What’s the trade-off between comparative effectiveness and cost–effectiveness analysis?

The US healthcare system, with its multipayer health insurance, is the most expensive system in terms of healthcare spending as a fraction of gross domestic product and per capita spending, while it leaves 47 million Americans without health insurance to bear the subsequent impact: the life expectancy of Americans was ranked 42nd in the world [1–3]. On the other hand, countries such as the UK, Canada, Australia and Germany, all with a single national payer system, have been constantly battered with the difficulty of balancing the costs of healthcare and equitable access to advanced medical technology, such as surgeries and pharmaceuticals. With the aging of the global population, rising healthcare costs and the rapid development of expensive medical technology, all healthcare systems are faced with the challenge to provide rationales for the continuation and expansion of current health services.

Cost–effectiveness analysis (CEA) has been used as a guide to resource allocation in health, by comparing the incremental costs with the incremental benefits associated with new health services compared with existing ones or reference cases [4,5]. CEA has been increasingly common in medical and health literature in the past decade. More recently, comparative effectiveness (CE) has been suggested as a separate method to guide the allocation of resources [6,7]. CE is focused on the relative effectiveness of clinical treatments: instead of comparing the new treatment versus the placebo (as is done in many applications for new drugs to the US FDA in the USA), CE requires the comparison of the new technology/drugs to the existing or prevailing one. While CE has been proposed as a separate guideline from the CEA, there will inevitably be competing interests and emphases placed on them by the stakeholders. As a result, the trade-off between CE and CEA must be carefully explored to ascertain the potential consequences for possible misuse.

Costs of costing information

The cost component in CEA provides additional information on the incremental costs per unit increase in effectiveness, which is not present in CE. However, the additional information on costs is not free of costs. One particular consequence of the inclusion of costs in analysis is the increased uncertainty in information. Direct and indirect medical costs are notoriously difficult to measure, and the price index in medical care was unreliable in measuring the overall healthcare spending [8]. When we subtract one uncertain cost from the other uncertain one, and then divide it by the difference between another two uncertain numbers in medical effectiveness, the final results (i.e., the incremental cost–effectiveness ratio [ICER]) often have wide varying ranges even for the same treatment across studies performed by reputable researchers. While decision-analysis techniques, such as the Markovian approach, can provide a confidence interval for ICER through simulations, it cannot diversify away the systematic uncertainty involved in cost estimation. In this sense, separating clinical effectiveness from CEA may provide an attractive alternative to many payers and agencies who are not certain about the usefulness of results derived from CEA.

Perspective matters

There is little controversy surrounding perspectives in clinical effectiveness as a whole. However, this is not the case for costs. For a single-payer system, the perspective of costs is more clear cut, as the payer (usually the national health provider or the national health insurance program) is the most relevant party whose perspective should be adopted. Since it is a national program, the perspective is more aligned with the societal perspective that the economics literature in health and medicine uniformly advocates. However, it is less clear cut in a multipayer system such as the US system. Naturally, the costs incurred by the payer are most pertinent to the payer, but this is not necessarily aligned with the societal perspective. For example, a gain made by a managed care company through excluding certain coverage can be a cost to the society, because the society has to pay the social welfare costs of caring for the untreated patients, and such societal costs are not accounted for in CEA considered from the perspective of the managed care company. There is no inherent incentive for managed care companies to align their services to the societal perspective, which makes the comparability of CEA a disastrous event when all the results are gathered together in a league. Again, due to the inherent characteristics of costing, CE presents a more comparable case to payers across public and private sectors, as well as national boundaries.

Does the market work?

If the market works perfectly, once the information on CE is available, the resource will be allocated or re-allocated by the most efficient use of the new technology. This is not necessarily the case in healthcare, because consumers are often unaware of the true costs of healthcare as they are only faced with a fraction of costs through deductible payments or copayments in the presence of health insurance. In addition, consumers are less informed about the clinical effectiveness of treatments than service providers, which creates a problem when the provider has an incentive to induce demand to increase its profit margin. Therefore, the omission of the cost components in CE in resource allocation is likely to result in allocations that are blinded to the budgetary impact of the inclusion of the services. As discussed earlier, the perspective from a private payer will not be necessarily aligned with the best interest of the society. With the distortion of price due to insurance, there will be no guarantee that cost-effective strategies by the private payer would be similarly cost-effective to the society.

On balance, we are left with a paradoxical paradigm: while CE has clear advantages in uniformity and comparability of information, it is lacking a key component of a societal/national perspective, without which the private market cannot allocate resources in the best interest of society. Methodologically, CE and CEA can be viewed as two separate guidelines. However, like the tiered formulary in prescription drug plans, we may need to consider the tiered formulary in political decision-making in healthcare: with CE as the cornerstone, CEA can be built upon with more copayments from government funding.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.