Wednesday, January 6, 2016

CMS’ Standardized Plan Option Could Reduce Discrimination

Blog_OpenEnrollment_2016

In their recently proposed rule, the Centers for Medicare and Medicaid Services (CMS) created a “standardized option” for plans at the bronze, silver, and gold metal levels. Insurers could elect—but are not required—to create a standardized option for sale on the exchange.

Standardized options would have a fixed deductible, out-of-pocket limits, and standard copayments and coinsurance for a number Essential Health Benefit (EHB) services. For instance, a beneficiary would pay the same for a primary care appointment, emergency room visit, or surgery regardless of which standardized plan the beneficiary selects within a metal tier.

The rationale for the creation of the standardized option is that consumers find the large variety of cost-sharing structures difficult to navigate. The evidence supports CMS in this regard.

A study of Medicare Part D demonstrated that individuals regularly choose plans on the basis of premiums (which are readily available and easily comparable), and not expected out-of-pocket costs (which would involve a much more in-depth assessment of the benefit design). After Massachusetts decided to standardize their insurance options in 2010, and a group of economists found that standardization simplified consumers’ decisions and improved their overall welfare. Other state-based marketplaces, like California, have also adopted standardized plan designs.

While simplifying the vast array of plan options for consumers is an important goal in and of itself, there may be other important reasons to support plan standardization. Namely, plan standardization may be a solution to another emerging issue: the manipulation of benefit designs to discourage enrollment by those with chronic health conditions.

Plans have been shown to contain benefit designs that could make care unaffordable for individuals with certain health conditions. I was the lead author on a recent New England Journal of Medicine study that demonstrated how some insurers were placing all drugs used to treat HIV on the highest cost-sharing tiers, a process called “adverse tiering.”

A beneficiary in a plan with adverse tiering could expect to pay more than $3,000 more annually for their HIV medications, compared to a beneficiary in other plans. After the study was published, the Department of Health and Human Services (HHS) subsequently finalized a regulation labeling adverse tiering as a form of discrimination starting in 2016.

The ways that insurers may discriminate through benefit design manipulation extends beyond drugs, however. In a separate, recently published study in The Journal of the American Medical Association, we demonstrated how insurers were also excluding all of certain specialists that treat individuals with chronic conditions, including rheumatologists, endocrinologists, and psychiatrists. For someone who needs access to these specialists, this could mean paying nearly all of the cost of their out-of-network care. The possibilities for insurance discrimination through benefit design are nearly limitless — and with the failure of risk-adjustment programs to adequately compensate insurers for providing care to the chronically ill, these discriminatory practices could multiply.

Standardization As A Solution: Analysis Of The Data

How might standardization prevent discriminatory practices? The very definition of standardizing benefit designs means that insurers no longer have the flexibility to set cost-sharing amounts for certain health services (and discourage enrollment by individuals with certain chronic health conditions).

To test how plan standardization affects discrimination, I examined the 2015 health plans in the six most populous states that run their own state-based exchanges. This is before the regulation addressing adverse tiering goes into effect in 2016. Three of these states (California, New York, and Massachusetts) had state-based policies that required the creation of standardized plans, and three of these states (Washington, Maryland, and Colorado) did not.

For each state, the lowest premium silver plan was examined for each insurer, focusing on formularies and summaries of benefits and coverage to assess cost-sharing for HIV medications. I repeated the methodology of our New England Journal of Medicine study that first described adverse tiering — by incorporating cost-sharing for Nucleoside Reverse Transcriptase Inhibitors (NRTIs), negotiated drug prices from Humana to estimate drug costs for private insurers, and premiums for a 30 year-old non-smoker not receiving tax credits or cost-sharing subsidies.

Adverse tiering was used as an example of benefit design discrimination, and was defined identically as before: the placement of all NRTIs, both generic and brand-name, in tiers with coinsurance or copayment levels of at least 30 percent. Significant differences between standardized and nonstandardized plans were evaluated by using two-sample t-tests for expected costs, and by using chi-squared tests for categorization of adverse-tiering plans.

