Value-based insurance design for prescription drug coverage increases drug adherence in patients with chronic disease, though their effect on clinical outcomes and health spending remain uncertain.
Value-based insurance design (V-BID) is an insurance cost-sharing model in which patients pay less for medications deemed to be of higher value. Our objective was to determine the association between V-BID and medication adherence, clinical outcomes, healthcare utilization, and spending in patients with or at risk for cardiovascular chronic diseases, compared with no differential lowering of drug co-payments.
We searched PubMed, MEDLINE, EMBASE, CINAHL, Cochrane Controlled Trials Register, Current Controlled Trials, and reference lists of included studies and relevant reviews up to September 2012. Two reviewers independently identified primary research studies with the following study designs: randomized controlled trial, interrupted time series, and controlled before-after studies. Two reviewers independently extracted data and assessed quality.
Ten studies were identified: 1 high-quality randomized controlled trial, 1 interrupted time series analysis, and 8 controlled before-and-after studies. Heterogeneity in study populations and interventions, overall low study quality, and lack of standard error reporting precluded meta-analysis. All reported improvement in medication adherence for medications subject to V-BID, of between 2 and 5 percentage points. Impact on clinical outcomes was unclear, with only 1 study reporting on this, noting no difference in the primary outcome, but a reduction in adverse secondary outcomes with V-BID. Of the four studies that examined the impact of VBID on healthcare expenditures, V-BID tended to increase overall prescription drug spending, though three of the four studies reported similar overall healthcare costs due to decreased non drug medical spending.
V-BID is associated with improved medication adherence but its effects on clinical outcomes, healthcare utilization, and spending remain uncertain.
Am J Manag Care. 2014;20(6):e229-e241
Prescription drug costs comprise 10% of all healthcare expenditures in the United States,1 reaching $263 billion in 2011, and they are projected to rise to nearly $500 billion annually over the next 10 years.2 Patient co-payments and/or deductibles were established in part to lower costs to the insurer and to reduce inappropriate overconsumption. The amount of cost sharing has traditionally been based on drug acquisition cost, where expensive brand name medications have higher levels of co-payments than less expensive generic drugs.3 Past studies have shown that patients may reduce drug use in response to cost sharing, but tend to reduce both appropriate and inappropriate therapy,4,5 indicating that they may not be able to differentiate high- from low-value therapies.6 This leads to increased medical complications and healthcare utilization4,7 and higher overall healthcare costs.
Fendrick and colleagues introduced a novel cost sharing design in 2001, initially termed “benefit-based co-pay,”8 and subsequently renamed “value-based insurance design” (VBID). 3 The underlying premise is that a drug co-payment is based on the clinical benefit, or value, of that drug rather than its acquisition cost. More specifically, Fendrick and colleagues define V-BID as “decreasing cost sharing for interventions that are known to be effective and increasing cost sharing for those that are not.”9 The goal of the V-BID system is to encourage the use of the most effective drugs, while constraining costs by keeping cost sharing intact for drugs deemed to be of lower value.6
Since 2001, numerous private insurers have implemented V-BID for prescription drug coverage. We conducted a systematic review to determine the impact of V-BID on medication adherence, clinical outcomes, and health-system costs in patients with cardiovascular-related chronic diseases, compared with those with usual drug coverage. We defined V-BID as the selective lowering or waiving of drug co-payments for medications that were deemed to be of high value in the treatment of chronic diseases. Our search focused on cardiovascular (CV)-related chronic diseases, given that long-term medication use (in addition to lifestyle changes) is the mainstay of treatment for these conditions, and that a large body of evidence shows that selected medications are effective in reducing morbidity and mortality. For example, treatment of hypertension reduces the risk of major CV events, with similar benefit seen across many antihypertensive drug classes used.10 Similarly, intensive glycemic control in diabetes with oral hypoglycemic agents or insulin reduces complications of retinopathy and nephropathy.11 V-BID is one method by which evidence-based appropriate treatment can be promoted for these diseases.
Data Sources and Searches
A study protocol was created that outlined the objectives,eligibility criteria for study inclusion in the systematic review, and plan for data abstraction, synthesis, and quality assessment. We searched the following databases from inception to September 12, 2012: PubMed, MEDLINE, EMBASE, CINAHL, Cochrane Controlled Trials Register, and Current Controlled Trials. Key terms included hypertension, cardiovascular diseases, stroke, diabetes mellitus, dyslipidemia, cost sharing, deductibles and coinsurance, co-pay, and value-based pricing or plan. The search strategy can be found in the eAppendix. The search was limited to English and French language studies, and no limitations were placed on patient characteristics, study duration, or outcome. Reference lists of all included studies and relevant narrative reviews were manually searched.
