By Kelsey Berry
It’s time to talk about discrimination again — this time, in insurance benefit design.
A recent study in NEJM by Jacobs and Sommers has coined the term “adverse tiering” to describe the use of drug formularies by insurers “not to influence enrollees’ drug utilization but rather to deter certain people from enrolling [in the plan] in the first place.” [emphasis mine] Evidence of adverse tiering includes the placement of all drugs for certain condition in the highest cost-sharing tiers of drug formularies. This practice, it turns out, occurs fairly frequently – at least when it comes to a common HIV medication, nucleoside reverse-transcriptase inhibitors (NRTIs). Jacobs and Sommers analyzed the placement of NRTIs on formularies for 48 plans in 12 states using the federally facilitated insurance marketplaces, and found evidence of adverse tiering in 25% of plans. Their conclusion? Many insurers may be using benefit design to dissuade sicker people from enrolling in their plans. This raises concerns about adverse selection, as well as discrimination on the basis of health status – a practice the ACA was meant to address via community rating and guaranteed issue requirements, among others.
The study provides an important data point as we continue to assess whether the ACA is living up to our goals for health care reform. I believe we’ll see several more studies of this nature coming down the line, drawing attention to insurer practices that fail to comply with regulations, that are creative interpretations of vague requirements, or that aren’t addressed in existing regulations and may require new scrutiny. As we digest these, I’ll raise two important points for consideration:
(1) First, as Jacobs and Sommers point out, adverse tiering – and perhaps some other benefit design features that dissuade sicker people from enrolling in exchange plans – seem like instances of discrimination on the basis of health status. However, state insurance regulators have expressed difficulty in finding an ideal standard for identifying discriminatory benefit design. With regard to prescription drugs, CMS proposed performing outlier analysis on QHP cost-sharing as part of QHP certification, a process which would identify plans as suspect if they have an “unusually large number” of drugs subject to prior authorization and/or step therapy requirements in a particular category and class.” On this logic, if a few more plans in Jacobs and Sommers’s study had placed NRTIs in the highest cost sharing tier, those that had done so may no longer be considered outliers – does this imply that they do not discriminate? Kate Greenwood has discussed this topic on Bill of Health before, see “Market-Norm or Necessity.”
Another criteria for identifying discrimination has been used in a different health area — mental health policy. The 2008 Mental Health Parity and Addiction Equity Act regulates insurance benefit design to prevent plans from discriminating against individuals with mental health and substance abuse disorders. The criteria used to identify potentially discriminatory benefit design under this law involves checking whether behavioral health benefits are offered at parity with general medical-surgical benefits, in a single plan. In other words, regulators perform an intra-plan comparison, rather than an inter-plan comparison, to identify “discriminatory” benefit design. The advantage of this check versus outlier analysis is that there is a clearer built-in standard — medical-surgical benefits, themselves highly regulated — against which to compare benefits for a certain group of patients, rather than a contingent standard set by market norms. However, even this compliance check is imperfect: plans still have flexibility with regard to designing medical-surgical benefits, and it’s not always clear that what works for benefits in the medical-surgical category is adequate or even comparable to behavioral health benefits. This area is the subject of my ongoing work in mental health policy, and I’d be glad to discuss with anyone interested in the topic.
(2) For the ethicists in the room, the question on everyone’s minds: even if we can reliably identify discrimination in health plan benefit design on the basis of health status, does this constitute unfair discrimination? I think it’s safe to say that, based on Jacobs and Sommers, individuals with HIV in adverse-tiering plans are getting the raw end of the deal – they certainly are paying more out-of-pocket (an estimated $3000 annually, on average) than those with HIV not in adverse-tiering plans. But, we’ll have to go a little farther to demonstrate why this discrimination and its effects are unfair – I’ll leave it to the comment box to raise some ideas.