“It is written: Man shall not live by bread alone.”
No matter what you think of the source of that quote, the idea that there may be limits to “aligning incentives” has some merit. In healthcare settings, physicians seem to be supportive of being fairly compensated for their work, but also seem to be quite skeptical about the use of “carrot and stick” style economic rewards to influence clinical practice.
Case in point is this interesting paper describing the results of a randomized clinical trial that used blood cholesterol-level control to assess the relative merits of a) rewarding just the patients vs. b) rewarding just the doctors vs. c) rewarding both patients and doctors vs. d) usual practice, or a control group.
The study took place in three marquee institutions, involving 340 primary care physicians who were already taking care of 1503 adult patients with 1) elevated cholesterol levels who 2) either had coronary artery disease or were at high risk for coronary artery disease.
About half of the patients were already on cholesterol-lowering pills.
The purpose of the study was to determine if real money could be used to increase the rate and level of prescribing a statin drug aimed at achieving levels of cholesterol control that were consistent with national guidelines.
The 358 patients in the first group (a above) were cared for by 58 physicians; these patients rewarded by participation in a daily lottery system that gave $10 or $100 if the right number was hit and a wireless-enabled pill bottle indicated that it had been opened.
The 64 docs in the second group (b above) caring for 433 patients with no incentives were rewarded with $256 for each patient per quarter who successfully lowered their cholesterol to target levels.
In the third group (c above) 346 patients got $5 or $50 if they hit the lottery, while their 58 physicians got $512 per patient at target.
A fourth comparison group of patients and physicians served as the control group with no economic incentives.
All patients received their statin drugs in a radio-enabled pill bottle that signaled each time the container was opened.. This allowed researchers to track medical usage.
12 months later, compared to the control group, the only patients that lowered their cholesterol in a statistically significant manner were the ones in the third “shared incentives” group. What’s more, while the drop was greater than would expected through chance alone, the absolute change was relatively small and wouldn’t be expected to result in a big change in the likelihood of a future heart attack. Last but not least, while the shared incentives group opened their pill bottles more frequently, the average level of medication compliance for all groups was less than 50%.
The authors correctly point out that the usual care control group of patients (N=366) being cared for by their control physicians (N=58) were exposed to the wireless-enabled pill bottles and that the lowering of their cholesterol levels made the three intervention groups look bad by comparison.
Your correspondent’s take?
While notions of “pay for performance,” “value, not volume,” and “skin in the game” are attractive notions to policymakers and health leaders, their top-down impact at the one-on-one doctor-patient level defies linear economic logic. The PHB suspects that the physicians caring for these patients had already talked to their patients about starting or increasing the cholesterol medicines and that that quality care had already occurred independent of any fancy monetary incentives. In other words, they were already doing their best.
On an unrelated note, simply monitoring medication compliance with the radio-controlled pill bottle seemed to have an outsized impact on the study. This author wonders if that can’t be used to help patients who are already trying to do their best.
This study should give pause to anyone who thinks that physicians can be manipulated with more money. They live by more than bread alone.