The differences between nonstandardized plans and standardized plans were profound. Plan standardization appeared to eliminate adverse tiering completely. Forty-five plans were examined in total, and none of the 19 standardized plans were categorized as adverse-tiering plans, while six of 26 nonstandardized plans (23 percent) were adverse-tiering plans (p<.05, see Table 1). While Massachusetts and New York both contained standardized and nonstandardized plans, the samples of plans each state were too small to make a meaningful comparison between standardized and nonstandardized plans within each state individually.

Furthermore, plan standardization led to large financial savings for those with HIV. The average annual cost of a brand-name HIV drug was roughly three and a half times more expensive in nonstandardized plans compared to standardized plans ($2,449 vs. $681, p<.001, see Figure 1).

The average annual cost of a generic HIV medication was more than three times more expensive in nonstandardized plans compared to standardized plans ($787 vs. $258, p< .01). The average annual cost of an HIV regimen was more than six times more expensive in nonstandardized plans compared to standardized plans ($5,443 vs. $896, p<.001). Even when premiums, deductibles, and out-of-pocket maximums are included, nonstandardized plans are on average $2,550 more expensive than standardized plans for HIV-positive beneficiaries ($7,161 vs. $4,611, p<.001).

There was no significant difference in monthly premiums between nonstandardized and standardized plans ($273 vs. $309, p= .11), although it is possible with a larger sample size, a difference in premiums would have been observed.

Table 1: Standardized Plans And Costs Of HIV Medication Regimen, By State And Plan

Douglas_Jacobs-Figure-2

*An Adverse-Tiering Plan places all Nucleoside Reverse Transcriptase Inhibitors in drug tiers corresponding to copayments or coinsurances of at least 30 percent.

** Average annual total cost for HIV-positive beneficiary includes plan-specific monthly premiums, deductibles, and out-of-pocket maximums.

The cost of Atripla was used to generate the average costs of an HIV Regimen. For the few plans that did not cover Atripla (n=3), The Sustiva and Truvada was used instead, which is similarly priced.

California, Massachusetts, and New York all had plan standardization policies. All plans in California are must be standardized plans, whereas Massachusetts and New York allow nonstandardized plans in addition to the required standardized plan offerings. All Adverse-Tiering Plans were nonstandardized plans (P<.05).

Figure 1: Average Annual HIV Medication Costs In Standardized Plans vs. Nonstandardized Plans

Douglas_Jacobs-Figure_1

* Average annual total cost for HIV-positive beneficiary includes plan-specific monthly premiums, deductibles, and out-of-pocket maximums.

The cost of Atripla was used to generate the average costs of an HIV Regimen. For the few plans that did not cover Atripla (n=3), The Sustiva and Truvada was used instead, which is similarly priced.

All differences are between standardized plans and nonstandardized plans are significant (all differences had P< .001, except for differences in the annual cost of a generic drug, which had P<.01).

The Standardized Option And Drugs: Regulatory Implications

This analysis provides evidence that plan standardization has the potential to reduce discriminatory benefit design practices like adverse tiering and reduce cost barriers for important medications that treat chronic conditions. Even though this analysis focused only on HIV medications, separate analyses have demonstrated adverse tiering occurring for other chronic conditions like rheumatoid arthritis, multiple sclerosis, cancer, schizophrenia, and even conditions as common as diabetes and asthma. As such, the creation of a national, standardized option represents an important step forward for a diverse array of beneficiaries.

However, there are some important differences between standardized plans in California, Massachusetts, and New York, and how the standardized option will be implemented on a federal level.

First, there are significant differences in the definitions of cost sharing for specialty tier drugs in California, Massachusetts, and New York, and the proposed standardized option specialty tier. For standardized plans in 2015, California had a specialty tier cost-sharing amount of 20 percent, and Massachusetts and New York had a maximum copayment of $70 for their most expensive tiers. The cost-sharing amounts of the most expensive tiers were effectively a cap on medication costs.

While standardized plans were typically still free to select what drug tier was assigned to what medication, even if a standardized plan in these states assigned the highest standard tier to all HIV medications, it still would not break the threshold to qualify for adverse tiering (which is 30 percent coinsurance). This is in stark contrast to Colorado, where most of the nonstandardized plans charge 30-50 percent coinsurance for higher drug tiers, which can correspond to more than $700 a month on average for an HIV regimen.