Using a 2-step process, 2 reviewers (KLT and LB) first independently reviewed all identified abstracts for eligibility, then extracted and reviewed the full-text articles for the abstracts that were thought to meet eligibility criteria, or where there was any uncertainty. Studies published exclusively as abstracts without accompanying full-text articles were not considered. Disagreements were resolved by consensus or with the aid of a third party (BJM). Eligibility criteria were the implementation of V-BID as described above, inclusion of at least a proportion of adults with 1 or more CV-related chronic disease (hypertension, dyslipidemia, diabetes, coronary artery disease, heart failure, or stroke), and 1 or more outcome of interest (medication adherence, clinical outcomes, healthcare utilization, or costs). We included randomized controlled trials (RCTs), nonrandomized controlled trials, interrupted time series (ITS) analyses, and controlled before-and-after studies, based on the Cochrane Effective Practice and Organization of Care Group [EPOC] taxonomy of healthcare policy studies.12 Studies were excluded if the patient populations were exclusively children or adolescents.
Data Extraction and Quality Assessment
Data were extracted on study characteristics (including funding, inclusion/exclusion criteria, design, and methods), baseline characteristics of the intervention and comparator groups, and primary and secondary outcomes for each study. Two reviewers independently extracted data from all studies fulfilling eligibility criteria; disagreements were resolved by consensus. The quality of included studies was assessed using the Cochrane Collaboration risk of bias tools.12
Data Synthesis and Analysis
The significant heterogeneity in study populations and interventions across studies, the overall low quality of the studies, and lack of reported standard error or standard deviation for outcome measurements precluded pooling of results in a meta-analysis. Furthermore, though adherence was reported for all studies, definitions varied slightly among them. Results for each study are summarized individually.
Our search identified 2186 citations, of which 43, along with an additional 4 identified through manual review of reference lists, were included for full-text review (Figure). Of those, 10 studies evaluating 9 V-BID interventions met inclusion criteria: 1 randomized controlled trial,13 1 interrupted time series study,14 and 8 controlled before-and-after studies15-22 (Table 1). All studies were completed in the United States. Of the 10 studies, 2 studies13,14 selectively lowered or waived co-payments for CV medications only (antihypertensives and statins); 3 studies15,19,22 waived co-payments for diabetic medications only, and 5 studies16-18,20,21 waived co-payments for a combination of both CV and diabetic medications. Two studies19,22 waived copayments for brand name drugs only, while the other 8 studies reduced cost-sharing for both generic and brand name drugs.
Six studies14,16,17,19-21 included disease management cointerventions. Five of these14,16,17,19,20 introduced V-BID alongside a separate voluntary disease management programa while 1 study21 tied the V-BID intervention to the disease management program; refusal to participate in disease management in this study also precluded patients from participating in V-BID, making it impossible to distinguish whether outcomes could be attributed to V-BID or the disease management intervention in this study. In 3 of the 6 studies14,16,17,20 with disease management co-interventions, both the intervention and comparison groups could voluntarily participate in disease management programs. Therefore, disease management participation is expected to influence outcomes only if participation rates differed between these 2 groups. Two of these 3 studies16,17 reported similar disease management participation rates in the V-BID and comparison groups, while the third study14 did not report participation rates at all. Four of the 6 studies14,16,17,20 with disease management co-interventions reported only overall results and did not report separate outcomes for participants of disease management programs. A total of 2 studies15,22 ensured that participating insurance plans in the intervention and control groups did not administer disease management programs simultaneously; and 2 studies13,18 did not mention the presence or absence of concurrent disease management programs at all.
Quality of Studies
The only randomized controlled trial13 was rated high quality and the interrupted time series study14 was rated moderate quality. The remainder of the studies, all controlled before-and-after studies, were considered poor quality (Table 2) primarily because of their study designs and their risk of confounding due to concurrent disease management programs for which outcomes were not adjusted.14,16,21
Five studies14,16-18,20 reported significant differences in baseline characteristics (such as age, gender, and baseline medication co-payments) between the intervention and control groups, which were partly addressed in 3 of the 5 studies17,18,20 through propensity score matching, though this did not entirely eliminate differences across groups in all studies.18,20 In another study,21 though the baseline characteristics were similar among groups, the control group was comprised of patients who declined to participate in disease management programs (and therefore were ineligible for V-BID), while the intervention group included patients who agreed to participate in (and therefore also received) V-BID. Five studies15,16,19-21 also reported statistically significant differences in baseline medication adherence between the intervention and control groups.
