The methodology used in both studies will be briefly highlighted. Both studies used similar sampling strategies (for hospitals, patients, and nurses), data collection methods (primary nurse survey data, secondary large database sources for hospital and patient data), and analytic methods (regression modeling). 

There are consistent and unique findings across the two studies. The impact of nursing staff mix on 30-day hospital death rates has been consistently demonstrated. In 2002, a 10% increase in the proportion of registered nurses in the staff mix was associated with 5 fewer deaths for every 1000 patients. This impact was larger in the 2006 study – a 10% increase of registered nurses was associated with six fewer deaths for every 1000 patients. More years of nurse experience was found to be related to lower mortality rates in the 2002 study but not in the 2006 study. In the 2006 study, additional predictors of lower hospital mortality rates included higher proportions of baccalaureate prepared nurses and higher nurse-reported use of care maps or protocols to guide patient care.

Hospitals and hospital systems interested in minimizing mortality rates should employ higher proportions of nursing staff that are registered nurses and baccalaureate-prepared. They should also aggressively support the development and use of current care maps to guide patient care.