Similar to nonstandardized plans in Colorado, the federal government has proposed a 40 percent coinsurance for their specialty tier. This is not a trivial difference — in Colorado the annual cost of an HIV regimen was $8,512 annually compared to $1,727 in California. If the insurer creates a new standardized plan, it could tier most all medications used to treat HIV on the specialty tier (because of possible loopholes in previous federal regulations outlawing adverse tiering described here). In order to both realize a reduction in adverse tiering and to increase affordability, CMS should lower the cost-sharing amount for the specialty tier.

CMS should also define what constitutes a “specialty” drug, for the purposes of the standardized options. The specialty tier was initially designed as a way to designate those drugs that required additional assistance (often from pharmacists) with the drug’s administration. Even so, plans have increasingly assigned this tier to many different drugs, seemingly to discourage usage, or to pass more of the cost of the medication onto consumers.

Other plans took this a step further, and practiced adverse tiering. As long as CMS is creating the standardized option as a way to make shopping easier for consumers, they should clarify drug tier terminology so that “specialty tiers” can only be used on “specialty” medications. This would be a less burdensome requirement than plan standardization in New York, which requires that certain drugs be covered on certain tiers, and would be consistent with the regulatory goals of simplifying plan choices and nondiscrimination.

Implications Beyond Drugs

So far this post has focused on drug tiering, which represents one of the few areas that insurers would maintain the ability to change their benefit designs within the standardization option framework. Other cost-sharing requirements of the standardized benefit design, from emergency room visits, to skilled nursing facilities, to outpatient primary care visits, have to be consistent among all standardized options.

The importance of this cannot be overstated. As long as insurance companies have the incentive to discourage enrollment on the basis of health status, they can do so with by changing cost sharing for services in all aspects of their benefit design, but within the standardized option framework — they can’t.

However, again there seem to be exceptions. While standardized options (like all plans) still have to cover all of the essential health benefits categories, CMS has not specified the cost sharing for all required benefits.

The major categories that are not specified include durable medical equipment, emergency transport, mental/behavioral health inpatient services, substance use disorder inpatient services, habilitative services, maternity care, and children’s dental and vision services. As a result, insurance companies would be free to set cost-sharing amounts for these services on their own. This is particularly problematic because some of these nonstandardized services will be disproportionately used by high-need patients (like habilitative services for those with autism).

For dual purposes of improved shopping and nondiscrimination, CMS should set the cost-sharing amounts for all benefits. The state-based standardized plans reviewed in this analysis already do this. It is misleading to shoppers for a plan to be labeled “standardized,” while including numerous nonstandardized elements in each plan.

The objective is to make sure shoppers can compare plans on premiums and provider networks, knowing that the rest of the benefit design has largely identical cost-sharing requirements. It has become clear that failing to standardize all benefits could allow insurers to set cost-sharing amounts that discourage enrollment by those with certain chronic diseases.

Promoting Standardization

Standardized options appear to be an excellent way to ease health plan decisions and reduce or eliminate discriminatory benefit designs. The adoption of such plans should be encouraged to improve choices for consumers. One option would be a requirement for insurers to create standardized options, as California, Massachusetts, and New York have already done. In their proposed rule, CMS appears to have decided against this route for 2017.

A less burdensome way to encourage their adoption would be to display standardized options “first” on Healthcare.gov. In one behavioral study, over 60 percent of consumers selected their health plan without leaving the initial screen. Health insurance companies could be incentivized to create standardized plans to reach this large market share.

Indeed, CMS has stated that they are considering ways that standardized options could be displayed on Healthcare.gov, and so preferentially displaying standardized options first would fit this mold. If significant numbers of insurers create standardized options in 2017, CMS would face less resistance making the standardized “option” a “requirement” in subsequent years.

With some additional improvements to the final regulation—such the reduction in cost sharing for specialty tier medications, specialty tier definition, and standardization of the entire benefit design—standardized options could drastically improve the health insurance landscape. If standardized options are effectively promoted, beneficiaries will experience simplified shopping experience and de-facto nondiscrimination protections; two elements that could make standardized options a valuable policy for consumers and health insurance exchanges alike.



from Health Affairs Blog http://ift.tt/1MT34T6

No comments:

Post a Comment