Precision benefit design is an extension of the concept of value-based insurance design (VBID), which aims to link patient benefit design and coinsurance to the likely benefit of a treatment net of its cost. Whereas VBID measures the value of a treatment for the average patient, the advent of precision medicine raises the possibility of implementing a VBID-style approach but measuring a treatment’s value—and concordant benefit design—based on the specific patient genotype information. While precision benefit design is conceptually appealing, to implement this approach in practice would require overcoming at least 3 key challenges.
Advances in precision medicine offer the promise of tremendous gains in patient health over the coming decades. The US National Library of Medicine and the Precision Medicine Initiative define precision medicine as a new approach for “disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person.”1 Headlines in the popular press as well as discussions at academic conferences tout the potential benefits of precision medicine, while at the same time critiquing the nascent but growing evidence base and the high cost of treatment.2 For instance, the average cost of many new precision treatments, especially oncology treatments, routinely exceeds $100,000 per year.3 As a result, decision makers are increasingly faced with challenging choices, leading to calls for better incorporation of economic analysis into how precision medicines are evaluated and covered,4 especially in oncology.5 There are multiple initiatives already underway to address this challenge.6
One approach to address the cost/access issues, as proposed by Dr Fendrick and Ms Shope, is to implement precision benefit design. If a treatment is only indicated for patients with specific biomarkers, under a precision benefit design patients without those biomarkers may be denied coverage for the treatment. In other cases, precision benefit design’s implementation may not be so black and white. New immuno-oncology treatments often work better for patients with higher levels of the programmed death-ligand 1 (PD-L1) expression on cancer cells7; in these cases, precision benefit design could change a patient’s copayment depending on their PD-L1 levels.
More broadly, precision benefit design is an extension of the concept of VBID, which aims to link patient benefit design and coinsurance to the likely benefit of a treatment net of its cost.8 While VBID has not been used extensively to date, the Centers for Medicare and Medicaid Services are piloting a VBID model among Medicare Advantage plans. Whereas VBID measures the value of a treatment for the average patient, the advent of precision medicine raises the possibility of implementing a VBID-style approach but measuring a treatment’s value—and concordant benefit design—based on the specific patient genotype information.
While precision benefit design is conceptually appealing, to implement this approach in practice would require overcoming at least 3 key challenges: (1) ensuring formulary design is patient-centered; (2) ensuring that value is measured using scientific best practices; and (3) overcoming informational barriers.
Challenge 1: Patient-Centricity
For precision benefit design to truly improve the value of care patients receive, formulary design must be patient-centered. In its simplest form, the concept of a precision benefit design works well under the assumption that either (a) patients care only about a single outcome of interest, or (b) that precision medicines are always blockbuster treatments that are superior to existing treatments across all or most outcomes that matter to patients (eg, efficacy, safety, ease of administration, cost).
Neither of these assumptions typically hold in the real world. In practice, patient preferences are heterogeneous and vary across a multidimensional outcome space. Heterogeneity in patient preferences makes measuring the treatment value using standard cost-effectiveness analyses problematic.9 Moreover, many (although not all) patients want to be involved in shared decision-making with their physicians,10 and precision benefit design should enhance, rather than limit, these collaborative decisions.
The issue is further complicated by the fact that there is rarely a single treatment that dominates all others across all dimensions. For instance, an examination of cancer treatments across tumor types found that, when making treatment selections solely based on efficacy, the preferred treatment varied depending on whether efficacy was defined based on a median overall survival, mean overall survival, 1-year survival rate, or number needed to treat.11 When including costs and other treatment attributes of importance to patients in the decision problem, this multidimensionality further complicates the issue. Thus, precision benefit design should take into account patient preferences, in addition to more traditional cost-effectiveness comparisons, to ensure each patient gets the treatment of highest value to them.