Impact on Medication Adherence
All 10 studies reported the impact of V-BID implementation on medication adherence, either on its own (n = 4) or when given in conjunction with disease management (n = 6) (Table 3). Baseline adherence (ie, medication possession ratios) across studies was generally in the range of 60% to 80%, though 2 studies13,19 reported much lower baseline adherence rates of between 30% and 40%. Overall, V-BID was associated with an increase in medication adherence, as measured by medication possession ratio or proportion of days covered, of approximately 2 to 5 percentage points compared with the control group for medications subject to V-BID. In the randomized controlled trial,13 the increase was 4.4 to 6.2 percentage points across the medications subject to V-BID. Similar outcomes were seen across all studies, with follow-up periods ranging from 3 months13 to 3 years post V-BID implementation.19,20 Choudhry and colleagues’ interrupted time series study14 showed that following an immediate improvement in medication adherence in the V-BID group, the rate of decline in adherence was similar between the 2 groups over time.
Absolute increases in medication adherence were also similar between studies that evaluated V-BID only and those that also implemented voluntary disease management programs (Table 3). In the 2 studies19,21 that reported on the combined effect of disease management and V-BID, medication adherence improved by 2.7 to 6.5 percentage points in patients who received both V-BID and disease management programs. Kim and colleagues21 added that improvement was seen only if the disease management program was intensive in nature, such as with telephone counseling with a nurse to set goals and care plans, but not when the program consisted only of health education mailings.
Impact on Clinical Outcomes
Only the 1 randomized controlled trial13 examined the impact of essential drug coverage on clinical outcomes. The trial showed that in patients discharged from the hospital after a myocardial infarction, the rate of first major vascular event or revascularization was similar between those with full drug coverage for essential CV medications (angiotensin converting enzyme inhibitors, beta-blockers, and statins) and those with “usual drug coverage” (mean co-payments for the medications of interest ranging between $12.83 and $24.92 depending on the class). However, rates of total major vascular events or revascularization (hazard ratio [HR] 0.89; 95% CI, 0.80-0.99), and first major vascular event (HR 0.86, 95% CI, 0.74-0.99) were significantly lower in the full drug coverage group compared with the usual coverage group.
Impact on Health Utilization and Expenditures
One controlled before-and-after study21 examined changes to healthcare utilization before and after implementation of V-BID, but showed conflicting results, with a reduction in hospitalizations in the high-risk group who participated in V-BID and an intensive disease management program, but an increase in hospitalizations in the low-risk group participating in V-BID and a less intensive disease management program when compared with controls who chose not to participate in either program.
Three of the 4 studies13,19,21 that examined the impact of V-BID on healthcare expenditures found an increase in prescription drug expenditures overall (Table 4), while the fourth study found no statistically significant difference. 20 As expected, patient-borne prescription drug costs decreased (relative spending 0.70, 95% CI, 0.65-0.75) with V-BID implementation in the randomized trial.13 Total healthcare spending, including both drug and non-drug expenditures, was similar in the V-BID group compared with the usual care group in 3 studies,13,19,20 including the randomized trial.13 In the fourth study,21 total healthcare spending was lower in those patients who participated in V-BID and intensive disease management, but higher in the V-BID and less intensive disease management group compared with those who did not participate in either the V-BID or disease management programs.
To our knowledge, this is the first systematic review to examine V-BID. We found 10 studies that evaluated 9 interventions which compared V-BID with no differential lowering of drug co-payments in patients with CV-related chronic diseases. Although V-BID was consistently associated with an increase in medication adherence of 2 to 5 percentage points across all studies, including the sole randomized trial, the evidence of impact of V-BID on clinical outcomes was far more limited, with only 1 study13 evaluating this. Though this study showed no difference in the primary clinical outcome, there was a decrease in adverse clinical secondary outcomes. Furthermore, the combined role of disease management and V-BID is unclear. Only 2 low-quality studies, at high risk of residual confounding21 and selection bias,19,21 reported outcomes separately for patients participating in V-BID and disease management compared with those who chose not to participate in either, allowing assessment of the impact of combining these programs.
