Conceptual Framework: A nursing outcomes effectiveness framework guided this study.  Specifically, hospitalized patient outcomes are influenced by patient characteristics, their clinical condition, the treatments they receive, and the context of care delivery.

Subjects: Inclusion criteria were ³ 60 years of age, hospital admission during study period (July 1, 1998 to June 31, 2002), and at risk for falling or received the nursing treatment of Fall Prevention.  The majority were Caucasian (93.5%), female (52.6%), and married (53.2%) with mean age of 73.7 years (8.5). 

Methods: A nursing outcomes effectiveness database was developed from 9 data repositories from a Midwest tertiary care hospital. 10,187 hospitalizations were extracted and analyzed. Propensity scoring and generalized estimating equations (GEE) were used to determine the unique contributions of nurse to patient ratios (e.g. 1 to 3; 1 to 6) on identified outcomes. Following calculation of the propensity score, other nursing factors (e.g. skillmix) were included in the final model.    

Results: The odds of a complication increased by 38% when the nurse to patient ratio was greater than 1 (nurse) to 3 (patients) compared to less than 1 to 3. Number of units resided on during hospitalization, RN skill mix, and a drop in the average RN staffing on a unit all resulted in increased odds of having a complication during hospitalization.  Nursing factors that consistently influenced study outcomes across all statistical models were RN skill mix; RN staffing below the unit average, and number of units patients resided on during the hospitalization. 

Conclusions: Staffing levels and skill mix contribute significantly to patient safety.