The Patient Protection and Affordable Care Act (ACA) introduced restrictions on underwriting and new federal subsidies to expand access for Americans through the individual and small group health insurance markets.
The nature of specific statutory provisions in the law require premium stabilization mechanisms, Risk Adjustment (RA), Reinsurance, and Risk Corridors (aka the "3 Rs"). The latter two programs were designed to stabilize volatility and smooth out disruption of new enrollees and their unknown medical costs as the market achieved pricing normalcy. RA, by contrast, is a perpetual program engineered to prevent adverse selection and compensate carriers based on risk.
At the time of this policy brief, the individual market has seen a material decrease in the number of participating health plans in the health insurance exchange marketplace for plan year (PY) 2017. Issuers are either exiting or otherwise reducing their geographic footprint in the marketplace, citing a range of concerns. The higher-than-expected costs of the risk pool and subsequent failure of the Risk Corridor program to limit losses has been a clear factor influencing plan exits. Another commonly referenced concern, and the focus of this brief, is the RA program itself, which some believe is inaccurately redistributing premium revenues, resulting in plans being overcompensated for higher-risk enrollees. Others in the plan community are satisfied that, while imperfect, the RA process is functioning as intended.
On September 28, 2016 Leavitt Partners hosted a forum focused on technical and policy considerations pertaining to the Risk Adjustment (RA) program in the individual and small group insurance markets administered by the Centers for Medicare and Medicaid Services (CMS).
This forum consisted of representation from several health plan and issue-based associations including: Association of Community Health Plans (ACHP), America's Health Insurance Plans (AHIP), Blue Cross Blue Shield Association (BCBSA), Consumers for Health Options and Insurance Coverage in Exchanges in States (CHOICES), Health Plan Alliance (HPA), Hospital and Health Plan Coalition, and the National Alliance of State Health CO-OPs (NASHCO). Health plans participating in this discussion included Anthem, Cambia Health Solutions, New Mexico Health Connections, Health New England, and others.
The 16 forum participant organizations explored areas of policy consensus and dissent pertaining to the vitality of the marketplace, technical considerations for the RA program, and timing of prospective adjustments.
General Market Sentiments
There was general agreement among participants regarding the plans' concern for the core health of the new individual insurance market. All participants acknowledged challenges faced by new and incumbent participants, but had varied opinions regarding the source of such challenges. Specific points of consideration include:
General Market Views – The majority of participants acknowledged common interest in a robust marketplace with appropriately aligned incentives for participation that promote competition, choice, and stable premiums. A few participants expressed that their members were either exiting or otherwise "staying out" of the market until there is a greater degree of predictability and stability. There was broad concern about the status of the individual and small group marketplaces to date, particularly in view of the plan departures and premium increases for PY2017.
Other Factors – The RA program was not designed to address the problems induced by inadequate premium rates. Congress' decision to defund the Risk Corridor program was highlighted as a factor that exacerbated the overall market situation. Other participants reviewed a litany of additional challenges adversely affecting the market, including the late authorization of transitional plans, inadequate Special Enrollment Period (SEP) enforcement, third-party payments, and insufficiently balanced risk pools. A few participants believe the RA-related challenges are eroding insurer participation, limiting competition, reducing product options, and increasing premiums in the market.
Directional Accuracy of the Formula – Several participants believe that the RA formula is "directionally accurate" in payment transfers from issuers with lower-than-average risk to those with higher-than-average risk, as intended. Specific cited evidence includes paid claims that correlate to RA transfers, and that the 2014-2015 transfers had measures of consistency. This view is aligned with CMS' position. Some participants stressed that directional correctness was insufficient given the magnitude of the financial implications for various plans and that significant adjustments are urgently needed to stabilize the markets and enable the ongoing participation of an adequate number of insurers to ensure healthy competition.