Our review reveals a substantial evidence gap on the impact of V-BID on clinical outcomes, health utilization, and healthcare spending, consistent with prior narrative reviews by Choudhry and colleagues23 and Fairman and Curtiss.24 One possible reason that V-BID has not resulted in overall cost savings is that an increase in co-payments for low-value medications has not yet been applied together with a decrease in co-payments for high-value medications, as described in the original V-BID model.8 As a result, any cost savings with V-BID interventions depend on improved medication adherence leading to improved clinical outcomes and resulting in decreased healthcare utilization and medical spending. Melnick and Motheral,25 using a plausibility calculation method, argued that net cost savings with V-BID is highly unlikely, as large reductions in drug co-payments result only in small increases in medication use (low price elasticity), and that avoidable adverse events due to improved medication adherence are rare.
Despite the limited evidence, approximately 20% of private insurance plans offered by large American employers include V-BID.24 Moreover, in 2007, 80% of large employers indicated an interest in implementing V-BID over the ensuing 5 years.23 There has also been substantial US governmental interest and political activity regarding V-BID, and in 2009, the Seniors’ Medication Copayment Reduction Act was created, requesting a demonstration program to test V-BID in Medicare beneficiaries.26 Most recently, the Patient Protection and Affordable Care Act, under Section 2713(a), specifically addresses V-BID, stating “The Secretary may develop guidelines to permit a group health plan and a health insurance issuer offering group or individual health insurance coverage to utilize value-based insurance designs.”27 V-BID need not be limited to drug insurance alone, and could potentially be applied to other health services. In 2010, new regulations in the US required private health insurance plans to cover high-value preventive services that were given a rating of grade A or B by the United States Preventive Services Task Force. This includes breast and colon cancer screening, diabetes screening, and routine vaccinations.28 There has also been interest in applying VBID to the areas of diagnostic imaging,29 gastroenterologic procedures such as endoscopy,30 and oncology.31 However, actual implementation and subsequent evaluation of VBID has yet to extend beyond prescription medications. The limitations in evidence do not seem to substantiate the widespread interest and implementation of V-BID and should be considered experimental.
There were limitations to our review. First, our review evaluated only the impact of V-BID and not whether medications that had their co-payments reduced were correctly classified as “high value” in their respective V-BID programs. As with any systematic review, our study was limited by the quality of the underlying studies. Except for the 1 randomized trial, the literature base in V-BID at the time of our study had a moderate to high risk of bias, specifically in study design (in particular with selection bias) and reporting.24,32,33 Our focus on CV-related chronic diseases may limit generalizability to other conditions, though medications for these diseases constitute a large proportion of the medications funded by outpatient prescription drug benefit plans. In addition, private insurers implemented V-BID in 9 of the 10 studies, potentially also limiting generalizability of the findings, especially with respect to public drug insurers. Lastly, current studies in V-BID target populations that already have good medication adherence, and this may limit the ability of V-BID to increase compliance.25 A notable exception was the randomized trial, 13 with a baseline adherence of about 44%, likely because a variety of insurance plan sponsors were included—employer, union, and government insurers—thereby better representing the general population. Further research on the impact of V-BID on clinical and economic outcomes is required, particularly in populations that might benefit most,23 such as those at highest risk for clinical adverse events, those with low baseline compliance, and those facing financial barriers to drug adherence.
If V-BID is expected to lower costs, consideration must be given to increasing cost sharing for low-value medications. The major challenge is in classifying and defining a medication as low value, given that the evidence for services or medications being of low value is far less established than for high-value medications.34,35 Additional research into this, as well as the most appropriate patient populations to target, is required to inform further use of V-BID,36 particularly within publicly funded drug formularies.
Value-based insurance design is a novel approach to encourage adherence to high-value medications. Though it appears to be associated with improved medication adherence, the effects on clinical outcomes and overall health utilization and expenditures remain uncertain. Further high-quality research is required before more widespread implementation of V-BID can be encouraged.
