as beds per staff member). A cross-sectional survey of 2917 registered nurses in a subsample of 31 trusts (measured patients per ward nurse).

Outcome measure Risk-adjusted mortality rates for adult patients (administrative data).

Results For medical admissions, higher mortality was associated with more occupied beds per RN (RR 1.22, 95% CI 1.04 to 1.43, p=0.02) and per doctor (RR 1.10, 95% CI 1.05 to 1.15, p <0.01) employed by the trust whereas, lower HCSW staffing was associated with lower mortality (RR 0.95, 95% CI 0.91 to 1.00, p=0.04). In multivariable models the relationship was statistically significant for doctors (RR 1.08, 95% CI 1.02 to 1.15, p=0.02) and HCSWs (RR 0.93, 95% CI 0.89 to 0.98, p<01) but not RNs (RR 1.14, 95% CI 0.95 to 1.38, p=0.17). Trusts with an average of ≤6 patients per RN in medical wards had a 20% lower mortality rate compared to trusts with >10 patients per nurse (RR 0.80, 95% CI 0.76 to 0.85, p<0.01). The relationship remained significant in the multivariable model (RR 0.89, 95% CI 0.83 to 0.95, p<0.01). Results for surgical wards/admissions followed a similar pattern but with fewer significant results.

Conclusions Ward-based RN staffing is significantly associated with reduced mortality for medical patients. There is little evidence for beneficial associations with HCSW staffing. Higher doctor staffing levels is associated with reduced mortality. The estimated association between RN staffing and mortality changes when medical and HCSW staffing is considered and depending on whether ward or trust wide staffing levels are considered.