Variability in Readmission Rates

Submitted by hci3-usr on Friday, October 7, 2016 - 12:42

Newtown, CT- October 7, 2016

What’s the difference between a provider-centric measure and a patient-centric measure? A recent study in Health Affairs clarifies it In the study, David Chin and colleagues argue correctly that the majority of the variability in readmission rates post-discharge, at least for condition-related hospitalizations, occurs within a week of discharge. Beyond that, the picture becomes a little more fuzzy and, in the authors’ words, “most readmissions after the seventh day post-discharge were explained by community- and household-level factors beyond hospitals’ control.” Importantly, the authors also acknowledge that this seven-day cutoff may not be ideal in comparing facilities for readmissions following procedural episodes. Four years ago we published an Issue Brief in which we looked at the frequency and costs of readmissions caused by potentially avoidable complications for patients undergoing total knee replacement in both Medicare and commercial data sets. We found that while there was a significant drop in readmissions after the first week, that trend reversed itself for commercial patients in the fifth week. We also found that the cumulative number of readmissions after the first week and up to the 180 days was significantly higher than in the first week and that the overwhelming majority of those readmissions were linked to the procedure. Of course, those readmissions weren’t necessarily related to what happened during the hospitalization, and therein lies the difference between provider-centric and patient-centric measures.

What this means to you – As a patient, what matters is what happens during my episode of care, whether that’s an on-going condition, an injury or an illness. And what I want to know is whether the care I’m going to get will help me get better. From that viewpoint, any yardstick should look at complications that occur during the entire episode period, and that’s effectively what employers and other purchasers are doing when they contract for an episode of care. They are implicitly creating a warranty period, and encouraging the physicians, hospitals, physical therapists, visiting nurses and all other care providers to work together, as a team, to achieve better outcomes. Measuring that team’s performance, whether it’s a formal or informal team, should be done by looking at the rate of avoidable complications during the entire episode time window. In other words, what matters is a patient-centric measure. However, if I’m a physician trying to understand which hospital will be a better environment for my patient, or if I’m a payer or purchaser trying to figure out which hospital I should include in a preferred network for a specific surgery or the management of a condition, a provider-centric measure is the way to go, and David Chin and colleagues rightfully point out that a shorter time window of measurement can better capture the effects of the provider’s zone of influence. Both views are important and both fulfill an important function, and we should never confuse one for the other, and certainly not let those that will abuse the findings of the Chin study to compress the time frame for all measures. And for us, the patient-centric view will always be the default because what matters is not whether the measurement of the hospital’s performance is accurate, but whether the patient’s outcomes for the episode of care were the best they can be.


Francois Sig