When we saw Monday's New York Times report on how Medicare's experiment in disease management was not cutting costs, we asked Robert Berenson MD, Senior Fellow at The Urban Institute, to comment. Here's what he had to say:

Disease management for patients with chronic conditions is supposed to pick off the "low hanging fruit" of cost containment, both in Medicare as well as among self-funded private employers and commercial health insurers. In each of these settings, about 5 percent of subscribers/beneficiaries are responsible for more than 40 percent of the health costs, and 20 percent of the patients run up 80 percent of the costs. Many of these patients have one or more underlying chronic conditions, such as congestive heart failure (CHF) or diabetes.

Surely, the thinking goes, we can easily find savings by applying better management techniques. We can use telephone or web-based communication tools for better surveillance. We can teach patients self-management skills. And we can do all this, the thinking went, without having to interfere with the physician-patient relationship. Indeed, presidential campaigns typically assume substantial savings from disease management-type initiatives targeted to the high spending associated with patients with chronic conditions.

Based largely on anecdotal reports of great success by third-party disease management vendors and effective lobbying, Congress in the Medicare Modernization Act of 2003 commissioned a pilot test of disease management for patients with CHF and diabetes. Unfortunately, as the New York Times April 7 article by Reed Abelson, "Medicare Finds How Hard It Is to Save Money" documents, this three-year pilot failed to save money.

Every once in a while a seasoned (or crotchety) policy analyst gets to say, "I told you so." And it's my turn to say it now. Shortly after this Chronic Care Improvement (CCI) pilot was authorized, I testified before the Ways and Means Subcommittee on Health, warning the lawmakers that business techniques could only go so far in managing disease without the involvement of doctors:

"Although the CCI program may be a good start, in my opinion it is insufficient for truly addressing chronic care needs in Medicare because it lacks a focused physician component. The Administration emphasizes that the new program creates a 'business platform' that will permit innovation, but the CCI program ignores the reality that beneficiaries look to their personal physicians for responsibility for their health care—and not business platforms—whether health plans, disease management companies, or other third-party vendors."

Some elements of the disease management industry say the pilot failed to save money because CMS implemented it poorly. It's more likely that the failure belongs more with the "physician bypass" logic of the approach. Implementation problems notwithstanding, if this approach cannot achieve robust savings on congestive heart failure—the "poster child" of a condition responsible for avoidable, expensive hospitalizations and supposedly amenable to disease management interventions—one has to question whether it can possibly work to reduce spending for patients with assorted other chronic conditions. In fact, some disease management companies now are exploring how to change their approach to become an extension of the physician's office, rather than primarily an external party trying to engage the patient.

For more than 30 years Medicare has tried and failed in many other demonstrations to reduce costs associated with frail elderly and non-frail seniors with multiple chronic conditions—but without the direct involvement of patients' physicians. It is time to try to include, rather than bypass, physicians in chronic care management and coordination. That is where the so-called "patient-centered medical home" comes in. But that is another story.

(New America's Health Policy Program director Len Nichols and Berenson are co-directing a study on Medicare reform, specifically about how to make Medicare a value-based purchaser. They will release their first round of papers this July.)