Dr Campbell was supported by an Alberta Innovates-Health Solutions (AI-HS) Clinician Fellowship award. Dr Barnieh was supported by an Alberta Innovates-Health Solutions Trainee award. Drs Manns and Hemmelgarn were supported by career salary support awards from Alberta Innovates-Health Solutions. Dr Hemmelgarn was also supported by the Roy and Vi Baay Chair in Kidney Research. Dr Tonelli was supported by a Government of Canada Research Chair. Drs Manns, Hemmelgarn, and Tonelli were also supported by an alternative funding plan from the Government of Alberta and the Universities of Calgary and Alberta.
aDisease management programs consisting of targeted health education mailing, disease workbooks, telephone outreach, and counseling with a nurse to set care plans, periodic monitoring, or any combination of the above.Author Affiliations: Department of Medicine, University of Calgary, Alberta, Canada (KLT, LB, BM, DJTC, BRH, BJM); Department of Community Health Sciences, University of Calgary, Alberta, Canada (LB, FC, DJTC, BRH, DL, BJM); Interdisciplinary Chronic Disease Collaboration, Calgary, Alberta, Canada (LB, FC, DJTC, BRH, MT, BJM); Institute of Public Health, University of Calgary, Alberta, Canada (FC, BRH, BJM); The Libin Cardiovascular Institute, University of Calgary, Alberta, Canada (BRH, BJM); Department of Medicine, University of Alberta, Edmonton, Alberta, Canada (MT).
Source of Funding: This research was supported by an interdisciplinary team grant from Alberta Innovates-Health Solutions, the Interdisciplinary Chronic Disease Collaboration.
Author Disclosures: The authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article. The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.
Authorship Information: Concept and design (KLT, LB, BM, FC, DJTC, BRH, MT, DL, BJM); acquisition of data (KLT, LB, BM, FC, DJTC, BRH, MT, DL, BJM); analysis and interpretation of data (KLT, LB, BM, FC, DJTC, BJM); drafting of the manuscript (KLT); critical revision of the manuscript for important intellectual content (KLT, LB, BM, FC, DJTC, BRH, MT, DL, BJM).
Address correspondence to: Braden J. Manns, MD, MSc, Foothills Medical Centre, 1403 29th St NW, Calgary, AB T2N 2T9. E-mail: firstname.lastname@example.org.REFERENCES
1. Health Expenditures. CDC website. http://www.cdc.gov/nchs/fastats/hexpense.htm. Published 2012. Accessed March 18, 2013.
2. National Health Expenditures Projections 2011-2021. CDC website. https://http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/NationalHealthExpendData/NationalHealthAccountsProjected.html. Published 2012. Accessed March 18, 2013.
3. Fendrick AM, Chernew ME. Value-based insurance design: a “clinically sensitive” approach to preserve quality of care and contain costs. Am J Manag Care. 2006;12(1):18-20.
4. Goldman DP, Joyce GF, Zheng Y. Prescription drug cost sharing: associations with medication and medical utilization and spending and health. JAMA. 2007;298(1):61-69.
5. Lohr KN, Brook RH, Kamberg CJ, et al. Use of medical care in the Rand Health Insurance Experiment. Diagnosis- and service-specific analyses in a randomized controlled trial. Med Care. 1986;24(9 suppl):S1-S87.
6. Chernew ME, Rosen AB, Fendrick AM. Value-based insurance design. Health Aff (Millwood). 2007;26(2):w195-w203.
7. Tamblyn R, Laprise R, Hanley JA, et al. Adverse events associated with prescription drug cost-sharing among poor and elderly persons. JAMA. 2001;285(4):421-429.
8. Fendrick AM, Smith DG, Chernew ME, Shah SN. A benefit-based co-pay for prescription drugs: patient contribution based on total benefits, not drug acquisition cost. Am J Manag Care. 2001;7(9):861-867.
9. Fendrick AM, Chernew ME, Levi GW. Value-based insurance design: embracing value over cost alone. Am J Manag Care. 2009;15(10 suppl):S277-S283.
10. Turnbull F, Neal B, Ninomiya T, et al. Effects of different regimens to lower blood pressure on major cardiovascular events in older and younger adults: meta-analysis of randomised trials. BMJ. 2008;336 (7653):1121-1123.
11. Hemmingsen B, Lund SS, Gluud C, et al. Targeting intensive glycaemic control versus targeting conventional glycaemic control for type 2 diabetes mellitus. Cochrane Database Syst Rev. 2013;11:CD008143.
12. EPOC Author Resources. Cochrane Effective Practice and Organisation of Care Group website. http://epoc.cochrane.org/epoc-authorresources. Published 2013. Accessed March 17, 2013.
13. Choudhry NK, Avorn J, Glynn RJ, et al. Full coverage for preventive medications after myocardial infarction. New Engl J Med. 2011;365(22): 2088-2097.
