Diabetes Disease Management Saves Money in a Medicare Advantage Plan

In a prior post, the Disease Management Care Blog was set astir by a scientific meeting abstract showing that disease management saved money. Meeting abstracts typically make their way to print, and the DMCB is pleased to report that's what happened here. Authored by James Rosenzweig, Michael Taitel, Gordon Norman, Tim Moore, Wendy Turenne and Pei Tang, the manuscript is assertively titled "Diabetes Disease Management in Medicare Advantage Reduces Hospitalizations and Cost." It's available on-line at The American Journal of Managed Care.

The DMCB donned it scrutiny spectacles to see if its initial zeal was warranted. It was.

The setting for the study was Medicare Advantage (MA) "in one region of a large national health plan." To be eligible for disease management, the MA enrollees had to meet four criteria between May of 2005 and April 2006: 1) have insurance claims for diabetes and 2) have insurance claims for coronary artery disease, and 3) have an urgent care or emergency department or hospital visit for a diabetes related condition and 4) not have another highly significant condition like HIV, be in hospice or be on kidney dialysis.

526 persons met criteria and "were selected by the plan to be considered for the study." They were then "unequally randomized" (so that a majority could be placed in disease management). 462 were offered and 356 consented to be in disease management (106 didn't), while 64 persons were assigned to the control group. The average age in both groups was around 74 years, about 43% were female and the groups didn't significantly differ in baseline measures of health care costs.

The disease management was a standard mix of a baseline telephonic clinical assessments, nurse phone calls every 2-4 weeks, reminders, mailed educational materials/newsletters, home biometric monitoring and routine/as-needed-urgent physician reports. Based on some of the authors' affiliations, it looks like the company responsible for delivery of the disease management intervention was Alere, which, in turn, used patient education materials from the Joslin Diabetes Center.

In addition to being prospective and randomized, this was also an intention to treat study with an analysis that included utilization data from the 106 patients assigned to disease management but didn't give consent to participate. Compared to the baseline year (May '05 to April '06), the costs of the group assigned to disease management declined by $984 per member per year (PMPY), while in the control group's costs increased by $4547 PMPY. The difference was statistically significant. The savings were accompanied by reductions in diabetes as well as all-cause admissions and emergency room visits. While comparison to controls was not possible for the clinical measures, there was an across-the-board and statistically significant pre-post improvement for the disease management group in foot exams, A1c testing, LDL testing, microalbumin tests, retinal exams, use of ACE/ARBs and use of aspirin.

The DMCB thinks this is a compelling study. The methodology is solid, involving randomization in a prospective study of two parallel groups that were otherwise similar using an intention to treat analysis. It's hard to argue with it's conclusion that traditional disease management saved money. Given the extent of the reduction in costs, it's likely that the savings exceeded the costs of the program and/or the fees that were charged to the insurer. Kudos to Joslin, Alere and the unnamed MA plan for going the extra step and moving these data into the public domain.

While "greedy insurers" and their allegedly overpaid MA plans are a favorite piñata for the Obama Administration, it appears that doing the right thing is possible. As discussed in this CBO report, many MA plans offer disease management programs, but their impact was not known. Thanks to this report, we know more.

As the DMCB has noted before, "disease management" has evolved since it's earliest days. Modern "remote" or telephonic interventions are making more effective use of behavior change theory and the industry is far more sophisticated in targeting patients with the greatest likelihood of benefit. While that's not explicitly spelled out by the authors, the DMCB thinks both factors played a big role in this study's success.

The DMCB will close with this link to Don Berwick's first public appearance as CMS Administrator (it starts at the 9 minute mark and ends about 6 minutes later). The DMCB has seen this dismal phenomenon before: energetic, edgy, bright and sometimes confrontational physicians being elevated to positions of organizational authority, such as Chief Resident, Section Chief, Chairman or Vice President and then sadly becoming a mouthpiece for The Man. The DMCB forgives Don this time for parroting his bosses' unquestioning faith in electronic health records as a panacea for all that ails healthcare. With time, however, it hopes we see the familiar, skeptical and pushy Don that believes in outcomes, not vacuous political nostrums. He can start by looking at this AMJC paper by Rosenzweig et al and re-examine the role of the newer, smarter and more sophisticated versions of disease management as one option to increase quality and lower costs among the Medicare's enrollees.