Deployment of Preventive Interventions have been proven many times, but rarely implemented – May 2018

Deployment of Preventive Interventions — Time for a Paradigm Shift

New England Journal Of Medicine 2018; 378:1761-1763, DOI: 10.1056/NEJMp1716272
Katherine Pryor, M.D., and Kevin Volpp, M.D., Ph.D.
From the Center for Health Incentives and Behavioral Economics, and the Department of Medical Ethics and Health Policy, Perelman School of Medicine, University of Pennsylvania, Philadelphia.

In 2002, Knowler et al. reported results of a landmark study — a large, randomized, controlled trial comparing a behavioral intervention with medical therapy in the prevention of diabetes.1 Over a mean follow-up period of 2.8 years, the lifestyle-modification program, known as the Diabetes Prevention Program (DPP), reduced the incidence of diabetes by 58% as compared with placebo among people with elevated fasting and post-load plasma glucose concentrations. Metformin reduced the incidence of diabetes by 31% as compared with placebo.

Despite these findings, insurers have been slow to provide coverage for DPP-like interventions. In 2016, the Centers for Medicare and Medicaid Services piloted the program and determined that it improved the quality of patient care and reduced net Medicare spending, prompting a goal of expanding the DPP nationwide by 2018. Although coverage of metformin has been ubiquitous since it was introduced in the United States in 1995, many private insurers started covering the DPP only recently.

Financial incentives for tobacco cessation during pregnancy provide another example of an effective behavioral intervention that hasn’t been translated into practice. Smoking during pregnancy is a leading cause of maternal and neonatal morbidity and mortality, particularly among socially disadvantaged women and their children, and has long been a public health target. In the United States, such smoking rates have decreased only marginally in recent decades. A Cochrane review concluded that financial incentives are the most effective intervention in this population and can lead to quit rates up to four times higher than those achieved with other interventions. But such incentives haven’t been implemented in routine care of pregnant women.

Why are highly effective preventive interventions adopted slowly, if at all? The first issue is that, historically, far more resources have been devoted to treating disease than to preventing it; in 2015, only 3% of health care dollars were spent on preventive services. However, ongoing shifts in health financing are creating incentives for providers to pay more attention to modifiable risks such as antenatal smoking. Hospitals participating in accountable care organizations, for example, save thousands of dollars for each neonatal intensive care unit stay they prevent.

Second, treatments determined by the Food and Drug Administration (FDA) to be safe and effective are usually covered by insurers regardless of their cost, but preventive services have been held to a higher standard: they are often assessed on the basis of whether they generate a positive return on investment and save money in the short term. This disparity leads to overprovision of treatments and underprovision of preventive services, a trend that is exacerbated by high turnover in many health insurance markets. Because insurance contracts tend to be only 1 year long, insurers don’t want to spend money to prevent disease in members who may be covered by a different insurer in the near future.

Even Medicare — which typically covers beneficiaries for life — holds preventive services to a higher standard, applying cost-effectiveness analyses when making coverage decisions about preventive services but not treatments. This double standard has resulted in coverage of cost-ineffective therapies with prices of up to hundreds of thousands of dollars per quality-adjusted life-year, including treatments of questionable benefit (such as Avastin bevacizumab for metastatic breast cancer after the FDA withdrew support for this use).2 A recent study showed that reallocating current Medicare expenditures toward “dominant” (cost-saving and health-increasing) interventions would result in efficiency gains and improvement in the aggregate health of Medicare beneficiaries at no additional cost. 3

Third, behavioral interventions often represent unfamiliar territory for providers. Writing a prescription is generally easy and routine, and medications are heavily marketed and seen as being easier to broadly disseminate with predictable efficacy. But this assumption doesn’t always hold true. The diabetes-prevention trial, for example, found a less heterogeneous effect in the behavioral-intervention group than the metformin group: the DPP was associated with a substantial reduction in the incidence of diabetes regardless of patients’ baseline risk, but only the highest-risk patients in the metformin group saw a similar benefit.4

Fourth, many providers seem largely unaware of the high rates of medication nonadherence among their patients and don’t have effective tools for improving adherence. Prescribing a medication is simple for a provider, but taking a medication does not appear to be simple for many patients. Outside of clinical trials, adherence to medications is often low. In the year after a heart attack, for example, only 40 to 45% of patients take their medications as prescribed.