Small v. Large Issuer Considerations – Participants cited several viewpoints regarding the experience of smaller versus larger insurers. On the one hand, RA transfers for smaller carriers were roughly balanced between payments and receipts, indicating that such companies were not more likely to face payouts as a result of their size. In its most recent risk adjustment payment report, CMS found than an issuer's claims experience was far more predictive of an issuer receiving a risk adjustment payment than the issuer's size. On the other hand, some participants expressed the viewpoint that smaller carriers had much more difficulty in anticipating the average statewide risk profile when establishing premiums, and relatively small differences in actual plan enrollment could result in relatively large swings in their amount of RA transfers. Some participants pointed to these dynamics as creating greater volatility in payment transfers for smaller plans.
Timing – Participants' views varied regarding the speed with which adjustments should be implemented. Some participants believed certain technical changes (outlined in the following section) should be made immediately for future years. Some participants expressed a desire for certain changes to be administered retrospectively—particularly where simple solutions to acknowledged problems are available. Certain participants felt that, if CMS cannot implement such changes quickly, then the individual states should act to do so.
Other participants maintained incrementalism as a principle, affording the market an opportunity to internalize changes, appropriately adjust pricing and risk models, and gauge the efficacy of changes to the formula. A commonly referenced concern was unintentionally instigating greater volatility by instituting changes with uncertain effects. Some participants also indicated that states can choose to operate their own risk adjustment program or use the federal program. These participants put forward the viewpoint that any fixes and improvements to the federal program should be made by CMS, and not by individual states, to avoid inappropriate manipulation of payments and changes which could undermine the primary RA policy objective of preventing adverse selection among issuers.
Technical Considerations
The nature of RA renders a wide range of both policy and technical considerations. In some cases, policy and technical matters could be reconciled, while others are inconsistent.
Current Notice of Benefit and Payment Parameter Commentary – Early in the forum, participants explored the various adjustments in the proposed Notice of Benefit and Payment Parameters (NBPP), posted to the Federal Register on September 6, 2016. We observed minimal controversy to what has been proposed, with commentary on the following provisions:
• Pharmacy Data Inclusion – Participants were widely supportive of the use of pharmacy claims data as contributing to risk score evaluation in the RA model, believing this could improve the general accuracy of the model itself. Opinions differed on how expansive the pharmacy inclusion should be, as well as whether pharmacy data reinforces some of the bias in the formula. Some participants thought that pharmacy data could also help identify risk factors that would under-represent the risks of populations for newer plans that had less experience with members.
• Partial Year Enrollment – Participants were generally supportive of the proposed change to account for partial-year QHP enrollees. While not a notable point of contention or discussion, a few participants highlighted the need to pause pursuit of alternative models for assessing the risk and impact of these partial year enrollees until the durational factors could be better studied.
• High-Cost Risk Pooling – While principally aligned with the adjustments proposed in the NBPP, participants were divided on the benefits of establishing a high-cost risk pool for PY2018. Arguments characterizing the opposition cited the insufficient benefit derived from administering a complex, national calculation that would only net incremental benefit because of the minimal number of applicable individuals. Further, there was concern that a national model could create cross-subsidization complexity amongst and between states, unnecessarily amplify cost variation, and be antithetical to the program's objective to focus on intra-state risk stabilization. A minority of participants were supportive of the proposed language. It was noted that, if this change is to be made, doing so on a national basis would be the only feasible approach, given the very low incidence of such large claims (i.e., $2 million and greater).
• Data Recalibration Considerations – CMS proposes to continue using Truven MarketScan data, blended on a three-year average, as the calibration baseline for the PY2018 model and to continue consideration of using EDGE server data for this purpose in later years. There was a strong sentiment from participants that the RA formula would be well served by using EDGE server data to calibrate the model, as it more accurately reflects the nature of the marketplace and the risk. Participants have grown wary of the effects from calibration that reflect the large group market, given the marked inconsistencies in risk and patient behavior between these markets.
