|
MARKETWATCHShould Patients Receive Secondary Prevention Medications For Free After A Myocardial Infarction? An Economic Analysis
Taken in combination, aspirin, beta-blockers, angiotensin-converting enzyme (ACE) inhibitors, and statins (combination pharmacotherapy) greatly reduce cardiac events. These therapies are underused, even among patients with drug insurance. Out-of-pocket spending is a key barrier to adherence. We estimated the impact of providing combination pharmacotherapy without cost sharing ("full coverage") to insured patients after a myocardial infarction (MI). Under base-case assumptions, compared to standard coverage, three years of full coverage will reduce mortality and reinfarction rates and will save $5,974 per patient. Our analysis suggests that covering combination therapy for such patients will save both lives and money.
A NUMBER OF medications have demonstrated efficacy for the secondary prevention of coronary heart disease (CHD)–related events.1 Taken in combination, aspirin, beta-blockers, angiotensin-converting enzyme (ACE) inhibitors, and statins have been estimated to reduce the relative risk of CHD mortality by 80 percent, compared with placebo.2 These medications continue to be greatly underused.3 The cost of prescription drugs borne by patients is an important cause of medication underuse, even among patients with drug coverage.4 In a study of patients enrolled in fifty-two health plans, doubling patient copayments resulted in a 34 percent reduction in the use of lipid-lowering agents and a 26 percent reduction in use of antihypertensives.5 These data suggest that reducing patients’ out-of-pocket costs could increase the use of preventive cardiovascular drugs. In fact, the incremental cost of "full coverage" (that is, eliminating patient cost sharing) for secondary prevention medications provided to patients after myocardial infarction (MI) could be entirely offset by the cost savings of additional clinical events that are averted. A similar concept was demonstrated in a recent analysis of the use of ACE inhibitors in diabetics.6 Accordingly, we sought to assess the clinical and economic implications of providing aspirin, beta-blockers, ACE inhibitors, and statins ("combination pharmacotherapy") to post-MI patients without cost sharing.
Analytic model. We created a model to estimate anticipated changes in event rates and health care spending if combination pharmacotherapy were provided without any out-of-pocket cost to patients age sixty-five and older who have some drug coverage and were discharged after hospitalization for MI.7 We conducted our analysis from the perspective of a typical insurer that provides coverage for both medications and medical care. To assess the potential benefit of full coverage, we used observed post-MI rates of death, reinfarction, nonfatal stroke, readmission for congestive heart failure (CHF), and medication adherence, and estimates of the treatment effect of combination pharmacotherapy to calculate the expected number of events that would occur if post-MI patients did not receive any secondary prevention. From these untreated rates and estimates of how much eliminating cost sharing improves adherence, we calculated the number of events that would be observed with full coverage. The incremental costs of expanded coverage were estimated by comparing the cost of drugs and clinical events currently faced by insurers with those that would be incurred with full coverage. Drug costs were calculated as the product of the actual drug costs, the proportion of costs typically paid by insurers, and the proportion of patients that adhere to therapy. Event costs were calculated as the product of event costs and estimated event rates. As such, insurers’ costs increase as they pay a greater proportion of medication costs and as patients fill more prescriptions; insurers’ costs decrease as patients experience fewer events. Baseline inputs. We estimated the ability of combination pharmacotherapy to reduce CHD-related events from the literature.8 We estimated current adherence from a World Health Organization (WHO) study of patient adherence with long-term therapies.9 We obtained the change in drug usage resulting from elimination of copayments from Dana Goldman and colleagues.10 Current rates of post-MI reinfarction, stroke, and CHF were measured using Medicare claims data.11 Cost data for combination pharmacotherapy were obtained from a major online pharmacy, drugstore.com. Patients were assumed to pay an average of 32 percent of medication costs at present.12 The cost of care attributable to post-MI events were derived from Allison Rosen and colleagues.13 Costs associated with CHF were obtained from Lawrence Liao and colleagues.14 We ignored the cost of death, re-infarction, stroke, or CHF that did not result in hospitalization.
Base-case and conservative-case analyses.
