Under Medicare's newest pay-for-performance program, hospitals receive rewards or penalties for meeting process performance and patient-satisfaction goals. The program, better-known as value-based purchasing, or VBP, assumes correctly that patients will fare better if hospitals improve the processes by which they deliver care.
A lot of research went into developing the VBP plan. Numerous studies showed a strong relationship between using best practices and better outcomes. What was much less clear in the medical literature was whether improving a patient's perceptions about the quality of care would have a similar impact. Recent research is divided on the question, an issue raised at last month's meeting of the Medicare Payment Advisory Commission
An analysis by Modern Healthcare
would seem to bear out the contention that something is amiss with VBP, at least as it is currently structured. Our reporters compared its rewards and penalties to the other new Medicare performance program initiated under the Patient Protection and Affordable Care Act: penalties for readmissions.
Under the VBP program, hospitals were either rewarded or penalized up to 1% of their total Medicare reimbursement based on a suite of 13 process measures, including the government-approved hospital survey of patient satisfaction. The reward system was weighted 70% for process measures and 30% for patient satisfaction.
The second program, whose first-year results were announced in November, penalized hospitals up to 1% of reimbursement for excessive rates of readmission, a hard outcomes measure.
Our comparison found there was only a weak correlation between the VBP rewards/penalties and readmission penalties. In fact, more than 41% of the hospitals that scored best on readmissions—they received no penalty—were penalized for their performance on processes and patient satisfaction. On the other side of the spectrum, more than 42% of the hospitals that received a maximum 1% penalty on readmissions were rewarded for their process and patient-satisfaction performance.
There are numerous possible explanations for the divergence between process performance and outcomes at nearly half of all hospitals. Mortality and overall health, which have yet to be included in a Medicare rewards program, may be more important outcome measures than readmissions. The process measures used by Medicare, gleaned from reimbursement records, may not be the most important in driving outcomes. Timing may also be a factor—how much can in-hospital processes really determine readmission rates; it may be that overall system performance needs to be rewarded or penalized, not just the hospitals.
The weak link between patient satisfaction and final outcomes also needs to be taken into account. Are questions about the cleanliness or noise levels in hospitals really relevant? Does a zero on “How often did nurses explain things in a way you could understand?” translate into worse outcomes? Writing late last month in the New England Journal of Medicine, defenders of patient-satisfaction surveys—some of whom are financially involved in developing their own survey methods—vigorously defended patient experience surveys “when designed and administered appropriately.” Going farther, some analysts have criticized Medicare's rewards and penalties for being too small to influence provider behavior. The implication is that the government should move quickly to raise the stakes.
That would be a mistake. Medicare and other payers using pay-for-performance and other incentive programs must be sure they are rewarding and penalizing the right processes—ones that truly improve outcomes. And if hospitals and providers are going to be asked to throw patient satisfaction into the mix, the questions asked must not confuse feeling good with getting better.
Merrill Goozner, Editor