Economics of Prevention Workshop
August 27-28, 2015

Executive Summary

On August 27 and 28, 2015, the National Institutes of Health (NIH) Health Economics Common Fund Program and Office of Disease Prevention (ODP) sponsored a workshop on the Economics of Prevention. The goals of the workshop were to showcase research supported by the Health Economics Common Fund Program, discuss the state of the research field, and identify gaps and opportunities to be addressed in future research.

The workshop was organized in four main sessions: (1) use of preventive services, (2) targeting and personalization in prevention, (3) behavioral economics—insights for prevention, and (4) evaluating preventive interventions. Participants included leading researchers in economics and preventive medicine and staff from the NIH and other federal agencies within the U.S. Department of Health and Human Services.

Keynote Address: Prevention, Harm Reduction, and Policy

The keynote address, delivered by Richard Frank, Assistant Secretary for Planning and Evaluation of the U.S. Department of Health and Human Services, distinguished between harm reduction and use reduction prevention policies. Although both approaches aim to minimize the social costs of harmful consumer products or behavior, harm reduction policies seek to reduce the average harm to users, whereas use reduction policies seek to reduce use itself. Examples of harm reduction policies include needle exchange programs, which provide clean needles to intravenous drug users to limit the spread of infectious disease, and e-cigarettes, which might be a less harmful alternative to smoking.

Harm reduction policies are controversial for ideological and evidentiary reasons, and few analyses comprehensively consider the net effects of harm reduction programs. The extent to which harm reduction policies reduce average harms, and whether they increase or decrease rates of use, is unclear. Modeling the outcomes of harm reduction programs is challenging because these programs change incentives for complex social behaviors that are strongly affected by social context. It is likely that the optimal solution for multiple domains might be a combination of harm reduction and use reduction strategies.

Session 1: Use of Preventive Services

A variety of incentive strategies have been suggested to encourage greater use of high-value preventive services. Presenters discussed studies of incentives for increased use of prevention targeting employees, providers, and those in specific types of health plans, as well as reductions in out-of-pocket costs. Overall, the findings indicate that reducing the cost that individuals pay for the services or offering them monetary incentives for participating in prevention programs appears to increase their use modestly. The impacts of increased physician reimbursements and switching from traditional to high-deductible health plans on preventive services utilization are minimal.

Participants discussed the possibility that even a modest cost for preventive services might change consumer perceptions and create barriers to seeking services. Understanding the barriers to preventive services use, in addition to possible incentives, is an important research topic.

Session 2: Targeting and Personalization in Prevention

TRisk stratification tools, biomarkers, and other targeted or personalized approaches might help prevent negative health outcomes while minimizing the harms of overtreatment. However, the effectiveness and cost-effectiveness of these approaches is uncertain. This session featured modeling studies that elucidated circumstances under which greater targeting would and would not be cost-effective, a theoretical overview of the relevant economic factors, and of how risk-stratified analyses of clinical trials can lead to different, more nuanced conclusions than traditional approaches.

The cost-effectiveness of targeted prevention strategies is likely context-specific and sensitive to a number of assumptions. Risk stratification, on the other hand, is a general approach that can be applied in many different contexts. Workshop participants identified the incorporation of treatment harms into risk models as an important topic for further study.

Session 3: Behavioral Economics—Insights for Prevention

Behavioral economics, a method of economic analysis that applies psychological insights into human behavior to explain economic decision making, might provide useful insights for prevention. Presentations in this session explored different approaches to promote healthful decisions among clinicians and the public. Evidence suggests that social interventions, such as peer comparisons, are generally more effective than educational interventions in changing behavior. Interventions that make the healthful choice the easy choice are particularly effective. These include improved default options, active choice to require an opt-in/opt-out selection without a default, and providing services in ways that are physically and temporally compatible with individuals’ daily routines.

Behavioral interventions can provide inexpensive and scalable means to steer individuals toward desired behaviors; however, the effects of individual interventions are typically modest. A combination of multiple strategies will likely be needed to achieve optimal outcomes. Several additional research questions include the durability, social acceptability, and cost-effectiveness of behavioral interventions.

Session 4: Evaluating Preventive Interventions

Models are useful tools for predicting and evaluating possible outcomes of public health policies. Because achieving public health outcomes and assessing cost-effectiveness require a long timeframe, it is often not feasible for governments to wait for empirical observations before implementing policies. Models can help separate the effects of policies from potential confounding factors, account for heterogeneity in individual characteristics, and provide a variety of outcome measures and cost estimates to satisfy a range of information needs.

Nonetheless, models have several inherent limitations. For example, models cannot predict the future or determine causal relationships. Models rely on accurate data inputs and on humans to make final policy judgments. The results of modeling studies, such as a microsimulation model of the cost-effectiveness of mammography screening in several real and hypothetical scenarios, can inform the development of cost-effective prevention policies and help policy makers anticipate needed modifications as the landscape changes.

The Future of the Economics of Prevention

It is important to identify ways in which prevention policies can be better informed by a strong evidence base. Researchers have begun to identify effective and ineffective approaches, although many evidentiary gaps remain. Sharing models and improving input data, such as by identifying the characteristics and needs of different populations, may lead to methodological improvements that expand the evidence base.

Workshop participants agreed that in the future conducting many small, inexpensive randomized controlled trials and observational studies can help identify promising policies to study in larger contexts. Results of these trials should be interpreted cautiously, and interventions should be scaled up in a stepwise manner to determine their reproducibility in different contexts. Combining multiple promising interventions may further augment desired outcomes.