We performed two primary analyses. In the base-case analysis, we used the inputs described above (see Exhibit 1
Sensitivity analyses. We performed four sets of sensitivity analyses. First, one-way sensitivity analyses varied inputs using the ranges presented in Exhibit 1  . Second, two-way sensitivity analyses assessed the simultaneous impact of our estimated treatment effects and other primary assumptions.16 Third, we assessed the impact of changing the combination pharmacotherapy regimen to include medications thought to have greater efficacy in reducing post-MI events or increasing medication adherence, but at a higher cost.17 Finally, we evaluated the impact of extending the time horizon of full coverage for up to ten years.
Under base-case assumptions, full coverage is expected to increase compliance from 50 percent to 76 percent. For every 100 post-MI patients, this would result in 1.1 fewer deaths, 13.1 fewer nonfatal MIs, 1.2 fewer nonfatal strokes, and 6.6 fewer readmissions for CHF than with current coverage (Exhibit 2  ). Expanded coverage would cost insurers an average of $644 more per patient but would avert $6,770 in event-related costs on average. Therefore, insurers would save $5,974 per patient.
Under conservative-case assumptions, full coverage would increase compliance from 50 percent to 63 percent and result in 0.4 fewer deaths, 5.7 fewer nonfatal MIs, 0.5 fewer non-fatal strokes and 1.8 fewer nonfatal CHF readmissions per 100 post-MI patients. Coverage would cost insurers an average of an additional $550 per patient but would avert an average of $1,731 in event-related costs. Therefore, under conservative assumptions, full coverage for combination pharmacotherapy would save lives and money.
Sensitivity analyses.
Our findings are robust to all of our base-case assumptions in one-way sensitivity analyses. Using conservative-case inputs, full coverage is cost saving if at least 2.5 percent more patients adhere to therapy, if event costs and event rates are at least 30 percent of baseline inputs, and if drugs costs are $6,100 or less for three years of therapy. Similarly, full coverage will save money if treatment effects are at least 38 percent of conservative-case estimates; insurers’ costs will be reduced if combination pharmacotherapy reduces the relative risk of nonfatal reinfarction by 44 percent or more, compared with placebo (Exhibit 3
This analysis suggests that providing full coverage for combination pharmacotherapy to insured post-MI patients could simultaneously save lives and money and that the magnitude of such savings could be substantial. Based on our base-case analysis, eliminating out-of-pocket drug costs for combination pharmacotherapy for the 423,000 Americans with drug insurance who will experience their first MI in 2006 would save 4,736 lives and would save insurers more than $2.5 billion.19 Our results highlight the reductions in treatment costs that might be achieved by reducing barriers to the use of preventive medications. A similar benefit has been demonstrated for other conditions, including diabetes and hyperlipidemia, which like MI have treatments that are highly effective, relatively inexpensive, and greatly underused.20 Accordingly, large employers and insurers have begun experimenting with the selective reduction of copayments and have found very favorable short-term economic returns.21 Insurer incentives. Because of the fragmented nature of the U.S. health care system, many insurers feel little incentive to adopt similar "benefit-based" formularies.22 Patients frequently switch insurers because of changes in employment or adverse selection. Thus, insurers face the possibility that they might bear much of the cost of preventive therapy while other insurers reap the savings from averted clinical events. However, our analysis suggests that insurers garner benefits within the first year of full coverage after an MI and should have sufficient incentive to provide full coverage for combination pharmacotherapy, at least in the short to medium run. Because the implications of insurance "churning" are largely irrelevant in systems with single-payer comprehensive coverage, our results have even clearer implications for Medicare. Assuming that Part D covers an average of 37 percent of beneficiaries’ drug costs, under base-case assumptions, we estimate that providing post-MI Medicare beneficiaries with full coverage for combination pharmacotherapy will save more than $5,600 per patient over a three-year period.23 Analytic assumptions. It is reassuring that our results are robust to many of our model assumptions. Full coverage will reduce costs over longer and shorter time horizons than we primarily considered, if treatment effects are smaller than assumed and if adherence rates are lower than projected. Similarly, although conservative assumptions decrease the magnitude of the potential improvements in health and savings, full coverage still saves lives and money. Nevertheless, our models rely on some inputs whose magnitudes are uncertain. The true benefit of combination pharmacotherapy has not