Challenge 2: Staying Aligned With the Latest Science
A second challenge is using the latest scientific evidence to determine what “value” means. For instance, payers typically only measure estimates of treatment-related cost and clinical benefit to the patient themselves, but treatments may also have broader costs and benefits. Treatments may affect patients’ ability to return to work and care for family, for example, or impact potential caregiver burden in the case of debilitating diseases. Further, and especially in the context of cancer, using median survival as the primary efficacy outcome may miss the fact that patients place a high value, not just on outcomes for typical patients, but on the low-probability but high-reward possibility of having an above-average survival benefit (ie, being in the right tail of the survival distribution).12 This concept is often referred to as the “value of hope,” and even the American Society for Clinical Oncology has included survival in the tail of the distribution within their updated value framework.13 Precision medicine treatments that extend survival may also provide an “option value” in that patients may live long enough to receive and benefit from the next treatment innovation.14 Various leading scientific bodies have proposed including these broader societal and patient-centered value components in cost-effectiveness analysis,15,16 and some studies have found that incorporating broader societal measures of value can have a significant effect on a treatment’s incremental cost-effectiveness ratio, particularly in oncology.17
In addition to which costs and benefits to include in any assessment of value, one should also consider the quality of the evidence underlying these value components. For example, most oncology clinical trials are of a relatively short duration and may measure surrogate endpoints. Thus, how a treatment affects patient survival beyond the trial’s duration is a key modelling assumption that affects a treatment’s estimated value. Further, it is not always clear how survival measured using surrogate endpoints (eg, progression-free survival, complete response rate) correspond to overall survival gains in the real world.18
A treatment’s value may also depend on the timing of when it is received and the patient preferences over this timing. Novel oncology treatments often initially rely on clinical trial evidence from later lines of therapy before moving to newer lines of therapy as additional clinical trials read out their results. Thus, patients and providers often face 2 options. One option would be to use the novel therapy in first line. However, if that treatment does not work for that individual patient, there will be little evidence of the effectiveness of older treatments after the use of the novel treatment. On the other hand, patients could use the older, less effective treatment first and reserve the newer treatment for second line. Although this older treatment will be less efficacious on average, those for whom treatment fails will have a viable option in second line, whereas patients who choose the more efficacious therapy may not. Precision benefit design would need to measure value based not only on individual treatments but on treatment sequences and allow patient preferences to inform treatment order in cases where there is scientific uncertainty.
Challenge 3: Accessing Needed Information
While there may be various forms of precision benefit design, tailoring benefits and coverage policy based on a patient’s genotype is only viable if information costs are sufficiently low for payers, physicians, and patients. For payers, a precision benefit design requires (at a minimum) tailoring benefit designs (ie, coverage policies and cost-sharing arrangements) based on the patient’s genetic information and lab results. While providers inevitably need to conduct lab tests to identify which patients are good candidates for precision medicines, the results of these lab tests may not be rapidly available to payers. One option would be to require that patients receive genetic tests before authorizing the use of a medication. Another option would be to integrate lab results into payer databases and to translate these results into cost sharing and access policies in near real time. Both of these approaches, especially the latter, could be administratively burdensome to payers and could adversely affect patient health outcomes if needed treatments were delayed due to bureaucratic approval processes.
Further, precision benefit design may impose additional expenses on providers. Providers may advise patients on the best available treatment but also need to advise patients on the relevant cost of these treatments. Precision benefit design adds a layer of complexity to the provider-patient decision-making process. Further, most patient’s current benefit design does not change over the course of a year, but, with full implementation of precision benefit design, a patient’s coverage may change based on the results of their genetic tests. While the dynamic approach is attractive from a value perspective, it does make it challenging for providers who advise patients on treatment options and seek reimbursement for physician-administered drugs. Physician costs for time spent interacting with payers is already 4 times as high in the United States as in Canada,19 and precision benefit design in any form could increase providers’ administrative costs even more.
Precision Benefit Design Requires Precision Value Assessment
To implement precision benefit design in the most rigorous manner, payers need visibility into the true value of these precision medicines. Value assessment should follow best practices from organizations such as the National Health Council,20 Second Panel on Cost Effectiveness in Health and Medicine,16 the International Society for Pharmacoeconomics and Outcomes Research,15 and other leading bodies.
First, payers need to understand what patients want from their treatments. To achieve this, payers should conduct patient focus groups, interviews, or surveys to get to the patient perspective on factors that matter to their treatment experience. Moreover, payers should work closely with patient organizations and hold discussions with the clinicians treating these patients as well as researchers.
Second, payers should begin by conducting a standard cost-effectiveness analysis. The cost-effectiveness analysis should follow best practices from leading scientific bodies,15,16 consider how treatment order affects relative efficacy, and include broader societal costs and benefits. The goal of this step is to determine whether—by traditional metrics—a treatment represents a high value.
Third, and based on the understanding of patients’ lived experience, payers should consider employing alternative methods, such as multi-criteria decision analysis (MCDA),21 to examine how different patient preferences affect estimated treatment value. Implementing a precision benefit design that helps to get high-value treatments to the average patient is a laudable goal, but a better goal would be to ensure that each patient receives a treatment that provides high value to them. MCDA can help quantitatively inform which treatments may provide the highest value to a specific patient, whose unique preferences were identified in step 1.
Fourth, precision value assessment must be updated over time. As new treatments enter the market and new evidence on existing treatments emerge, the estimated value of these new treatments may change. Thus, payers will need to periodically revisit their precision benefit design based on this new evidence.
Finally, payers must consider the administrative cost of implementing precision benefit design and regularly updating their precision value assessment as new evidence emerges. With the recent advances in information technology, the cost to design and administer precision benefit design may fall and could soon become a viable option.
Conclusion
Precision value assessment can help use the latest clinical and real-world evidence, information on patient preferences, and novel viewpoints on value to help payers create more flexible and customizable benefit design that can match appropriate patients more quickly with affordable, efficacious treatment options.
While precision benefit design is conceptually appealing, it must also account for flexibility in treatment response, patient heterogeneity and long-term evidence development about treatment efficacy to ensure appropriate access to such therapies based on disease progression, failed first- and second-line therapies, and new evidence. Ultimately, precision medicine and precision benefit design must have, at its heart, the goal of improving the lives of those to whom it matters most: the patients.
