Recent events have cast doubt on the ability of disease management to reduce medical cost and to affect the quality of care.
The scenario is familiar. The marketplace comes up with an innovation that promises to fundamentally change healthcare and bring about better quality of care at lower cost. Purchasers enthusiastically embrace the new strategy, and the industry that developed the innovation does very well financially—for awhile. Then, doubts begin to set in. The promised cost savings seem, at best, transient. Quality of care is not improving. And medical providers together with their patients begin to argue that care decisions should not be influenced by third parties driven by profit motives. I am, of course, talking about the rise and fall of the HMO in the 1990s.
Do recent events suggest that population-based disease management could suffer a similar fate? Lately, the news hasn’t been very encouraging, particularly after the Centers for Medicare & Medicaid Services announced that participants in the Medicare Health Support demonstration failed to meet the statutory requirements, and, as a result, the pilot program would terminate this year. An important factor contributing to this decision was the inability of the participating vendors to achieve sufficient cost savings to cover their fees. This result had been anticipated by several earlier reviews of the evidence, including one by the Congressional Budget Office1 and one by me and my co-authors in a recent issue of The American Journal of Managed Care,2 both of which found no conclusive evidence that disease management reduces healthcare cost.
But can disease management provide important benefits other than cost savings? Our review yielded substantial evidence that disease management improves quality of care and disease control. This finding would suggest that disease management has beneficial effects even if it does not save money; therefore, it is a good investment. This argument recently was voiced by prominent industry representatives, who state that value for money is a better evaluation criterion than return on investment.3,4
In this issue of The American Journal of Managed Care, Chan and Cooke examine the issue of the value of disease management by investigating the impact of a disease management program on pharmacologic treatment after myocardial infarction.5 They studied the program’s effect on patients’ utilization of recommended drugs, time to fill the first prescription after discharge, and adherence to the drug therapies. Based on a case–control analysis, they found no significant differences in utilization of and adherence to evidence-based drug regimens. In fact, the comparison group did slightly better on most measures than the intervention group.
Does this result allow us to conclude that the disease management program provided no value? Although the findings of this study might imply as much, it is possible that these results are due to problems in the study design rather than in the disease management program. In this study, the authors took advantage of a natural experiment that resulted from employers having to purchase disease management services as an add-on to the regular health plan coverage that they offered to their employees and their dependents. Employees whose employers opted to purchase the program constituted the intervention group, and employees whose employers declined the option served as the comparison group. Although this is a plausible approach and the 2 groups looked fairly equivalent at baseline, unobservable differences between the groups could have biased the results. Observable differences also may have caused bias, because the authors did not adjust the results for differences in available patient characteristics to improve the comparability of the intervention and comparison groups.
Further, the authors provided no information on preintervention levels of quality of care. Thus, it is possible that the intervention and comparison groups had different baseline or preintervention levels of quality of care. In that case, the similar postintervention levels would disguise differential changes in quality. Lastly, the study suffers from power limitations, as the small sample sizes would have been sufficient only to detect differences of a magnitude of 10-15 percentage points in utilization and adherence rates. So it is conceivable that the study’s limitations prevented it from detecting a smaller, but nonetheless significant, beneficial effect of the intervention.
The alternative explanation, however, is that the program indeed had no measurable effect on quality of care, and the study accurately reflected that. It certainly cannot be taken for granted that every disease management intervention will result in better quality of care, so we need to take seriously those studies that suggest the absence of an intervention effect. At this point, most of the evidence comes from evaluations of small-scale, provider-based programs and cannot be easily generalized to the large-scale, population-based programs that use a very different approach. The evidence for the population-based programs is limited and mixed. Some studies have found substantial improvements in quality of care,6-8 but others have not.9,10 Most notably, results from the Medicare Health Support demonstration project did not suggest that the intervention was able to improve quality of care provided to Medicare beneficiaries in the first year of the project.11
More research is needed to provide a conclusive answer to the question of whether or not population-based disease management can improve quality of care and, if so, which interventions have which effects under which conditions. Much in the same way that some (though not all) disease management programs might be able to reduce direct medical costs, the likelihood that the programs will affect quality of care and the magnitude of that effect will depend on a number of factors, such as the baseline level of quality, the design and execution of the program, and the targeted conditions and populations. Especially if the justification for investing in disease management shifts from reducing cost to providing value for money, we need to understand and quantify this value so that we can compare the cost-effectiveness of disease management with that of other possible interventions.
The study by Chan and Cooke clearly contributes to the knowledge base in this area.5 It also demonstrates that creatively taking advantage of so-called natural experiments can yield interesting results without the need for controlled trials. However, many more such studies are needed to fully understand the impact of disease management. The industry would be well served to contribute to this body of evidence by submitting its results to the scrutiny of the peer-review process. In the meantime, as the authors state appropriately, purchasers should adhere to the principle of caveat emptor and review their disease management programs both according to the results they deliver and the methods used to calculate those results.
Author Affiliation: RAND Corporation, Arlington, VA.
Author Disclosure: Dr Mattke conducts research and consults on projects for purchasers or operators of disease management programs.
Address correspondence to: Soeren Mattke, MD, DSc, RAND Corporation, 1200 S Hayes St, Arlington VA 22202. E-mail: email@example.com.
1. Congressional Budget Office. An analysis of the literature on disease management programs. October 2004. http://www.cbo.gov/showdoc.cfm?index=5909&sequence=0. Accessed March 13, 2008.2. Mattke S, Seid M, Ma S. Evidence for the effect of disease management: is $1 billion a year a good investment? Am J Manag Care. 2007;13(12):670-676.
4. Disease Management Association of America. Enhancing Health Care Quality Through Disease Management and Care Coordination. February 2007. http://www.dmaa.org/pdf/QualityandDM.pdf. Accessed March 13, 2008.
6. Fireman B, Bartlett J, Selby J. Can disease management reduce health care costs by improving quality? Dis Manag. 2004;23(6):63-75.
8. Villagra VG, Ahmed T. Effectiveness of a disease management program for patients with diabetes. Health Aff (Millwood). 2004;23(4):255-266.
10. Mattke S, Jain AK, Sloss EM, Hirscher R, Bergamo G, Oâ€™Leary JF. Effect of disease management on prescription drug treatment: what is the right quality measure? Dis Manag. 2007;10(2):91-100.