Medication use vividly illustrates the paradox of overuse, underuse, and misuse of health care in America. While a growing segment of the population overuses certain medicines — including opioids, central nervous system depressants, and stimulants — other medicines known to deliver clinical value are substantially underutilized by people who could benefit.
In one recent study, for example, the rates at which people discharged from the hospital after a heart attack were adherent to cardiovascular medications (antihypertensives, beta-blockers, and statins) were relatively low (41 percent, 49 percent, and 55 percent, respectively), even after the insurer waived patients’ out-of-pocket costs. In a separate study, among patients who started medicines for diabetes, bone density, and glaucoma, rates of adherence after six months were just 66 percent, 56 percent, and 47 percent, respectively.
The total economic impact of medication non-adherence — which contributes to costly health complications, worsening of disease progression, and preventable utilization — has been estimated to be as much as $290 billion. But despite growing recognition of the problem, little progress has been made to improve medication adherence at a population level. However, as explained in more detail below, several new market and policy dynamics have the potential to substantially mitigate barriers to—and indeed trigger new investment in—improving medication adherence. These developments include lower drug costs; a growing evidence base on the consequences of nonadherence and on interventions that improve adherence; new adherence-related quality measures; new drug benefit designs and other payment innovations; and new technologies and analytical services that facilitate measuring and improving adherence.
Why Non-Adherence Has Been A Persistent Problem
Several barriers have stood in the way of improving adherence. First, evidence relating adherence to clinical and economic outcomes has been inconclusive. While many studies (see Appendix B of linked document) have identified a positive association between medication non-adherence and adverse clinical outcomes, the validity of these studies has been questioned because the association between adherence and outcomes is thought to be confounded by hard-to-measure patient characteristics. As a result, public and private payers have tended to view medication adherence as an intuitively important quality goal, but one with uncertain effects on the total cost of care.
Second, many of the drugs used to treat or prevent prevalent chronic diseases have been costly, posing economic barriers for payers and patients alike. Thus, improving medication adherence has been an expensive proposition when the cost of increased utilization for relatively expensive medications is weighed against uncertain clinical and economic offsets.
Third, while adherence-enhancing interventions have been described, the most effective strategies have involved trusted health professionals taking the time to engage with patients regarding their goals and preferences and counsel them on the purpose and goals of therapy, the administration and duration of therapy, what to expect from the treatment, and how to handle common barriers such as side effects or cost. Yet, providers in a fragmented health care delivery system have been neither equipped to monitor and modify patients’ health behaviors related to medication use, nor accountable for this.
New Opportunities To Realize Improvements In Medication Adherence
As mentioned above, several changes in the health care system offer the opportunity to overcome these barriers to improving medication adherence.
Lower drug prices. First, the cost of common chronic medications has declined substantially in recent years, and will continue to do so. This is due primarily to the loss of patent protection for blockbuster drugs in some of the most commonly prescribed classes, including statins, antihypertensives, and antiplatelet drugs. The generic share of retail prescriptions in the U.S. has grown from 18.6 percent in 1984 to 74.5 percent in 2009, with a notable acceleration in recent years.
Within Medicare Part D, the federal program to provide drug benefits to Medicare beneficiaries, the average cost per day of treatment has declined from $1.50 per patient in 2006 when the program was implemented to $1.00 per patient in 2010; this figure is expected to reach $0.65 per patient by 2015. Lower effective drug prices for common chronic conditions have significantly improved the cost-effectiveness of several chronic therapies and have increased the likelihood that appropriate use of safe and effective medicines results in net cost savings from a payer and societal perspective.
Better evidence on the consequences of non-adherence and interventions that work. Our understanding of the effects of medication adherence and other quality and cost-related outcomes is also improving, thanks to the increasing number of well-designed studies investigating these connections. Recent papers by Roebuck et al. (2011), Stuart et al. (2011), and Choudhry et al. (2011) each identified positive associations between medication adherence and cost savings attributable to lower unnecessary medical utilization.
