Economic Grand Rounds: Medical Necessity: Its Impact in Managed Mental Health Care

All managed care techniques are predicated on the notion that only services actually necessary to treat a sickness or injury should be reimbursed. Clearly, delivering unnecessary services increases the cost of health services. Thus, for example, cosmetic procedures to reduce the signs of aging may be desirable, especially to the patient, but clearly not necessary to treat a sickness or injury.

To define what is necessary, most payers have developed a definition of medical necessity. In general, such definitions require that to qualify for payment, a service must be:

This definition of medical necessity is too limited to adequately serve the needs of those seeking mental health and substance abuse services, and its application may inadvertently increase future costs.

Sabin and Daniels (1) have discussed issues related to determining medical necessity in mental health and substance abuse services. They pointed out that insurance is designed to pay for the treatment of sickness or injury, but not for services for people who are "suffering from life." They argued that treatment is never medically necessary for shyness or other personal enhancement, loneliness and irascibility and other temperamental characterological issues, or lack of personal satisfaction. They also discussed the issue of moral hazard, that is, the insurance concept that a hazard is created when individuals alter their behavior so as to claim benefits, such as burning an insured property to receive reimbursement.

Insurance underwriters have expressed fears that if greater mental health and substance abuse benefits are available, people will increase their use of these services for problems in living rather than for psychiatric disorders, thereby creating moral hazard. Sabin and Daniels contended that insurance should pay for services designed solely to treat DSM-IV disorders and, as a corollary, services intended to help a patient return to or achieve normal functioning. This view is consistent with the generally used definition of medical necessity.

The actual day-to-day implementation of the concept of medical necessity is affected by variables other than just diagnosis. A prime example concerns how reimbursement can affect the determination of medical necessity of a given service by a clinician. If a provider's income is enhanced by increasing service utilization, such as in a fee-for-service reimbursement system, a subtle incentive may exist for the provider to overdiagnose or overrepresent to the behavioral managed care organization the seriousness of symptoms so that a client contending with a life problem may be called depressed.

Similarly, if a provider's income is decreased by increased utilization, a person's depression may be minimized by the therapist and described as contending with a life problem. The same contingencies apply to a behavioral managed care organization, whose self-interest will also influence the day-to-day application of the concept of medical necessity. The diagnosis-based approach to determining medical necessity can be tainted by nonclinical factors. Basing the determination on an assessment of day-to-day functioning related to the mental or substance use disorder may help correct for this influence of nonclinical factors.

The current formulation of medical necessity also has unanticipated negative cost consequences. For example, to deny outpatient psychotherapy as not medically necessary for a person who is unhappy because of marital problems may be very shortsighted if that individual decompensates for lack of treatment and is subsequently hospitalized for major depression. In this scenario, long-term costs may actually increase as a result of a short-term focus on medical necessity.

Utilization review of outpatient care to determine medical necessity also has unanticipated negative cost implications. Outpatient review is labor intensive for behavioral managed care organizations. Because of the large number of outpatient cases, more staff are required per covered member than for inpatient review. A compounding factor is the perceived need for frequent outpatient reviews, usually at intervals of three to ten visits. The assumption is that an inverse relationship exists between the number of reviews and actual outpatient utilization—that is, the more frequent the reviews, the lower the utilization. Yet the median number of outpatient visits in community mental health and substance abuse treatment centers is 3.5, and the modal number of visits is one (2,3). Private-sector data indicate that the median number of visits is 4.5 (4). Furthermore, outpatient benefits are often fairly limited, usually ranging from 20 to 30 visits a year, combined with a sometimes high copayment. Finally, the costs to appeal a denial of outpatient care can significantly increase the cost of service without increasing the number of services delivered.

As an alternative to simply performing medical necessity determinations, outpatient review ought to have a case management function. The goal of the review could be to encourage higher use of outpatient services, particularly for patients likely to relapse or for those who without such services would require more intensive levels of care.

The most adverse cost impact of implementing the current concept of medical necessity comes from its lack of recognition of the need for continuing treatment even when signs and symptoms of a mental or substance use disorder are not present. This lack is particularly true for long-term conditions that may be in remission but for which ongoing treatment is necessary to avoid more intensive levels of care in the future. Certain personality disorders and severe and persistent mental illness are examples. Another is substance abuse.

In some preliminary work with women with substance use disorders, Norman Hoffmann (personal communication, 1996) has found that a minimal course of treatment is required to produce any discernible effect in terms of abstinence. Once the "dose" of treatment exceeds this minimum, additional improvement is initially noted with greater doses, but returns diminish as the amount of treatment increases. Hoffmann likens this phenomenon to the dose-response curve found in other areas of medicine. His early findings with women suggest that more than 70 percent of those who had received at least five months of continuing care after primary treatment remained abstinent for the first year after treatment. Anything less than three months of aftercare resulted in negligible abstinence rates.

Looking at a unit of care, defined as four hours, several thresholds of care emerge. One is at about 55 to 60 units of service, and another at about 95 to 100 units. Women who do not receive at least 60 units are less than 50 percent likely to be abstinent one year after treatment. Of those who receive 60 to 99 units of service, almost 60 percent are abstinent, and almost two out of three of those who receive 100 or more units of service remain abstinent. Clearly, providing too little care may be inexpensive in the short term, but may be expensive to the patient and society in the long run.

Finally, the very use of the term medical necessity implies certain beliefs about mental health and substance abuse problems—that is, that they are solely medical disorders, which is not universally accepted. In addition, the concept has led behavioral managed care organizations to look only for a medical rationale for continuing stay at any level of care, such as the need for 24-hour nursing services as a prerequisite to inpatient treatment.

Notwithstanding the already limited benefits and high copayments in many mental health and substance abuse benefit structures, the current definition of medical necessity further limits providing needed mental health and substance abuse care. Even if parity with coverage for other illnesses is achieved, access to the behavioral health benefits will still be limited by the current medical necessity definition. A new approach is needed for determining what is necessary to treat, and required to pay for, mental health and substance abuse problems.

"Treatment necessity" or "clinical necessity" would require that to qualify for payment, a service must be:

The concept of treatment necessity would help to address the shortcoming of the current concept of medical necessity without encouraging the use of benefits for personal growth or for general welfare. The risk of moral hazard would be contained. The concept also has the merit of supporting relapse prevention, as well as providing services to prevent the need for higher levels of care. The definition, combined with benefit parity, would encourage more appropriate managed behavioral health care.

However, the overall cost implications of this new definition for payers should be assessed. It is unfortunate that commercial payers shy away from future savings strategies, partly because of the one- to three-year contracts many hold with employers. In these circumstances, savings generated as a result of today's decisions could accrue to a competitor three years from now. Notwithstanding resistance that may be encountered from payers, it is time to retool traditional medical necessity approaches to managed mental health and substance abuse care.

Dr. Ford is a project director at Health Systems Research, Inc., 2021 L Street, N.W., Washington, D.C. 20036. Steven S. Sharfstein, M.D., is editor of this column.