|Budget Issue:||Medi-Cal Managed Care Rate Setting Methodology for Two Plan Model|
|Program:||Department of Health Care Services|
|Finding or Recommendation:||Based on our review of DHCS’ new rate development methodology, which includes a risk adjustment mechanism, it seems reasonable and overall appears to be consistent with the Mercer recommendations. We also make several recommendations regarding TPM rate setting. Please see the narrative below.|
Concerns have been raised regarding the Two Plan Model (TPM) rate-setting methodology. In response, we provide a description of the new rate-setting methodology and related issues. We also offer our general assessment of this new rate methodology, outline issues for legislative consideration, and offer recommendations regarding TPM rate setting.
The TPM managed care model consists of a local initiative and a commercial plan operating within a single county. The local initiative must generally be locally developed and operated while the commercial plan is selected through a competitive bidding process.
2006 Rate Study. In September 2006, generally in response to federal requirements that Medi-Cal managed care rates be actuarially sound, the Department of Health Care Services (DHCS) contracted with Mercer Consulting to conduct a study of the Medi-Cal capitation rate development and reimbursement structure. Mercer’s study identified potential improvements to DHCS’ rate-setting methodology. Among Mercer’s recommendations was to develop a mechanism to make risk adjustments to the rates paid to health plans. Risk adjustment can be a cost-effective way to better align state capitated payments to health plans to account for differences in health plan risk related to the health needs of participating beneficiaries.
“Safety Net” Provider Participation Requirements. Safety net providers deliver health care to a significant number of Medi-Cal, uninsured patients, and other vulnerable populations. Examples of safety net providers include community clinics and hospitals that serve a disproportionate share of Medi-Cal and uninsured patients. Some safety net provider participation requirements apply for managed care plans that contract with the Medi-Cal Program. Both local initiatives and commercial plans must meet federal requirements that ensure that Medi-Cal beneficiaries covered in their plans can access medical services through Federally Qualified Health Centers (FQHCs), which are safety net providers that receive federal grants generally because they serve a high number of Medi-Cal and uninsured patients. State law requires that local initiatives subcontract with FQHCs and encourages commercial plans bidding for a state Medi-Cal contract to offer such subcontracts to FQHCs. Current California Medi-Cal regulations also require local initiatives to include other types of safety net providers in their delivery system. While these regulations are specific to local initiatives, many Medi-Cal commercial plans are required through their contract with the state to comply with these types of requirements.
State Enrollment Policy Encourages Safety Net Provider Use. The state also uses another mechanism to encourage safety net provider use by the health plans known as default enrollment. New Medi-Cal beneficiaries in certain counties who do not select a health plan are automatically assigned after 30 days to a health plan in their county. These are known as default enrollees. Plans that contract with a higher percentage of safety net providers and that score better on specified performance measures receive a greater share of these default enrollees.
The DHCS, which is in the process of implementing Mercer’s recommendations, used a risk adjustment mechanism to develop managed care rates for the first time for the 2009-10 rate year. The implementation of these risk adjustment changes is being phased in and consists of several key rate development steps which we describe below.
(1) “Plan-Specific” Rates. The DHCS begins the rate development process by developing a plan-specific rate for each individual local initiative and commercial plan. To develop this plan-specific rate, DHCS uses cost, utilization, and “encounter” data from each plan. (As described in more detail later in this analysis, encounter data is a term that refers to information about encounters between a beneficiary and health care provider, such as a clinic.) This plan-specific rate will constitute 80 percent of the final rate paid to each plan.
(2) Countywide Average Rate. The next step in the rate development process is for DHCS to develop a single countywide average rate from the two plan specific rates. This provides a baseline to appropriately apply risk adjustment factors later in the process. This county rate is intended to represent the general cost of doing business in that county.This is an important step to standardize plan rates so that factors for risk can be reasonably applied.
(3) Risk Adjustment Factors Determined With Medicaid Rx Model. Due to a lack of reliable and complete encounter data, an alternative tool is being used to assess the relative medical risk of Medi-Cal beneficiaries participating in the health plans. This national model, known as Medicaid Rx, has been tailored to California. This tool is being used to develop the risk factors that are applied to the county average rates that are calculated in the step described above. This tool relies mainly on pharmacy data to assess an individual’s health risks, which are combined for its beneficiaries to calculate a risk factor. Each plan’s risk factor is applied to the countywide average rate to establish a risk-adjusted rate for each plan. This will constitute 20 percent of the rate that will actually be paid to each plan.
(4) Final Rate Includes Only Some Risk Adjustment. The final step in the rate-setting process is to calculate the final rate for each individual plan. The administration has decided to phase in risk adjustment in order to protect plans from “rate shock.” Specifically, as noted above, the DHCS decided to reflect in the final rate only 20 percent of the adjustment that would otherwise be made for risk. This approachavoids huge year-over-year changes in plan rates and gives plans time to adjust their business practices to the new rates. (As noted above, the other 80 percent of the final rate is comprised of the plan-specific rate calculated in the first step.)