To read the initial viewpoint this article references, click here.
References
1. Genetics Home Reference, US National Library of Medicine, US Department of Health & Human Services. What is precision medicine? ghr.nlm.nih.gov website. https://ghr.nlm.nih.gov/primer/precisionmedicine/definition. Published August 14, 2018. Accessed August 16, 2018.
2. Cohen JP. Taking a wider view of precision oncology. Forbes. https://www.forbes.com/sites/joshuacohen/2018/08/02/taking-a-wider-view-of-precision-oncology/#345d1aef2022. Published August 2, 2018. Accessed August 16, 2018.
3. Mailankody S, Prasad V. Five years of cancer drug approvals: innovation, efficacy, and costs. JAMA Oncol. 2015;1(4):539-540.
4. Stern AD, Alexander BM, Chandra A. How economics can shape precision medicines. Science. 2017;355(6330):1131-1133.
5. Lakdawalla D, Shafrin J, Lucarelli C, Nicholson S, Khan ZM, Philipson TJ. Quality-adjusted cost of care: a meaningful way to measure growth in innovation cost versus the value of health gains. Health Aff (Millwood). 2015;34(4): 555-561.
6. Berndt ER, Goldman D, Rowe J. Introduction. In: Lu JR, Eggleston K, Tung-Chieh Chang J, eds. Economic Dimensions of Personalized and Precision Medicine. Chicago, IL: University of Chicago Press; 2018.
7. Carbognin L, Pilotto S, Milella M, et al. Differential activity of nivolumab, pembrolizumab and MPDL3280A according to the tumor expression of programmed death-ligand-1 (PD-L1): sensitivity analysis of trials in melanoma, lung and genitourinary cancers. PLoS One. 2015;10(6):e0130142.
8. Chernew ME, Rosen AB, Fendrick AM. Value-based insurance design. Health Aff (Millwood). 2007;26(2):w195-w203.
9. Kravitz RL, Duan N, Braslow J. Evidence-based medicine, heterogeneity of treatment effects, and the trouble with averages. Milbank Q. 2004;82(4):661-687.
10. Degner LF, Kristjanson LJ, Bowman D, et al. Information needs and decisional preferences in women with breast cancer. JAMA. 1997;277(18):1485-1492.
11. Karweit J, Kotapati S, Wagner S, Shaw JW, Wolfe SW, Abernethy AP. An expanded portfolio of survival metrics for assessing anticancer agents. Am J Manag Care. 2017;23(1):54-61.
12. Shafrin J, Schwartz TT, Okoro T, Romley JA. Patient versus physician valuation of durable survival gains: implications for value framework assessments. Value Health. 2017;20(2):217-223.
13. Schnipper LE, Davidson NE, Wollins DS, et al. Updating the American Society of Clinical Oncology value framework: revisions and reflections in response to comments received. J Clin Oncol. 2016;34(24):2925-2934.
14. Thornton Snider J, Batt K, Wu Y, Tebeka MG, Seabury S. The option value of innovative treatments for non-small cell lung cancer and renal cell carcinoma. Am J Manag Care. 2017;23(10):e340-e346.
15. Lakdawalla DN, Doshi JA, GArrison LP Jr, Phelps CE, Basu A, Danzon PM. Defining elements of value in health care—a health economics approach: an ISPOR Special Task Force report [3]. Value Health. 2018;21(2):131-139.
16. Sanders GD, Neumann PJ, Basu A, et al. Recommendations for conduct, methodological practices, and reporting of cost-effectiveness analyses: second panel on cost-effectiveness in health and medicine. JAMA. 2016;316(10):1093-1103.
17. Shafrin J, Skornicki M, Brauer M, et al. An exploratory case study of the impact of expanding cost-effectiveness analysis for second-line nivolumab for patients with squamous non-small cell lung cancer in Canada: Does it make a difference? Health Policy. 2018;122(6):607-613.
18. Lakdawalla DN, Shafrin J, Hou N, et al. Predicting real-world effectiveness of cancer therapies using overall survival and progression-free survival from clinical trials: empirical evidence for the ASCO Value Framework. Value Health. 2017;20(7):866-875.
19. Morra D, Nicholson S, Levinson W, Gans DN, Hammons T, Casalino LP. US physician practices versus Canadians: spending nearly four times as much money interacting with payers. Health Aff (Millwood). 2011;30(8):1443-1450.
20. National Health Council. The Patient Voice in Value: The National Health Council Patient-Centered Value Model Rubric [press release]. National Health Council website. https://www.nationalhealthcouncil.org/patient-centered-value-model-rubric-released. Published March 28, 2016. Accessed August 16, 2018.
21. Baltussen R, Niessen L. Priority setting of health interventions: the need for multi-criteria decision analysis. Cost Eff Resour Alloc. 2006;4(1):14.