For example, Roebuck et al. examined the impact of medication adherence on medical spending in four therapeutic areas: congestive heart failure, hypertension, diabetes, and dyslipidemia. They found that combining increases in spending on medications with corresponding decreases in medical spending associated with increased adherence (e.g., from fewer hospitalizations and emergency department visits) yielded benefit-cost ratios ranging from 3.1:1 for dyslipidemia to 10.1:1 for hypertension.
These studies — which report promising results regarding the impact on medication adherence in different therapy areas and for different patient populations — sought to control for patient characteristics and address other methodological concerns that had previously subjected the literature on adherence and medical outcomes to criticism. The evidence base for adherence, and the quality of that evidence, is improving.
Changes to drug benefit designs. The Patient Protection and Affordable Care Act of 2010 (ACA) closed the so-called “doughnut hole”—the coverage gap in the Medicare drug benefit—by offering additional medication discounts for beneficiaries. According to the Centers for Medicare and Medicaid Services (CMS), this policy has now saved Medicare beneficiaries $3.4 billion on prescription drug costs, which equated to $837 per beneficiary, on average, in the first three months of 2012.
An increasing number of private payers have also implemented “value-based insurance designs,” in which the cost-sharing requirements of highly cost-effective medications is lowered to encourage adherence. The Mercer National Survey of Employer-Sponsored Health Plans demonstrates that value-based insurance design use is increasing and that 81 percent of large employers plan to offer it in the near future. Evidence to date suggests that such changes to pharmacy benefit designs will not cure the problem of non-adherence alone, but by mitigating the cost barrier from a patient’s perspective, these benefit designs may increase the impact of other interventions.
Consensus-based adherence quality measures. Much recent progress has been made by researchers and quality measurement bodies on the development of new performance measures related to medication adherence. The Pharmacy Quality Alliance (PQA), a non-profit stakeholder organization that develops performance measures related to medication use, now has developed, tested, and endorsed ten adherence measures covering a range of therapy areas.
Drawing on early research on new methods for measuring the “proportion of days covered” (i.e., the proportion of days on which a patient had a given medication available over a defined time interval), PQA measures have been endorsed by the National Quality Forum. Three such measures (for statin, oral diabetes, and antihypertension medications) were adopted by CMS in 2011 for quality reporting and payment purposes in the Medicare Advantage program (discussed below). PQA continues to develop additional adherence measures covering new therapy areas and to make adjustments to existing measures, driven by increasing demand by health care payers who view adherence as an important intermediate outcome in health care.
New health technologies and related analytic capabilities. The proliferation of health technologies will enable new strategies for improving adherence. For example, “e-prescribing” technology, which replaces prescriptions delivered by paper, phone, or fax, has the potential to reduce the incidence of medications being prescribed but never obtained by patients. A recent study reported that a significantly higher proportion of prescriptions are initiated (i.e., written, filled, and actually picked up at the pharmacy by patients) when doctors use e-prescribing technology compared to when they use other traditional prescribing methods. This suggests that even greater adherence benefits may accrue in the future as the adoption rates of e-prescribing and other complementary health technologies rise in the era of the “meaningful use” of health IT.
The coupling of “big data” and advanced analytic techniques enabled by health technology can also be used to make further progress on medication adherence rates, including monitoring adherence at a population level, identifying patients at risk of non-adherence, and developing better evidence on what types of interventional strategies designed to improve adherence work best and for which patients. Such innovations can lead to more personalized, more targeted adherence interventions that aim to prevent non-adherence from occurring rather than merely treating it once it does.
Changes in payment policy. Finally, public and private health care payers are now experimenting with new provider payment models—such as patient-centered medical homes, accountable care organizations (ACOs), and episode-based payments—designed to slow spending growth and achieve progressively better care quality. These initiatives will emphasize the need to maximize the returns available from effective use of medications since they represent the most common medical intervention in health care.
