Findings:
Cost to the State
The managed care pilots were predicated on the theory that the state could increase access to health care and reduce costs at the same time by enrolling Medicaid clients in HMOs. TDH's estimate of savings would appear to support this theory. But the underlying studies cited by TDH reveal the omission of significant cost factors that tend to reduce the savings attributable to managed care to a statistically insignificant level.
TDH reports FY 1997 savings due to Medicaid managed care of approximately $35.6 million, including $14.5 million paid to the state through profit sharing arrangements between TDH and the managed care organizations (Summary Report, p. 4). This is about 4 percent of total program costs in these areas. Intuitively, this makes sense because the state set the HMO capitation rate at a level 4 percent below the estimated FFS costs in each area. But several problems cloud this simple picture.
[Table 2] A small number of voluntarily enrolled blind and disabled clients account for half of the Rudd and Wisdom savings estimate
First, the often cited 4% discount off FFS costs is not as straightforward as it appears. In its 1998 study, Rudd and Wisdom discovered that actual FFS costs in the managed care areas were significantly lower than they appeared from the NHIC data used as the basis for HMO capitation rates.
According to Rudd and Wisdom: "If the FFS claims for managed care counties were overstated, as we believe they were, then the [HMO] premium rates were set at a higher level than would have been the case if the data were not in error. Our claims adjustment methodology has resulted in a reduction of the historical FFS claims and a corresponding reduction in the projection of FFS claims that would have been incurred had managed care not been implemented. This has led to a significant reduction in the estimated savings realized from managed care" (Rudd and Wisdom, p. 4).
Second, the cost of the managed care organization itself (the capitation rate and NHIC administrative expenses multiplied by the number of enrollees) is not the only cost in the managed care system. New costs-like the cost of the enrollment broker-have been introduced into the system. Old costs-like the cost of the Vendor Drug Program or the amount the state reimburses Federally Qualified Health Centers (FQHCs) for services-may have changed.
And the major FFS cost savings program-selective contracting-no longer applies to the areas where managed care rolls out. Plus, some services in the managed care areas remain under the FFS system, including the Vendor Drug Program and the costs of individuals who first present themselves to Medicaid for enrollment at the time they enter the hospital. While each of these individual items may not represent huge additional costs or lost savings, taken together they can eat up the amount TDH projects the state to have saved.
Two available studies compare the cost of managed care with the cost of traditional Medicaid. Each one addresses these external costs and their effect on savings differently.6 Rudd and Wisdom included few external factors (see table 3), and determined that managed care saved the state $28.1 million in FY 1997. Since Rudd and Wisdom assumed no profit sharing from managed care organizations, their savings estimate increases to $42.7 million with profit sharing added. This is the most optimistic figure available, so we use it as the starting point to see whether the savings are really there.
Vendor Drug and the Blind and Disabled Population
According to the Rudd and Wisdom study, more than half of the savings ($21.6 million) is attributed to lower costs for the voluntary enrollment of a relatively small number of blind and disabled people. In fact, Rudd and Wisdom predicted that in FY 1997 the state saved $10.2 million serving about 1,800 blind and disabled Medicaid enrollees in Tarrant HMOs alone. Statewide, disabled/blind Medicaid clients accrue costs averaging about $6,460 annually7. Rudd and Wisdom estimates that FY 1997 costs in Tarrant for blind/disabled clients without managed care would have been about $6000, while the costs for individuals enrolled in managed care dropped to $2600 (57% less). No medical management program could claim to have so severely cut back costs for a high utilization group without affecting quality of care, unless the costs were shifted elsewhere in the program.
[Table 3] Two major studies include different program costs
And this may be exactly what is happening here. Rudd and Wisdom did not include Vendor Drug program costs in its projections, although the three prescription limit is lifted for managed care enrollees and utilization might be expected to increase. In a separate study of the Vendor Drug Program under managed care, Rudd and Wisdom concluded that managed care reduces Vendor Drug costs, but the actuary elected to study all populations except blind and disabled-those individuals most likely to benefit from access to additional prescriptions. It is possible, in fact, that access to additional prescriptions is one of the biggest advantages attracting blind and disabled Medicaid enrollees to join an HMO. Rudd and Wisdom recommended that TDH apply its findings conservatively (that TDH assume no cost effect), in part because an analysis from the Bureau of Budget and Support Services (reviewed by Rudd and Wisdom) indicated that the average Vendor Drug cost for managed care clients is actually higher than that for FFS clients. This is logical, because one way to appropriately reduce emergency room and hospital utilization is to ensure that patients get the medications that will control their condition outside of a hospital setting.