Finally, concerns about scalability are often a barrier in the deployment of proven behavioral interventions. Consider financial incentives for antenatal smoking cessation: such a program would require an intensive schedule of in-person visits for biochemical assessment of abstinence. Although such assessments could be built into the standard prenatal care schedule in which urine collection during office visits is common, the program would still require a shift in what providers do during visits. Assessing smoking status and counseling against ongoing tobacco use are already part of the routines of antenatal care providers, but overseeing a reward system tied to smoking cessation would be new. There is no readily available infrastructure for clinics to manage such a program, and developing one might require a third party. Health plans could be the third party that assesses cotinine test results and administers rewards, but this would need to be done in a way that minimizes delays and administrative complexity.

These barriers signal a need to rethink and optimize the infrastructure and platforms on which health services are currently delivered. For example, leveraging Web-based technologies or wireless devices would address many scalability concerns and help facilitate adoption of certain behavioral interventions. Consider the DPP-like behavioral intervention: it is labor- and time-intensive for both staff and participants, with requirements including supervised physical-activity sessions, individualized coaching, and case managers. There are geographic limitations in availability, and only a small fraction of people with prediabetes enroll. However, online versions of DPP-like interventions now exist and feature greater schedule flexibility, a personal coach, and online peer-support groups, eliminating the need for in-person assessments. Online programs result in weight loss similar to that seen in the standard DPP.5 Web-based platforms have been used successfully in contingency management for both chronic disease and substance abuse. In these programs, biochemical markers such as carbon monoxide and blood glucose or vital signs such as blood pressure can be assessed by means of virtual observation of patients using monitoring equipment in their homes. Such platforms facilitate important innovations in supporting management of a growing range of diseases and care for hard-to-reach populations.

For health care’s transformation from a volume- to a value-based framework to be successful, we think that putting coverage of preventive services and treatments on more even footing will deliver great value. Historically, preventive services have been adopted only if they have been proven to save money, whereas treatments have been evaluated on the basis of their benefits and risks, without consideration of costs. The slow movement toward coverage and implementation of behavioral interventions may accelerate substantially as population-based financing becomes the norm. Payment reform has the potential to bring about a paradigm shift whereby all services are evaluated using the same standard: Do they improve health at a reasonable price? Such a shift could increase insurers’ willingness to cover high-value preventive services and providers’ interest in designing ways to facilitate the uptake and deployment of those services on a broader scale — enabling us to achieve better health at lower cost.

References

• 1. Qaseem A, Snow V, Gosfield A, et al. Pay for performance through the lens of medical professionalism. Ann Intern Med 2010;152: 366-9.
• 2. Berwick DM. Era 3 for medicine and health care. JAMA 2016;315:1329-30.
• 3. Casalino LP, Gans D, Weber R, et al. US physician practices spend more than $15.4 billion annually to report quality measures. Health Aff (Millwood) 2016;35:401-6.
  https://doi.org/10.1377/hlthaff.2015.1258 available at sci-hub
  
  “A study of twenty-three health insurers found that 546 provider quality measures were used, few of which matched across insurers 10 or with the 1,700 measures used by federal agencies”
  “State and regional agencies currently use 1,367 measures of provider quality”
  “The current system is far from being efficient and contributes to negative physician attitudes toward quality measures.”
• 4. Higashi T, Shekelle PG, Adams JL, et al. Quality of care is associated with survival in vulnerable older patients. Ann Intern Med 2005;143:274-81.
• 5. Hemingway H, Crook AM, Feder G, et al. Underuse of coronary revascularization procedures in patients considered appropriate

candidates for revascularization. N Engl J Med 2001;344:645-54.