Value-based insurance design program in North Carolina increased medication adherence but was not cost neutral

By Matthew L. Maciejewski et al.
Health Affairs, February 2014

Value-based insurance design (VBID) … is based on the premise that higher medication and administrative expenses incurred by insurers will be offset by lower nonmedication expenditures that result from better disease control. This article examines Blue Cross Blue Shield of North Carolina’s VBID program, which began in 2008. The program eliminated copayments for generic medications and reduced copays for brand-name medications. Patient adherence improved 2.7–3.4 percent during the two-year study period. Hospital admissions decreased modestly, but there were no significant changes in emergency department use or total health expenditures. The insurer incurred $6.4 million in higher medication expenditures; total nonmedication expenditures for the study population decreased $5.7 million. Our results provide limited support for the idea that VBID can be cost-neutral in specific subpopulations…..

[S]tatistically insignificant short-term changes in expenditures fall short of demonstrating that VBID can bend the cost curve as anticipated by its proponents.

http://content.healthaffairs.org/content/33/2/300.full

In the space of a decade and a half, the concept of “value-based insurance design” has undergone a rapid transformation: from an idea few had heard of; to the latest health policy fad; to the subject of some very good research that refuted claims that VBID will cut costs. It is not unusual for unproven health policy notions to leap onto center stage. It is unusual, however, to see them subjected to unbiased, rigorous analysis so soon after they have been loosed on the public. For cultural anthropologists who like to study the mores of the US health policy community, the VBID fad provides an interesting case study because its treatment has to date proven to be the exception to the rule.

At the turn of this century, the label “value-based insurance design” had not been invented, and discussion of the underlying concept was limited to a few professional journals. Judging from several of the early papers on the concepthttp://www.ncbi.nlm.nih.gov/pubmed/11570020, its advocates were motivated by a desire to counteract the destructive effects of yet another health policy fad – the “consumer-driven” health policy.

In 2004, thanks largely to an article in the Wall Street Journal, the VBID idea burst out of obscurity and quickly became another hyped nostrum in the managed care pharmacopeia. The article described a successful experiment by Pitney Bowes, a self-insured Fortune 500 company, designed to reduce its total health care spending on employees with asthma and diabetes by reducing copayments on asthma and diabetes drugs. http://www.sph.umich.edu/vbidcenter/news/pdfs/2004_WSJ%20article%20%20(2).pdf

By 2005 the VBID label had been invented and a VBID “center” had been created at the University of Michigan. By 2006 the VBID label was appearing regularly in the professional literature. For the next several years the lay and professional literature published claims that VBID was not just good for patients but would also lower spendinghttp://content.healthaffairs.org/content/26/2/w204.full  By 2009 congressional Democrats had incorporated VBID into legislation that would become the ACA and, when the ACA was enacted in March 2010, VBID was engraved into US law. (The phrase appears in the third full paragraph of the ACA under a paragraph entitled, “Coverage of preventive services.” http://www.gpo.gov/fdsys/pkg/BILLS-111hr3590enr/pdf/BILLS-111hr3590enr.pdf  It reads: “The Secretary [of Health and Human Services] may develop guidelines to permit a group health plan and a health insurance issuer … to utilize value-based insurance designs.”)

But at about the time the ACA was enacted, research was beginning to demonstrate that VBID, at least as it applied to drugs, did not save money. A paper co-authored by, among others, people affiliated with Pitney Bowes concluded that Pitney Bowes’ VBID program did not cut costs http://content.healthaffairs.org/content/29/11/1995.full?ijkey=V1BuR15ZD….  A literature review published in Health Affairs in 2013 concluded VBID programs “were consistently associated with improved adherence (average change of 3.0 percent over one year),” but they did not lead to “significant changes in overall medical spending for patients and insurers.” http://www.ncbi.nlm.nih.gov/pubmed/23836741 The study by Maciejewski et al. quoted above, which examined the first VBID program run by an insurance company (as opposed to a self-insured company), reached the same conclusion. The insurance company, Blue Cross Blue Shield of North Carolina, lowered copayments for medications prescribed for hypertension, hyperlipidemia, diabetes, and congestive heart failure. Adherence rates rose 3 percent over the first two years (2008-2009), but total spending did not fall.

It is instructive to ask at this point what it is about the VBID proposal that distinguishes it from other long-lived managed care fads such as the HMO, “coordination,” utilization review, report cards, and electronic medical records, all of which were and still are promoted as cost-containment ideas. A closer look at the VBID’s evolution suggests three features account for VBID’s different treatment:

•  It was clearly defined;

•  its costs, including the indirect costs, were much easier to measure; and

•  the great majority of all costs were paid by the same payer (a self-insured employer and, more recently, an insurance company).

By contrast, many managed care fads are poorly defined, their indirect costs (for example, the administrative costs that utilization review imposes on providers) are harder to track, and the costs (as opposed to the savings) are absorbed by multiple parties.

A few payers have recently begun to experiment with applying the VBID concept beyond prescription drugs and using the “stick” approach (higher copayments for “lower valued” services) rather than the “carrot” approach (lower copayments for “higher valued” services). If that practice becomes widespread, the VBID concept may become more amorphous and its costs more difficult to track. In that event, we may see a resurgence in evidence-free claims about the potential for VBID to cut costs. If that happens, researchers should insist that analyses of VBID must continue to include an examination of all costs generated by VBID, including patient out-of-pocket expenses and higher medical costs caused by any damage to patient health.