Skip to main content

Process Measures and Quality Improvement

By July 28, 2010Commentary

Maximizing quality in health care is a goal everyone can agree on.  How to get there is less clear.  While quality improvements ultimately should be reflected in better health outcomes for patients, one popular method for measuring quality is use of process-of-care measures–looking at whether a provider took certain steps in treating the patient.  Increasingly, hospitals, physicians and other providers are being compensated or penalized based on use of these measures.  A body of research is evolving to examine whether and the extent to which process-of-care is actually correlated with ultimate health outcomes.  According to research published in the Journal of the American Medical Society the answer in at least one case is that there may be no significant correlation. (JAMA Article)

The study looked at the Surgical Care Improvement Project, which required measurement and reporting of six process-of-care measures in the hopes of reducing surgical infections.   CMS reduces hospital reimbursement by 2% if they don’t report their performance.  One goal of the measures is to report the data publicly to help patients choose higher-quality hospitals.   Using data from about 400 hospitals, the researchers looked for an association between performance on the process-of-care measures and infections.

While the researchers found that performance on one composite measure was correlated with fewer infections, performance on another composite and on individual items was not.   The results of earlier studies showed even weaker correlation.  The outcome suggests that skepticism toward process-of-care measures is warranted, particularly if they are to be used to affect reimbursement or even patients’ choice of providers.  Very clear research evidence should exist that a process-of-care measure is not only correlated with but likely significantly causative of a better health outcome before the measure is used in these ways.  In addition, collecting and reporting this data imposes additional administrative costs which could be avoided if they have no real utility.

Leave a comment