Editor’s note: One of the authors of this post, Peter Neumann, will be discussing issues related to the post at a Health Affairs September 13 event, “Understanding The Value of Innovations In Medicine.”

In 1996, the U.S. Panel on Cost-Effectiveness in Health and Medicine recommended that analysts conducting cost-effectiveness analyses (CEAs) should perform a reference case analysis, following a set of standard methodological practices to improve comparability and quality. They further recommended that such analyses assume a societal perspective, reflecting the perspective of a decision maker allocating resources broadly across the entire population (such as consequences of interventions that fall outside the health sector).

Since publication of the original Panel’s report, researchers have published thousands of CEAs. However, as highlighted in the recently published report of the Second Panel on Cost-Effectiveness in Health and Medicine (Note 1), many if not most CEAs have failed to use a societal perspective. Only 341 (29 percent) of 1,163 cost per quality-adjusted life year (QALY) analyses published through 2005 adopted a societal perspective, for example.

Most published CEAs have omitted important elements—e.g., excluding costs related to patient and caregiver time or costs outside the health sector. Moreover, since the original Panel’s book, decision making bodies (e.g., the National Institute of Care Excellence) that have used CEAs to inform reimbursement decisions have generally taken a more focused “payer” perspective.

As the Second Panel recognized, a societal perspective is useful because it reflects the broad public interest. However, the Second Panel also acknowledged that no particular decision maker or budget holder has such a perspective. Thus, the Panel recommended that CEAs report two reference case analyses, one based on a health care sector perspective (which more closely resembles the purview of health payers) and one on a societal perspective. The Second Panel also recommended inclusion of an “impact inventory,” or structured table, listing each intervention’s health-related effects and nonhealth-related effects; the latter category includes such areas as economic productivity, social services, the legal and criminal justice systems, education, housing, and the environment.

This post examines why so few CEAs have incorporated a societal perspective. We offer several policy ideas that could help ensure that assessments of drugs and a broad array of medical technologies and services incorporate the entire range of costs and benefits.

US Value Assessment Frameworks And ‘Perspective’

The “perspective” or viewpoint of an analysis is a central issue for the “value assessment frameworks” that have been promulgated by various medical professional societies and other organizations, e.g., the American College of Cardiology (ACC)/American Heart Association (AHA), the American Society of Clinical Oncology (ASCO), the Institute for Clinical and Economic Review (ICER), Memorial Sloan Kettering Cancer Center (MSKCC), the National Comprehensive Cancer Network (NCCN), and Faster Cures/Avalere (FC). Because these frameworks have different objectives and intended audiences, the perspective underlying their approaches also varies. Some frameworks (e.g., those from ASCO, FC) aim to inform shared decision making between patients and physicians, and are thus oriented toward a patient perspective. Unsurprisingly, they take a narrower view on analytic perspective, though they still may omit components important to patients. For example, ASCO’s framework has included drug costs, but not other costs that are likely to be relevant to patients, such as those related to patient or caregiver time.

Other frameworks use population-based approaches that resonate more with payers and health systems and raise additional questions about perspective. ICER, for example, has developed a two-pronged approach. The first part comprises a conventional CEA, under which ICER back-calculates a “long-term value for money” price based on a cost-per-quality-adjusted-life-year threshold between $100,000 and $150,000, with some allowances for thresholds outside that range. Part two is a budget impact analysis in which ICER creates a budget “alert” based on a range of assumptions on a drug’s price and level of use, and reflecting anticipated national GDP growth rates and the number of new drugs approved each year.

The ICER approach essentially assumes a health system perspective, though it allows for consideration and discussion (if not full incorporation into cost-effectiveness ratios) of broader elements of value. (ICER’s appraisal committee is asked to consider elements of value that might not have been considered in the cost-effectiveness calculations — e.g., social values associated with severity of illness. It is assumed that these other considerations will help decision-makers consider the appropriate place in the range of cost-effectiveness ratios provided after considering all the other elements of value.) ICER’s CEAs focus on health costs, thus excluding non-health effects (though they do assume a lifetime horizon).

In its recent revised methodology, ICER noted it will conduct a scenario analysis to include work productivity when feasible, and will employ a new template to illustrate the components included in analyses, as the Second Panel recommended. Moreover, ICER’s budget impact analyses take a five-year time horizon, excluding any cost offsets beyond the five years, reflecting the shorter-term viewpoint of payers.

