A paper entitled "An increase in the number of nurses with baccalaureate degrees is linked to lower rates of postsurgery mortality" appeared in Health Affairs in March and  generated quite a lot of buzz on the Internet.

Its major finding was that hospitals in Pennsylvania that had 10% more nurses with BSN degrees were found to have 2.12 fewer deaths per 1000 postop patients than those that did not. The authors extrapolated this, saying that if all the hospitals they surveyed had the same percentage of BSN nurses as the best performers, 500 deaths may have been avoided.

The reduction in mortality rates was not significantly affected by staffing levels, skill mix or years of experience as a nurse.

The mechanism for the decrease in death rates was not explained but assumed to be better rates of "rescue" after the development of complications which also was significantly associated with the presence of more BSN nurses and not staffing levels, skill mix or years of experience as a nurse.

The abstract concludes, "The findings provide support for efforts to increase the production and employment of baccalaureate nurses."

The math in the paper is confusing. In 2006, 25,000 nurses responded and 1/3 (presumably about 8,333) of them were staff nurses in general hospitals. Those the information from those 8,333 nurses was the basis of the study. They go on to say that there were 134 hospitals with an average of 48 respondents. That computes to 6,432. That's a discrepancy of over 2000.

A cliché that is often used in comments about research papers is "the study raises more questions than it answers."

Here are a few.

Continue reading...