AHRQ on Readmissions

Regular readers know how we feel about the CMS hospital readmissions penalty program:  that it is poorly designed and can reward or penalize hospitals arbitrarily, without regard to the actual quality of care they are providing.  That doesn’t mean that monitoring readmissions isn’t important or that they can’t be an indicator of poor quality.   An Agency for Healthcare Research and Quality Statistical Brief gives some basic data on the nature of hospital readmissions.  The research is based on hospital data from 15 states for the year 2008.   (AHRQ Brief)    The researchers examined all hospitalizations and readmissions and divided them into those for surgical and non-surgical reasons and for patients with chronic or acute conditions.  They also stratified the results by age and by payer.  Overall, for surgical hospitalizations the readmission rate was 6.1% to 19.1% depending on payer and age group, with an overall average of 12.5% for acute conditions and 12.6% for chronic ones.  For nonsurgical admissions, the rate ranged from 8.6% to 28.4%, with the overall average being 18% for acute conditions and 22.7% for chronic ones.  Non-surgical hospitalizations likely have higher readmission rates because these are patients with diagnoses like heart failure, who generally are sicker and need more care overall.  Readmission rates may be higher for persons with chronic conditions for the same reason, although that appears to only be true for nonsurgical admissions.

In general persons covered by Medicaid have higher readmission rates than people covered by private insurance and, for older adults who are not over 65, those who have Medicaid coverage have higher rates than people covered by Medicare.  Both these findings are likely the result of the link that appears to exist between poverty and poorer health status.  For hospitals seeking guidance on where readmissions are likely to occur; it is obviously more likely among older patients, among those covered by Medicaid, among those with a hospitalization related to a chronic condition and a non-surgical hospitalization.  What the data doesn’t reveal, at least in this brief, is geographic variation in rates or factors, or whether the readmission was clinically justified and not the result of poor prior care.  Those are the questions that should be addressed in any program to rationalize readmissions and deter ones that could have been avoided by better quality of care.