Deciphering the Cost Impact of Managed Care in Medicaid

What happens to financial resources when managed care strategies are engaged by Medicaid to control costs in seriously mentally ill populations? Findings from the study in this month’s Journal by Shern and colleagues offer answers that pose a virtual rashomon for the interested observer. The study takes a societal perspective in the examination of the distribution of financial resources when patients receive care in three of Florida’s Medicaid plans: two managed care plans in the Tampa Bay area and one fee-for-service plan in Jacksonville.

In the 1990s, Florida had initiated a model program in the Tampa Bay area by means of a 1915(b) waiver that required psychiatric patients receiving Medicaid to enroll either in a full-service health maintenance organization (HMO) or in a behavioral health carve-out that consisted of a partnership between a private for-profit national managed care firm and local community mental health centers. This arrangement was one of the earliest attempts by government to apply managed care principles that had been incubated in the private sector (1 , 2) to a public sector venue. The arrangement allowed the investigators to compare the distribution of costs for enrollees of the managed care services in the Tampa Bay area to enrollees of the fee-for-service plan in Jacksonville.

The study employed bimonthly interviews with 630 adults from the three plans over a 1-year time period between 1997 and 1999. In face-to-face interviews, patients reported what services they had used, including general health, mental health, substance abuse, pharmacy, dental, vision, rehabilitation, and food stamps. Data pertaining to living situation, legal involvement, and out-of-pocket expenses and payments made by friends and family were incorporated as well. Time contributions from family and friends were valued at the minimum wage. This was a tremendous effort by the investigative team. Rarely have studies collected such detailed information from enrollees, and the follow-up rate for the interviews was impressive: 92% of the original study group was contacted at least once, for an average of five interviews (a perfect follow-up rate would have been an average of six interviews).

Annual costs per enrollee per year were grouped into three payer categories: 1) the Medicaid plan, 2) other (non-Medicaid) public sources, and 3) personal sources, such as family or friends. The central analysis focused on how these costs were distributed among the three financing strategies: the HMO, the behavioral health carve-out, and the fee-for-service plan. The results of multiple regression models that controlled for demographic, clinical, and financing conditions revealed that private costs increased robustly in the managed care plans compared to the fee-for-service plan. Although managed care strategies may have reduced costs relative to fee for service for the Medicaid payers, it was at the personal expense of the families and friends of the enrollees in the managed plans. The uncovering of this cost shifting is pertinent in light of previous findings that the clinical status of this study population and their satisfaction with their mental health outcomes were comparable across the three financing conditions (3) .

This increased burden on family and friends, which involved mainly time spent, was not trivial. Personal costs constituted almost 25% of the total expenditures for this population, and the toll on families and friends of enrollees in the HMO and carve-out was approximately four to six times greater than similar payments in the fee-for-service group. The regression model also revealed that there was an interaction between the enrollee’s living situation and the financing strategy: enrollees covered by managed care who were living independently had substantially higher personal costs than such enrollees in the fee-for-service plan.

Reactions to the findings of Shern et al. will vary depending on one’s perspective. For many clinicians, managed care is viewed as a cause of all evils, including now the shifting of costs away from Medicaid to family and friends. To be fair, the contracts between Florida Medicaid and the behavioral health carve-outs and HMOs did not have a profit incentive for the managed care companies built into them. The entities delivering services, however, were at full risk for all comprehensive community mental health services. One cannot say that a profit motive on the part of managed care companies drove the shift of costs from Medicaid to the families. The risk was shifted to the direct provider organizations, and their incentive appears to have been to shift the cost. Managed care is simply a tool; it is not necessarily good or bad by itself.

Indeed, a recent study of the addition of mental health and substance abuse benefits to the Federal Employees Health Benefits Program (4 , 5) found that managed care had an opposite effect in the private sector when managed care was associated with parity insurance coverage. In contrast to the Florida Medicaid study, the Federal Employees Health Benefits Program study found that parity coverage resulted in a significant decline in out-of-pocket costs for individuals who used mental health services. Managed care kept total costs down, which was important to the Federal Employees Health Benefits Program, whereas it improved the value of insurance for federal employees. For private patients accustomed to high out-of-pocket costs, parity and managed care together reduced out-of-pocket spending. For Medicaid patients, who are accustomed to making few or no out-of-pocket payments, managed care increased personal spending. The Federal Employees Health Benefits Program study is now used to argue that, at least in the private sector, mental health parity is feasible as long as managed care is in place and that it reduces costs for patients. Shern et al. have shown us that in the public sector, at least in Florida, the opposite seems true. Here, when a population of seriously mentally ill enrollees in Medicaid is subjected to managed care, dollars saved by Medicaid are balanced by increased personal costs and burdens for patients and their families.

From the perspective of the Florida Medicaid administrators and managed care plans, dollars were saved. They might also suggest that the added involvement of family and friends of the managed care enrollees affords the opportunity for closeness and involvement that may not have been present under fee-for-service financing. As psychiatrists who are familiar with populations such as the one in this study, we were disturbed to learn that managed care strategies specifically shift costs to family and friends. We have long known that these stakeholders are at risk for stress and subsequent psychological and physical illnesses by virtue of their relationship to affected patients (6 , 7) . We cannot see any justification or rationale for this burden from a clinical or public policy perspective.

However, the authors themselves point out correctly that further research is needed to delineate the nature of these personal expenditures under managed care. What changes when the plan shifts away from a fee-for-service reimbursement strategy? Are the extra efforts by families and friends appropriate to the patient’s clinical status? Are there consequential health costs incurred to families and friends as a result of the increased burden, as has been found in the past (8) ? We doubt that the cost shifting contributes to the satisfaction of either the patient or their loved ones. If we are correct, then the Florida Medicaid program should adjust the delivery of managed care services to accommodate the inequities uncovered by the findings of Shern et al.

From the perspective of Florida’s governing administration, the governor and cabinet, there is no escape from the findings of Shern et al. that money saved in Medicaid was shifted over to private citizens. Perhaps the intent was to shift the burden to Medicaid beneficiaries and their families in spite of the fact that they qualify for benefits because they are indigent and often disabled. As this phenomenon is understood by taxpayers, it is unclear whether managed Medicaid in its current form in Florida will be seen as an acceptable approach to health care financing. We hope that it is not. At a minimum, this study reveals the importance of linking financing strategies to clinical insight and assessment of total societal costs. It is not enough to simply cut health care costs in one program to accomplish wanted efficiencies.

The findings from the study by Shern et al. remind us that health care services must be organized with a point of accountability that is concerned with all costs to society (9 , 10) . In 2001, the Florida Commission recommended that a statewide Coordinating Council for Mental Health and Substance Abuse Policy be created in statute as part of the office of the governor. Such a council would be in a perfect position to react and follow up on the findings of the study by Shern et al. (11) . From a national perspective, Frank and Glied (12) have called for a federal “mental health stewardship function,” a strategic leadership framework modeled after the Office of National Drug Control Policy that would serve as a voice for mental health. We have to do better for our patients and their families.