Scott MacStravic, PhD
When I read the interim report on the results so far of Medicare’s Health Support demonstration project, evaluating the impact of chronic disease management programs offered by vendors, providers and insurers, I found the same “mixed results” that have characterized previous scientific studies of health management by government insurers. While some of the projects delivered clearly positive results, there was a lack of “statistical significance” to most, due to small numbers of participants. Indeed, recruiting and retaining enough participants seemed to be a major problem for most project participants.
The report was lengthy, as you’d expect with a government study document, and about as exciting as you’d expect it to be. But included among the hordes of statistics was a table that provided what may be an explanation of the limited success achieved so far. The average fees for the fifteen project participants, covering a range of different and sometimes multiple chronic conditions being managed, was roughly $220 – per month! Moreover, the selection of patients who participated in the projects was sometimes such that their personal medical costs per year were not as great as the fees charged. [“The Evaluation of the Medicare Coordinated Care Demonstration: Findings for the First Two Years” Mathematica Policy Research, Inc.(www.mathematica-mpr.com/publications/pdfs/mccdfirsttwoyears.pdf)]
While I claim no great financial expertise, I still vividly recall Macawber’s discussion of the immutable low of economics in Dickens’ David Copperfield. If money out is more than money in, the result is misery and disaster. And if the fees on average equal roughly $2600 a year, and as much as over $5000, the projects started with a serious handicap. There certainly have been cases where a well-managed patient with a chronic condition generated more than $5000 in savings in a given year compared to patients whose conditions were out of control, these are more the exception than the rule.
This suggests two things that may be particularly difficult in government-sponsored studies. First is that providers of health management initiatives should carefully select whom they choose to manage, to ensure that there is room for net savings, given the prices they intend to charge. This may be difficult if payers insist on random selection for intervention and control groups in order to achieve scientific validity.
Second is that perhaps prices should be based on a “leaner” approach to health management so that prices can be set at a level where positive ROI is at least mathematically possible. HCOs everywhere are being forced by payment stinginess to become familiar with lean management of sickness care. If random selection in health management requires managing the health of low-cost as well as high-cost patients, then the ways by which health is managed should be varied for individual patients, on the basis of risk/reward potential.
There are widespread examples of “customization” of health management interventions based on risk/reward potential. The Duke University Health’s “Duke Prospective Health Program”, for example, offers all participants a set of basic support services, while reserving more intensive and expensive services such as case management, personal coaching and risk-specific programs like smoking cessation, to those who represent a real potential for positive savings. (www.dukeprospectivehealth.org)
HealthMedia, Inc. offers automated computer customization based on Health Risk Assessments that virtually ensure that every participant gets feedback tailored to the individual. Its proactive health interventions are widely used by insurers, including Kaiser Permanente, as well as over 200,000 employees whose productivity as well as health can be improved, thereby increasing the savings to employer clients. [“Revolutionizing Behavior Change: Achieving and Measuring Productivity Improvements” HealthMedia.com Oct 11, 2006]
The fact that thousands of employers are continuously and increasingly investing in employee health management, while most insurers invest in member health management may indicate any mix of misguided or successful investments in the non-government sector. Results by local governments, such as the dramatic savings achieved by Asheville, NC in reducing costs associated with patients diagnosed with diabetes, suggests that success is not impossible. [J. Mahoney & D. Hom Total Value/Total Return GlaxoSmithKline 2006] Once investors get the costs closer to the lean management level, we may see some positive results reported even by the federal government, to the relief of lots of taxpayers.