14. Choudhry NK, Fischer MA, Avorn J, et al. At Pitney Bowes, value-based insurance design cut co-payments and increased drug adherence. Health Aff (Millwood). 2010;29(11):1995-2001.
15. Chang A, Liberman JN, Coulen C, Berger JE, Brennan TA. Value-based insurance design and antidiabetic medication adherence. Am J Pharm Benefits. 2010;2(1):39-44.
16. Chernew ME, Shah MR, Wegh A, et al. Impact of decreasing copayments on medication adherence within a disease management environment. Health Aff (Millwood). 2008;27(1):103-112.
17. Farley JF, Wansink D, Lindquist JH, Parker JC, Maciejewski ML. Medication adherence changes following value-based insurance design. Am J Manag Care. 2012;18(5):265-274.
18. Maciejewski ML, Farley JF, Parker J, Wansink D. Co-payment reductions generate greater medication adherence in targeted patients. Health Aff (Millwood). 2010;29(11):2002-2008.
19. Gibson TB, Mahoney J, Ranghell K, Cherney BJ, McElwee N. Value-based insurance plus disease management increased medication use and produced savings. Health Aff (Millwood). 2011;30(1):100-108.
20. Gibson TB, Wang S, Kelly E, et al. A value-based insurance design program at a large company boosted medication adherence for employees with chronic illnesses. Health Aff (Millwood). 2011;30(1):109-117.
21. Kim YA, Loucks A, Yokoyama G, Lightwood J, Rascate K, Serxner SA. Evaluation of value-based insurance design with a large retail employer. Am J Manag Care. 2011;17(10):682-690.
22. Zeng F, An JJ, Scully R, Barrington C, Patel BV, Nichol MB. The impact of value-based benefit design on adherence to diabetes medications: a propensity score-weighted difference in difference evaluation. Value Health. 2010;13(6):846-852.
23. Choudhry NK, Rosenthal MB, Milstein A. Assessing the evidence for value-based insurance design. Health Aff (Millwood). 2010;29(11):1988-1994.
24. Fairman KA, Curtiss FR. What do we really know about V-BID? quality of the evidence and ethical considerations for health plan sponsors. J Manag Care Pharm. 2011;17(2):156-174.
25. Melnick SJ, Motheral BR. Is value-based value wasted? examining value-based insurance designs through the lens of cost-effectiveness. J Manag Care Pharm. 2010;16(2):130-133.
26. Fendrick AM, Martin JJ, Weiss AE. Value-based insurance design: more health at any price. Health Serv Res. 2012;47(1, pt 2):404-413.
27. Patient Protection and Affordable Care Act. Public Law No. 111-148, 124 Stat. 119. Vol 42 U.S. Code § 18001 (2010).
28. Administration announces regulations requiring new health insurance plans to provide free preventative care. Business Wire website. http://www.businesswire.com/news/home/20100714006468/en/Administration-Announces-Regulations-Requiring-Health-Insurance-Plans#.U4imMPldXas. Published July 14, 2010. Accessed March 16, 2013.
29. Kelly AM, Cronin P. Value-based insurance design: barriers to implementation in radiology. Acad Radiol. 2011;18(9):1115-1122.
30. Saini SD, Fendrick AM. Value-based insurance design: implications for gastroenterology. Clinical Gastroenterol Hepatol. 2010;8(9):767-769.
31. de Souza JA, Ratain MJ, Fendrick AM. Value-based insurance design: aligning incentives, benefits, and evidence in oncology. J Natl Compr Canc Netw. 2012;10(1):18-23.
32. Fairman KA, Curtiss FR. Making the world safe for evidence-based policy: let’s slay the biases in research on value-based insurance design. J Manag Care Pharm. 2008;14(2):198-204.
33. Fairman KA, Curtiss FR. Still looking for health outcomes in all the wrong places? misinterpreted observational evidence, medication adherence promotion, and value-based insurance design. J Manag Care Pharm. 2009;15(6):501-507.
34. Neumann PJ, Auerbach HR, Cohen JT, Greenberg D. Low-value services in value-based insurance design. Am J Manag Care. 2010;16(4):280-286.
35. Robinson JC. Applying value-based insurance design to high-cost health services. Health Aff (Millwood). 2010;29(11):2009-2016.
36. Ginsburg M. Value-based insurance design: consumers’ views on paying more for high-cost, low-value care. Health Aff (Millwood). 2010;29(11):2022-2026.