We note that the above risk adjustments were applied in such a way as to avoid increasing or decreasing the amount of funding that would be needed to be budgeted for these health plans after the plan specific rate was developed. Furthermore, only certain beneficiary groups and services were included in this risk adjustment.
Impact of Rate Methodology Changes to Plans Vary. The impact of the new rate methodology varies across health plans, resulting in rate increases for some and decreases for others. Because the risk adjustment has been applied as described above in a “budget-neutral” way, the new methodology effectively redistributes a fixed amount of funds between the plans. Therefore, risk adjustment creates both “winners and losers.”
Concerns Over Countywide Averaging. While health plans generally support the concept of risk adjustment, the local initiatives have expressed concerns over the countywide averaging component of the risk adjustment process. Specifically, these plans are concerned that this approach unfairly redistributes resources because it does not account for the specific obligations of local initiatives to support safety net providers.
The local initiatives have provided their own data that illustrates how the implementation of the county rate step redistributes resources between the plans. (To our knowledge, this data has not been verified by the department.) These local initiatives highlight data indicating that, on a statewide basis, local initiatives tend to have their rates reduced by this step in the calculation, while the commercial plans gain in this step of the process. While this is true, we note that this data also shows that some local initiatives benefit from this redistribution, while others do not. Furthermore, the data presented by the local initiatives does not reflect the final rates that would be paid to them as a result of the risk adjustment methodology. The further adjustments built into the rate-setting process would, in many cases, mitigate the impact of countywide averaging shown in the data.
Department Lacks a Timeline for Full Risk Adjustment Implementation. The department intends eventually to transition to fully risk adjusted rates, but does not have a timeline for full implementation of this change. Furthermore, the department has indicated that it plans to add other adjustments to the current rate methodology. These include “efficiency adjustments” to influence plan behavior and address inefficiencies in the base rate data not addressed with risk adjustment. For example, the department has indicated that it is considering making additional adjustments in the future that would provide better rates for plans with such outcomes for Medi-Cal beneficiaries as a reduced number of “preventable hospital admissions.” (This term generally refers to an admission to a hospital that could have been prevented through better primary care, such as by appropriately controlling asthma conditions for the participants in the plan.) The department indicates that it is currently in discussions with the health plans for the 2010-11 rate year regarding the timetable for a further phase-in of risk adjustment as well as potential efficiency adjustments.
Based on our review of DHCS’ new rate development methodology, which includes a risk adjustment mechanism, it seems reasonable and overall appears to be consistent with the Mercer recommendations. We also believe the administration’s decision to start the phase-in of risk adjustment at 20 percent is sensible, although a reasonable case could be made for a somewhat higher or lower adjustment percentage. In the long run, we believe risk adjustment will encourage efficiencyby ensuring that the rate paid to health plans are aligned with the health plan risks covered by each plan.
As noted above, the local initiatives have voiced concern that their use of safety net providers in their networks is not appropriately reflected in their rates under this new methodology. As we discussed previously, our analysis of the data presented by the local initiatives indicates that there would be both winners and losers in this group from the risk adjustment process. Also, as we noted, their data does not take into account the final rates set as a result of the risk adjustment methodology.
Moreover, the safety net data that we have reviewed does not demonstrate that local initiatives are necessarily being disadvantaged in this process relative to commercial plans. Based on our discussions with the department and plan representatives, and the limited data provided to us so far for our review, both the local initiatives and commercial plans seem to be making significant use of safety net providers in their networks. More detailed data regarding the use by beneficiaries of safety net providers for each of the plans would be needed to assess the claim that local initiatives are unduly disadvantaged in the rate-setting process. The DHCS has stated that it is continuing to study this issue.
Rate Development Process Is Not Transparent. This complex and technical rate development process described above for adjusting health plan rates for medical risk has not been adequately communicated to the Legislature and managed care plans. One legislative concern has been that the changes that DHCS has been making to the rate-setting process have not been formally identified in public documents provided to the Legislature, such as in the Medi-Cal local assistance estimate, the main document containing the Medi-Cal local assistance budget request. This information also has not been provided in quarterly reports that must already be provided to the Legislature by DHCS on managed care plans.
The managed care plans have raised concerns that they do not have access to data that would allow them to verify that their rates are being calculated in a manner consistent with the new risk adjustment methodology. For the same reason, they also are concerned that they lack information about some of the specific adjustments made by DHCS in setting their rates. The department has stated its intention to address these transparency issues by (1) providing plans with specific data on beneficiaries from the Medicaid Rx model as well as (2) information about some of the specific policy adjustments made by DHCS in their calculations of rates.
Better Encounter Data Is Needed. As indicated earlier, DHCS lacks reliable and complete encounter data on Medi-Cal patients who are covered under the health plans. As a result, DHCS has been using pharmacy and other Medicaid Rx data as a substitute for a full set of encounter data to adjust rates for medical risk. Part of the reason that such encounter data is difficult to obtain for Medi-Cal beneficiaries in health plans is because those plans in a number of cases use subcontractors to deliver medical services. The encounter data that may be collected by those subcontractors is not being provided to DHCS.
Based on our review of the DHCS’ rate development methods for the TPM and related issues, we recommend that the Legislature direct DHCS to: