One of the strongest trends in health care is the measurement of provider quality, used both to inform payers and consumers about potential factors to be considered when selecting a provider and increasingly, for pay-for-performance programs. It is obviously critical that the method of collecting and analyzing this data provides accurate and reliable information. The state of Massachusetts requested an analysis of hospital mortality rate calculation methods. This statistic is a very commonly used metric to describe hospital quality. The results of this study were published in the New England Journal of Medicine. (NEJM Website) (Article not online yet, published in 12/23/10 issue)
The researchers examined these calculations as performed by four commercial vendors on hospital mortality in Massachusetts. The vendors presumably used different methods to calculate the mortality measure. Some of the vendors calculated much higher mortality rates than others. The same hospital could have a substantially different rating resulting from the application of different methods. For example, twelve of twenty-eight hospitals had higher than expected hospital-wide mortality by one method but lower than expected mortality as calculated by another vendor. The same result would likely occur if the methods were applied nationally.
The significance of this study is obvious. Great caution needs to be exercised in viewing current published quality measures and a critical view needs to be taken in understanding how they were calculated. At some point, a very standard method needs to be agreed upon and mandated for use and regular audits performed to ensure that what is being reported reflects reality. If we expect payers and consumers to make decisions based upon quality measures, those measures have to be airtight.