According to A&M, blind and disabled clients account for the majority of vendor drug utilization, and the inclusion of the Vendor Drug Program in the cost savings equation reduces the estimated savings in Tarrant, Bexar, and Travis by a total of $11.9 million. Consumers Union supports continued access to adequate prescription care for all managed care enrollees, because this is a critical tool to prevent hospitalization. But any program cost analysis should include all aspects of the Medicaid program that might be effected by managed care. Given the potential for cost shifting, Consumers Union questions all the savings attributed to the blind and disabled population, but particularly the large savings estimated in Tarrant, Travis, and Bexar SDAs.
More Inclusive Analysis
Texas A&M attempted to incorporate many of the external costs (including vendor drug and TDH additional administrative costs), and concluded that managed care resulted in no statistically significant savings in Travis, Bexar and Tarrant service areas (see table 4). In fact, A&M researchers found that managed care may have cost the state more in Bexar than traditional Medicaid. In all five service areas together, A&M reported a total potential savings of $13 million-less than a third the savings reported by Rudd and Wisdom and probably not statistically significant for the billion dollar managed care program (in these areas) given margins for error in the study as a whole (see footnote 7 to table 4).
But, both the A&M study and the Rudd and Wisdom study omitted important cost factors, including the lost savings due to selective contracting, and the cost of the enrollment broker and the PCCM administrator. Consumers Union reviewed available information about the selective contracting program and the costs associated with new administrative tasks (see table 3) and attempted to assess whether these additional costs actually offset predicted savings (see table 4).
[Table 4] Managed Care adds significant new administrative costs, may increase vendor drug utilization, and reduces the savings from selective contracting
Selective Contracting
Prior to the implementation of managed care, the state initiated a program called LoneSTAR Select, designed to reduce hospital inpatient costs under traditional Medicaid by introducing price competition among hospitals. According to a 1998 Lewin study, hospitals negotiated FY 1997 discounts with the Medicaid program averaging 3 percent. More than 40 percent of Medicaid dollars pay for hospital care.
While selective contracting with hospitals resulted in a $47.9 million savings to the state in FY 1997, this was down from a savings of $57.4 million in 1995. Lewin attributed the nearly $10 million decline in annual cost savings primarily to the increased enrollment in managed care pilots (Lewin pp. 41-42). LoneSTAR Select discounts end when managed care rolls out for managed care enrollees (contracted discounts continue to apply to Medicaid recipients remaining in FFS) (Lewin, p. 21). This is important because, given the marginal nature of the projected savings due to managed care, the Lewin report indicates that much of the savings would have occurred anyway if all hospitalized patients were covered by traditional Medicaid under selective contracting.
Administration
Medicaid managed care creates new administrative tasks, and the costs must be included in any assessment of savings. TDH has expanded its managed care bureau in order to administer the complex system of new contracts. And, the multiple plan system creates the need for a new middle man-the enrollment broker, Maximus-to guide people through the much more complicated process of picking an HMO and a primary care doctor. TDH paid Maximus $13.6 million in FY 1998, the first full year of the Maximus contract. In the three initial contract months for FY 1997, the state paid $847,642. Additional TDH administrative costs amount to another $2.6 million. Consumers Union asked TDH for the amounts actually paid to Birch and Davis, the PCCM administrator, for FY 1997, but TDH refused to release the information and forwarded our request to the Attorney General for a determination as to whether it is public.
Taken together with the other ajustments, these administrative costs whittle away the savings estimated by Rudd and Wisdom (see table 4) to a level that is probably not statistically significant in the $1 billion dollar managed care program in these service areas.
Managed care was introduced in this state in order to increase access and quality of care at a reduced cost. Years later, we don't know if it has done either. The state has already started rolling out the Dallas and El Paso area pilots. These are the last of the very large metropolitan areas, and pose new and different problems even as the old problems remain. Now is the time to halt any new contracts for a while and look at what we are doing and where we are going.
|