ACOs. Newly created ACOs will have new incentives to focus on improving medication adherence at a population level. By our count, 20 of the 33 quality measures to which Medicare ACOs are accountable are related to the safe and effective use of medications.
In addition to quality measures, ACOs are likely to view medication adherence as a priority since they are eligible to keep a portion of cost savings attributable to slowing the growth of health care spending on behalf of the patients to which they are accountable. Hospitals seeking to reduce avoidable hospitalization rates, for example, may be more likely to invest in post-discharge care management programs that increase the likelihood that patients are on the right medications and that they take them as directed.
Notably, Medicare ACOs will be accountable for slowing the growth of health care spending associated with Medicare Part A (hospital) and Part B (outpatient physician) services, but not Part D (medications) spending. This means that ACOs can invest in steps to increase medication adherence, triggering increases in Part D spending for which ACOs are not accountable, and benefit from resulting slower growth in utilization rates for physician and inpatient services in key patient subpopulations and in certain therapy areas. This creates an important opportunity for ACOs to invest in increasing medication adherence and for CMS and researchers to investigate the policy incentives created by the current ACO regulations.
Medicare Advantage “Star” ratings. Health plans providing both medical and pharmacy benefits to about 12 million Medicare beneficiaries in the Medicare Advantage (MA) program also stand to benefit when they can help beneficiaries increase medication adherence. CMS now publicly reports and provides financial bonuses to MA plans on the basis of “Star Ratings”, which for 2012 are determined based on plans’ performance on 53 measures of quality. MA plans earning 5 out of 5 available “Stars” can earn 5 percent in premium bonuses, 4-Star plans can earn 4 percent, and 3-Star plans can earn 3 percent. A hypothetical MA plan in 2012 that improved its Star Rating from 3- to 5-Stars by improving its overall performance on the 53 Star measures would have earned an additional $16 per patient per month in bonus payments, on average. Bonus payments can be used to enhance benefits or to lower MA plans’ premiums to attract additional market share.
Three of the 53 measures used to calculate Star Ratings indicate whether beneficiaries take statin, oral diabetes, and hypertension medications as directed by their doctors. Adherence measures are weighted heavily because they are considered “intermediate outcome” measures; together, they account for 11 percent of the overall Star Rating for MA plans in 2012. The impact of adherence measures on overall Star Ratings doubles (to 22 percent) when the value of adherence measures is combined with other clinical quality measures in the Star Ratings that typically improve with greater adherence to appropriate medications (i.e., cholesterol control and blood pressure control measures).
This program is spurring new attention in and new investments in medication adherence programs among Medicare Advantage plans.
Implications For Policy And Practice
Despite the emerging trends outlined here indicating positive signs for the future of medication adherence, further steps are needed to achieve large-scale improvements in this important public health priority.
First, more evidence is still needed regarding the full range of therapy areas and patient populations in which increasing adherence leads to better quality outcomes and financial returns. Better evidence is also needed regarding how different types of patients respond to different interventional strategies in a range of contexts (for example, older frail patients on multiple medications with limited social support).
Second, additional adherence quality measures are needed encompassing new therapy areas and patient populations. Further refinement to existing measures may also be appropriate, such as the incorporation of techniques to adjust measure performance for underlying patient characteristics.
Third, more commitment is needed by public and private payers to include adherence quality measures as a basis for reimbursement or bonus payments for providers and pharmacists. Adherence-related quality measures and bonus payments can be expanded within Medicare to cover more beneficiaries (e.g., those in Stand-Alone Prescription Drug Plans), and similar policies can be adopted by Medicaid and commercial payers. Finally, further innovation and testing is needed of interventional strategies that combine effective approaches for targeting patients at the highest risk of non-adherence with efforts to personalize effective interventional strategies that work best for each patient.
While there is much work to do, the market and policy trends described in this paper suggest that the issue of medication adherence may finally be receiving the attention indicated by its potential value for improving health care.