been tested in a randomized controlled trial (RCT) and has been debated in the literature.24 Our models indicate that full coverage is cost saving as long as the relative risk reduction from combination pharmacotherapy is as little as 40 percent of our conservative-model inputs. By way of comparison, this is equivalent to the relative risk reduction over placebo expected from statin therapy alone without the other constituents of combination therapy as obtained from conservative-case estimates of the RCT literature.25 Similarly, the proportion of people who will remain adherent with therapy once out-of-pocket costs are removed has not been fully evaluated. Existing analyses of cost-related drug underuse are confounded by selection bias: People who are eligible for benefits or choose to pay for them might differ in important ways from those who do not. Accordingly, our results call for the more rigorous evaluation of whether full coverage for some conditions is as attractive as it appears to be. To overcome the limitations of existing data, these studies should be randomized or quasi-experimental in design; should evaluate a series of "candidate conditions" such as MI, diabetes, and CHF; and should be replicated in insurance markets of different structures. Although the primary focus of our analysis was the increment in adherence and the resultant improvement in outcomes that specifically results from the elimination of cost sharing, there are clearly other ways to improve outcomes and adherence. In our sensitivity analyses, we found that the number of lives saved and events averted might be higher for alternative therapeutic regimens, such as high-dose statins and extended-release metoprolol, while cost savings are maintained. Accordingly, an evaluation of full coverage should also assess the impact of which specific drugs are prescribed. In the case of combination pharmacotherapy, this will mean determining whether insurers should provide full coverage for all ACE inhibitors, beta-blockers, and statins or for only selected agents. Study limitations. Our analysis is subject to several other limitations. First, although we explored providing coverage for longer time frames in our sensitivity analysis, we did not consider a lifetime horizon, and we did not estimate changes in quality of life. Since many of the events that are prevented by combination pharmacotherapy occur in the months following an MI, evaluating a short time horizon is probably reasonable. It is, of course, likely that other conditions could arise because of averted death or reinfarctions; the clinical and economic costs associated with these could reduce the magnitude of benefit from combination pharmacotherapy in the long run. Second, we performed our analysis from the perspective of a typical health insurer, to make the business case for aligning insurers’ incentives and patients’ health outcomes in coverage decision making. As a result, we did not take a societal perspective, as is often recommended for cost-effectiveness analyses.26 Fortunately, the societal case for free post-MI drugs is much easier to make, because society does not require life-saving interventions to be cost saving. Removal of financial barriers to highly effective preventive medications would also likely lead to more equitable distribution of health care services. Finally, we did not evaluate the impact of providing insurance coverage to those who now lack it. Doing so would result in sizable increases in drug costs faced by insurers; however, it would likely also induce a much greater number of currently nonadherent patients to begin using these drugs than among the currently insured population that we assessed. IN CONCLUSION, ELIMINATING out-of-pocket costs for post-MI combination pharmacotherapy among insured patients has the potential to simultaneously save many lives and substantial medical costs. Our analysis supports calls to reconsider how drug formularies and insurance companies structure their benefits, especially for medications of proven efficacy that are greatly underused.
Niteesh Choudhry (nchoudhry{at}partners.org) is an assistant professor at Harvard Medical School and an associate physician at Brigham and Women’s Hospital in Boston, Massachusetts. Jerry Avorn is chief of the Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women’s and a professor at Harvard Medical School. Elliott Antman is director of the Samuel A. Levine Cardiac Unit at Brigham and Women’s and a professor at Harvard Medical School. Sebastian Schneeweiss is an associate professor at Harvard’s Medical School and School of Public Health. William Shrank is an associate physician at Brigham and Women’s and an instructor at Harvard Medical School. Elliott Antman is a member of the TIMI Study Group, which receives grant support for clinical trials from Pfizer, BMS, Merck, COR, Schering-Plough, and Sanofi-Aventis. Sebastian Schneeweiss is funded by the Agency for Healthcare Research and Quality (Grant no. 2-R01-HS10881).
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Study Data And Methods
Study Results
