Skip to main content

The Hospital Readmissions Non-reduction Program

By January 28, 2019Commentary

A couple of weeks ago we learn that the Medicare hospital readmissions reduction program may have actually increased mortality, not exactly a good ultimate health outcome.  Now we learn, via research in Health Affairs, that the previous estimates of readmission reductions were too high.  (HA Article)  Just to briefly recapitulate the program, there was a perception that there were a lot of unnecessary hospital readmissions within 30 days of a discharge, which was supposedly an indicator of poor quality and had Medicare paying for avoidable care.  The reduction program penalizes hospitals that have readmission rates in excess of “expected” ones for three common conditions.  Hospitals have incurred fairly substantial penalties since the program’s inception.  It has been heavily criticized for a variety of methodological issues, including penalizing hospitals for things beyond their control, like bad patient behavior.  But supposedly the program has been successful at reducing readmissions, albeit at the cost of a few patient lives.  One prior study suggested that the apparent reduction was due to gaming the “expected” number by increasing patient risk scores, kind of like what MA plans do.  In the current study, the authors look at changes in the system used by hospitals to submit claims to Medicare, which more than doubled the number of diagnosis codes hospitals could submit for a patient.  The change occurred after CMS announced the reduction program but before it was implemented.  And guess what, if you had used the more limited code set, you would have only seen about half the reduction that supposedly occurrred, as calculated using the large diagnosis code set, and consequent higher patient risk scores.  Oops.  But wait, there is more, if you correctly constructed control groups for reductions in readmissions by using non-targeted hospitals and non-targeted conditions, adjusting for baseline differences and using consistent risk-scoring methodology, most of the rest of the apparent reduction due to the program disappears.  Double-oops.  So CMS penalized a lot of hospitals for doing nothing wrong in regard to readmissions.  Time for CMS to declare victory and dump the program.  Maybe issue an apology to hospitals too.

Leave a comment