HCI3 Update from the Field: One Simple Lesson

Submitted by francois.debrantes@hci3.org on Monday, April 2, 2012 - 03:25

Newtown, CT – March 30, 2012

There's one simple lesson that keeps eluding many in our field: It's the design, stupid! Poorly designed incentive programs usually garner poor outcomes. For example, when CMS launched the Premier demo a decade ago, they tied marginal benefits to a tournament-style incentive program. The result, as reported in this week's NEJM, is underwhelming. The researchers' conclusion? It's the design, and they caution CMS on its current design of the hospital value-based purchasing model. In particular, Jha and colleagues argue that a hard-nosed focus on readmissions for three conditions might deter from other foci. We'd say that it's not the deterrence that's the problem, but rather the tendency to use these tournament-style programs, which have been mostly discredited. In a webinar a few weeks ago, CMS asked Adams Dudley to summarize the research on P4P to-date. One conclusion is that you should stay away from tournament-style designs. Our research clearly reinforces that point. It's somewhat ludicrous to believe that organizations will bust their humps all year, expending umpteen dollars, to beat a yet-to-be-determined target, for a measly 1% or 2% of Medicare revenue. Bob Berenson and colleagues suggest a better approach, one that we suggested three years ago – impose a warranty on hospital payments. This would make the incentive clear and quantifiable up front, encourage hospitals and physicians to reduce any costs that could count against the warranty (including patient safety failures and readmissions), and create a measurable cap on hospitalizations. This stuff isn't complicated folks and why CMS hasn't applied these lessons is puzzling. We're similarly puzzled by many of the design decisions made for the Bundled Payment for Care Improvement pilot. As it stands, many providers risk incurring a significant amount of insurance risk, which is antithetical to the very purpose of bundled payments, and might well deter many to participate. And the obstinateness with which some are clinging to a bad design is either arrogance or ignorance – neither is a good trait.

What this means to you – With close to a decade of work implementing simple pay for performance programs, such as our flagship Bridges To Excellence, and more complex incentive programs, such as PROMETHEUS Payment, we know and have shown unambiguously that incentive programs work well when well designed. And much research has shown they work poorly when poorly designed. Oftentimes expediency is the culprit. In the rush to do something, it's done badly. Frequently, however, it's due to timidity. Low rewards amounts are safe, even though we know they're ineffective. If you want someone to focus on a specific goal or set of goals, you create an incentive that will get them focused on January 1st, not December 1st. So ask yourself, and those you're trying to motivate, what will get them focused on January 1st. While these errors of design are regrettable, they're understandable. What's inexcusable, however, is a poor design caused by the arrogance of the designers. And even though we can take comfort in knowing they will eat their humble pie, there's no comfort in knowing that the collateral damage might be significant and set us all back. It's the design, stupid!


Francois de Brantes
Executive Director
Health Care Incentives Improvement Institute, Inc.
w: www.hci3.org