Quality Reporting and Surgery Outcomes

I don’t know why I take such great delight in seeing programs designed by know-it-all bureaucrats be shown to be completely fallacious.  Maybe it has something to do with that “designed by etc.”  Have a little humility.  The latest source of this perverse delight is a study published in the Journal of the American Medical association examining the effect of a surgery quality reporting program.  (JAMA Article)   The quality reporting program in this case was put together by the American college of Surgeons and focussed on mortality, complications, length of stay and other measures.   These were public reported.  The study used Medicare data on 11 serious surgical procedures and constructed a control group of hospitals that did not participate in this particular program.  It compared the participating and control group hospitals on thirty-day mortality, serious complications, re-operation, and readmissions with 30 days; as well as Medicare costs.  There were 263 participating hospitals and 2 control matched hospitals for each of the participating ones.  A large number of surgeries were performed and available for analysis.  Patient characteristics were generally similar at both sets of hospitals.  The participating hospitals, even after matching, tended to be larger, have more operating rooms and do more surgeries; characteristics which sometimes are said to lead to better quality in themselves.  While both participating and control hospitals showed improvement on the outcome measures, there was no significant difference in that improvement; in other words, participation in the reporting program did not appear to be associated with more improvement in outcomes.  And there was no significant difference in Medicare payments; no cost savings from participation.  Cost savings would be expected if there were fewer complications and readmissions.  Now one possible explanation is that the control hospitals had some other quality improvement efforts going on during the study period, but it certainly does not look like this particular program had any strong effect on outcomes.