Society’s Challenge With Narrow Perspectives

Cost-effectiveness analyses that that take a narrow perspective tend to undervalue interventions—whether they are drugs, vaccines, surgical procedures, or public health programs—that have broad social impacts. Numerous interventions, such as those for children, can have sizeable positive spillover effects on a patient’s family and friends (see here, here, here, and here). Treatments such as those for schizophrenia or alcohol use disorders may reduce costs associated with crime. The inclusion of productivity costs in CEAs substantially improves the reported cost-effectiveness of many interventions, including those for depression and pain.

Perspective is also important because of the signals it sends to product innovators (or any innovator who seeks to be rewarded for their investments). Value-based pricing and reimbursement approaches encourage firms to develop products according to the manner in which value is calculated. Pharmaceutical companies, for example, will have stronger incentives to develop a drug for depression that helps people get back to work, for example, if they are rewarded for doing so in the form of value-based pricing that reflects productivity effects.

Why Taking A Societal Perspective Is Challenging For Value Assessment Frameworks And Payers

While compelling arguments exist for value frameworks to assume a societal perspective, in practice it is challenging for them to do so. ICER’s reports, for example, are used by private and public payers, whose budgets do not pay for patient or caregiver time costs, or for costs that fall on non-health sectors. (ICER reports may also be used by other groups such as employers, clinical guideline committees, and patient groups.) Moreover, individual enrollees tend to remain with private health plans for relatively short periods of time. Studies have reported, for example, that 20 percent of Americans switch insurance plans each year. Given the purviews of organizations they tend to inform, value assessment frameworks can hardly be faulted for taking a narrow perspective in their CEAs or a short-term window in their budget impact analyses. The problem is not the value assessment frameworks so much as the misaligned incentives in the health marketplace in which they operate.

To be fair, there may be other reasons why societal perspectives are generally not conducted, including challenges in collecting data and the uncertainty that comes with assumptions around the broad impact of health interventions on non-health sectors. The Second Panel on Cost-Effectiveness in Health and Medicine discussed these challenges extensively and included two worked examples to illustrate how they can be addressed.

What To Do?

In the United States, no individual payer takes a broad societal perspective. Even the Medicare program, though funded by taxpayers as well as its beneficiaries, does not cover all health costs associated with interventions and does not pay for non-health costs. In theory, the US government might create a single health technology assessment (HTA) organization and mandate that it issue cost-effectiveness analyses using a societal perspective. However, proposing such a step would very likely engender concerns in some quarters about the perils of a lone bureaucracy in charge of analyses and the shortcomings of “one-size-fits-all” medicine.

In a sense, some stakeholders want it both ways: to retain the country’s decentralized, pluralistic health system, but also to have HTA bodies that take a societal perspective. There is an irony here: a single HTA entity with a broad purview would be opposed by the same parties who favor a societal perspective.

In this environment, there is still a useful role for CEA. Ideally, analysts conducting CEAs should follow the Second Panel’s recommendation to report studies using both a health and a societal perspective. As noted, this would help clarify for audiences the broad consequences omitted in narrow perspectives. At the very least, cost effectiveness analysts should clearly delineate their perspectives and list the components included and excluded, as the Second Panel recommends.

Beyond that, several larger reforms would help. First, there should be more focus on the displacement of inefficient services in the existing system, and not simply a focus on new technologies. (Notably, ICER’s updated framework highlights the importance of identifying lowvalue services as part of the evidence review process, and states that future reports will include a list of cost-saving measures in the health system in the relevant clinical area.) Second, experiments should continue on bundled payment and other value-based payment arrangements that create incentives for providers to manage patients as efficiently as possible regardless of the service mix used. Finally, the system would benefit from outcome-based contracting approaches to reward products which improve performance.

Note 1

Dr. Neumann served as a member and co-chair.

Author’s Note

Funding support was provided by Pfizer, Inc, to Precision Health Economics (PHE). In addition to his positions as Director of the Center for the Evaluation of Value and Risk in Health at the Institute for Clinical Research and Health Policy Studies at Tufts Medical Center, and Professor of Medicine at Tufts University School of Medicine, Dr. Neumann is a consultant to PHE.