• Estimation Bias Considerations – CMS has recognized and outlined proposed ways it could handle issues pertaining to the recognized "estimation bias" in the model, where the formula overcompensates for high-risk enrollees and under-compensates for low-risk enrollees. We did not observe any serious dissent to that opinion or an adjustment, with some members very much in favor of adjustments to the formula that would remedy this asymmetry. We do note that due to time limitations we were only able to discuss the issue and observed general consensus that improvement would be helpful. However, we did not have time to discuss in great detail the different proposals CMS has outlined on how to correct for this bias.
Induced Utilization Factors – There was consensus that CMS should review the accuracy and sufficiency of the current RA-induced utilization factors, which may not accurately reflect demand differences for each metal level. This tends to result in significant negative transfers for Bronze plans, making those products less attractive to offer for carriers.
Rating Area Considerations – A point of dissent involved a potential imbalance in the RA formula as it relates to different geographical areas within a state. A few participants felt the model should be adjusted to better reflect rural or urban geographies and apply the transfer formula based on factors in those specific areas. Other participants pointed out the statutory constraints requiring a statewide calculus for this adjustment. A few organizations suggested that the decision on whether to apply RA on a statewide or (smaller) market-area basis should be left to the individual states.
Information Sharing and Timing – As expected, all participants were in agreement that more should be done to share information early and often with the market on relative RA outcomes and status. All participants were appreciative of CMS's efforts to foster an open and transparent dialogue regarding RA improvements citing the recent CMS White Paper and subsequent issuer conference as good examples of positive efforts on this front.
Capping Risk Adjustment Payments – Payment caps, a topic with wide dissent, was discussed as a possible short-term RA improvement. The policy's advocates argue that, incremental changes to the formula could be made in a way that did not further exacerbate challenges introduced by RA, in tandem with market stabilization efforts. Other participants pointed out that such a cap would circumvent the core objective of RA in mediating payment transfers to plans with higher risk and does not align with the pricing assumptions used to develop the applicable premiums. One participant suggested use of a "credibility" approach, where RA transfers would be adjusted to reflect an insurer's size and proportion of new (unknown) members.
Efficiency Measures – Various participants repeatedly voiced that the RA risk transfer formula does not reflect plan efficiency. Some participants expressed that plans achieving below-average premiums through low provider payment rates, efficient administration, and/or savings from effective care management, tend to be penalized by the formula's statewide average premium. However, a couple of participants noted the difficulty in incorporating such efficiency factors in the RA determination. Admittedly, there was insufficient time to further explore this dynamic, but we believe there would be merit in an extended discussion on how to reconcile these interests.
Statewide Average Premium and Accounting for Administrative Cost Structure – Many participants appeared to support a change to the RA risk transfer formula that would remove some amount of fixed administrative costs from the statewide average premium, on the grounds that such costs are unrelated to health risk. Some participants pushed to remove all administrative costs, using plan-specific average claims costs instead of the statewide average, or applying separate statewide averages for the different metal-level plans. The sentiment expressed by these plans is that using premiums instead of claims costs for RA also exacerbates the problems to factor plan efficiency as noted above.
Conclusions
This issue brief is not intended to make specific policy recommendations. Leavitt Partners goal has been to capture the various policies where there appears to be consensus in the marketplace on matters pertaining to the RA formula. Some of these changes are already under strong consideration by CMS and have the support of market participants. Given the collective observations that the market continues to face serious challenges and that several exiting issuers cite RA as a contributing factor, there is merit in CMS working with the industry to continue refining and improving the formula.
The views, opinions and positions expressed within these guest posts are those of the author alone and do not represent those of Becker's Hospital Review/Becker's Healthcare. The accuracy, completeness and validity of any statements made within this article are not guaranteed. We accept no liability for any errors, omissions or representations. The copyright of this content belongs to the author and any liability with regards to infringement of intellectual property rights remains with them.