3M Commentary

Bundled Payment Incentives: Risk or Reward?

August 13, 2013 / By Norbert Goldfield, MD, Richard Fuller, MS

In a recent New England Journal of Medicine article, “Bundle with Care — Rethinking Medicare Incentives for Post–Acute Care Service,” Dr. Judith Feder notes that bundled payment systems create incentives for efficiency (which extend to the bundle but not beyond) and, at the same time create incentives to under-provide services [1]. In this blog post, I want to explore the way that incentives to under-provide services have been addressed in the two bundled payment systems currently in use (PPS and Medicare Advantage) and a bundled payment approach that is the subject of a current CMS demonstration (bundling post-acute care with or without inpatient hospital care). I’ll focus on these three bundled payment systems and the features related to incentives for maintaining or improving that are an integral part of each bundled payment approach (if any).

First, there are a wide variety of methods to maintain and improve quality of care. For example, though not a part of a particular payment system, professional standards are always very important in driving provider behavior, and malpractice lawsuits are an ever present threat. We have developed software to identify adverse events which might have been prevented.

PPS was implemented in 1982 (under TEFRA) and in 1983 as a Medicare prospective payment system, creating considerable concern that patients would be discharged “quicker and sicker.” There is no question that PPS provided hospitals with this incentive. However, PPS also contained two features that worked in the opposite direction. First, physicians were deliberately not included under PPS. Thus, physicians had the opposite incentive: to provide more services, as this created more billing opportunities for them. Second, PPS contained an outlier policy that limited hospital losses on very expensive patients, reducing the temptation to avoid these individuals or to discharge too quickly. Subsequent independent evaluations of PPS did not find a systematic pattern of discharging “sicker and quicker.”

Medicare Advantage plans also created quite a stir. Capitated payment systems such as this logically have the most powerful financial incentives to skimp on care because there is a single fixed payment amount for all services. In addition, there is no stop loss for Medicare Advantage plans, and payment for all providers including physicians is included in the capitation amount. While there have been instances where allegations of skimping on services have been made (including a 2002 movie, John Q., starring Denzel Washington), no systematic pattern of wide-spread abuse has been reported. It is one thing to have an incentive, it is quite another to act on it.

Event-based episodes like those proposed by CMS fall somewhere between PPS and Medicare Advantage in terms of the financial strength of the incentives to under-treat. Imagine a payment system like the second or third CMS bundling models, which bundles inpatient acute with all post-acute care, or bundles all post-acute care within a pre-determined window. This episode-based system would, unlike PPS, include physicians. However, it could easily also include an outlier policy like that used for PPS. In addition, the bundle of post-acute services would likely include readmissions. Readmissions are very expensive compared to other post-acute and post-acute services. Thus, it would be penny wise and pound foolish to skimp on services only to see the patient readmitted.

This is not to say that an event-based episode payment system that includes shared profits and losses is not a good idea. In fact, this is one reasonable approach to phasing-in the payment reform. It fits nicely with the approach that Medicare uses in its demonstration; Medicare continues to pay provider bills, adjusting the payments based on the difference between actual and expected total payments.

We recently had a manuscript, “Bundling Post-acute Care Services into MS-DRG Payments,” published by Medicare & Medicaid Research Review (MMRR). The article examines bundling payments into event based episodes. An event-based episode is an inpatient admission plus post-acute care for a pre-defined window such as 30 days post discharge. This method builds on Medicare’s DRG-based prospective payment system. This approach to bundling payment has been very successful. The paper is built on the premise that, if the current DRG system could be expanded to include post-acute care, hospitals would have incentives to better co-ordinate acute and post-acute care, improving quality while reducing cost. Bundling post-acute care services into MS-DRG payments means that the acute diagnoses and procedures which influence acute care costs need to be integrated with the chronic conditions which influence post-acute care costs. This study demonstrates that classification tools which are currently available can easily serve as the basis for an integrated payment system which includes both acute and chronic care. Therefore, this approach to integrated payment reform is a practical alternative for Medicare.

The article may be accessed here. You are free to print copies to share with colleagues or post elsewhere on the web. Dissemination for educational and scientific purposes is permitted without restriction and encouraged.

James C. Vertrees, PhD , is a Senior Economist at 3M Health Information Systems.


1. Feders J. Bundle with Care — Rethinking Medicare Incentives for Post–Acute Care Services. N Engl J Med 2013; 369:400-401. August 1, 2013. DOI: 10.1056/NEJMp1302730