Linda H. Aiken, Sean P. Clarke, Robyn B. Cheung, Douglas M. Sloane, Jeffrey H. Silber, "Educational Levels of Hospital Nurses and Surgical Patient Mortality," Journal of the American Medical Association (JAMA), Vol. 290, September 24, 2003, 1617-1623.

1. Scientific merit: The study has been subjected to the highest degree of scientific review by the National Institutes of Health (NIH) in its funding of the study, and by the reviewers and editorial staff of the Journal of the American Medical Association (JAMA), one of the world's most respected scientific journals in health and medicine. The authors of the paper come from different disciplines-nursing, medicine, sociology-and have a 15-year track record of publishing scientific papers in the top journals in the world on nurses, nursing care, and nursing care outcomes.

2. The research question: Among the 168 hospitals studied, the proportion of nurses with baccalaureate-level (BSN) education ranged from 0 to 77%. With variation in nursing educational composition being so great, health outcomes researchers have a legitimate interest in examining the potential consequences of such variation for patient outcomes. This research team already established in a 2002 paper in JAMA that variation across hospitals in patient-to-nurse ratios was significantly related to risk of dying following common surgeries. Examining the large variation in educational composition of the hospital nurse workforce is a logical extension of a longstanding program of research.

3. Contribution of nurse education to variation in mortality: There are many explanations for hospital deaths, the most important one being how sick the patient is on admission. This study documents that there is an effect of nurses' education on hospital mortality rates above and beyond the other important factors that help explain variation in hospital death rates. The study was designed to take into account how sick the patient was on admission (133 patient variables were used to determine the patient's risk of death), the size of the hospital, the technological capacity of the hospital, whether the hospital is a teaching facility that might bring additional resources to patient care such as resident physicians, whether the patient's attending surgeon is board certified, and patient-to-nurse staffing ratios. After taking into account all of these other factors, a 10% increase in the proportion of hospital staff nurses with baccalaureate degrees is associated with a 5% decline in mortality following common surgical procedures.

4. Nurse workloads and educational mix both important: The study clearly shows that the fewer patients hospital staff nurses care for, the lower the hospital mortality. The educational mix of nurses is significantly associated with mortality after taking into account patient-to-nurse staffing ratios. As noted in the paper, the lowest risk-adjusted mortality following common surgeries is in hospitals in which nurses care for four or fewer patients each and where the proportion of staff nurses with BSN education is 60% or higher.

5. Does the inclusion of master's prepared staff nurses inflate the effect of baccalaureate education? The study question focuses on the variation in educational mix of hospital staff nurses. As noted in the paper, proportions of RNs with different levels of education were analyzed separately--associate degree, diploma, baccalaureate, and baccalaureate plus master's degrees. The particular type of educational credential for nurses without baccalaureate degrees was not a factor in patient outcomes and thus the findings were reported in aggregated form. The average proportion of staff nurses with baccalaureate plus master's degrees was very small, only 1.7% of staff nurses. The effect of baccalaureate education with and without nurses with master's preparation was examined and there were no differences in the results providing no rationale for excluding some staff nurses with baccalaureate degrees from the study. The few master's prepared staff nurses do not account for the baccalaureate education effect. For simplicity, the researchers elected to compare the proportion of staff nurses without and with a baccalaureate degree.

6. Cause and effect: This paper confirms a significant association between educational levels of hospital nurses and variation in mortality. Longitudinal research will be required to establish causality and the direction of the causal path, i.e., whether a higher proportion of less educated nurses "causes" higher mortality or whether hospitals with high mortality cannot attract more educated nurses. Because other possible explanations for the association between education and mortality have already been ruled out by the study design, the direction of causality is in question, but the association is not.

7. Implications: The findings of this study can be helpful to all nurses in obtaining more support for nursing education, facilitating the career advancement of nurses, and in improving nurses' working conditions including moving to safer patient workloads.

September 30, 2003