Introduction: A New Look at Evidence About Reduced Cost of Medical Utilization Following Mental Health Treatment

Introduction

Insurance companies have commonly discriminated against the treatment of psychiatric illness. Although there has been some degree of parity regarding inpatient treatment (at least historically), very few insurance policies have customarily reimbursed outpatient psychiatric services at the same level that they reimburse outpatient nonpsychiatric medical treatment.¹ For many years, there has been a particular stigma attached to psychotherapy that appears to derive from several myths: 1) psychotherapy is not a real treatment; 2) psychotherapy is simply hand-holding that any nontrained professional could do; 3) there is no evidence supporting the efficacy of psychotherapy; and 4) if psychotherapy is available to the public as a component of a standard benefits package, “everyone” will use it and it will “break the bank.”

Because of these perceptions, it has always been an uphill battle to arrange payment for psychotherapy through third parties. One of the strongest arguments in the many skirmishes associated with that battle has been the so-called cost-offset argument. In brief, this is the argument that expenditures on medical and surgical care will be reduced following the provision of outpatient psychiatric services, especially psychotherapy. The classic article featured in this issue of the Journal has been one of the most influential contributions to this argument. Mumford and her colleagues performed a meta-analysis of 58 controlled studies of the effects of outpatient psychotherapy on subsequent medical utilization. Eighty-five percent of those studies reported a decrease in medical utilization following psychotherapy. Of these studies, 26 were naturalistic time-series studies that compared medical care utilization before and after psychotherapy, with each individual serving as his or her own control. Thirty-two studies were experimental in design, using either random assignment or some matching scheme to place patients with specific treatment conditions. Twenty-two of those 32 studies examined the effects of the psychotherapeutic intervention on patients who were hospitalized for medical crises. In this subgroup, inpatient hospitalization was reduced approximately 1.5 days below the average of 8.7 days for the control group. Five of the studies provided data that permitted an unbiased examination of the effects of psychotherapy on inpatient as well as outpatient medical care utilization. The average change was a decrease in inpatient utilization of 73.4% and a decrease in outpatient utilization of 22.6%.

In this same report, the investigators examined the claims files of the Blue Cross and Blue Shield Federal Employee Program from 1974 through 1978. During that time period, 53% of federal employees were insured by this program, thus providing the largest fee-for-service database available. Examination of the group of subscribers that used mental health treatment compared with those who did not found a significant cost-offset effect, primarily from the lowering of inpatient medical charges for the mental health group. Another significant finding from this analysis was that about 1.5% of persons covered actually received some form of mental health services in any one year during the 5-year period.

In drawing their conclusions, the authors stressed that it is erroneous to assume that the reduction in medical services linked to psychotherapy is a result of keeping “the worried well” from having access to medical outpatient services. They stressed that individuals who receive mental health services are physically sicker and suffer from more chronic disease than people who do not use psychiatric services. To assume that the impact of outpatient psychotherapy is simply a substitution of one outpatient service for another has no basis in the data.

Much has changed in the years since the appearance of this article. The cost-offset argument has fallen out of favor to a large extent. First, with the dramatic impact of managed care in reducing the length of hospital stays, the decrease in inpatient expenditures, which is where most of the savings lie in the cost-offset paradigm, is much less impressive. One of the most influential studies has been the Rand Health Insurance Experiment,² in which families at six sites in four states were randomly assigned to a variety of different health insurance plans with varying coinsurance rates. In this elegantly designed study involving 5,809 persons, no cost offset could be demonstrated for those who used mental health treatment. Although hospitalization was not included in the measurement of cost offset, the investigators also pointed out that since there was such small utilization of outpatient psychiatric services, it was difficult to show much of an impact of the treatment.

Another factor in the decline of the cost-offset argument has been the observation that in many cases the reduction in medical and surgical services cannot be demonstrated until the 12 months after the psychotherapy.³ Government policymakers, managed care organizations, corporations, and insurance companies tend to plan only for a 12-month budget year. An offset that will occur in the next budget year may be of little interest. Another concern about the cost-offset argument is that it has the potential to trivialize the serious impairment associated with psychiatric disorders.⁴ In other words, an assumption inherent in the notion of cost offset is that it makes good economic sense to provide insurance coverage for psychiatric treatments only because they can lower costs of medical treatments for illnesses that are more “real.” The psychiatric suffering in and of itself is not viewed as worthy of the same investment as nonpsychiatric illnesses. In fact, at least one state legislature has even introduced a bill that stipulates that mental health treatments can be covered only insofar as they reduce the cost of other medical treatment.

The other aspect of the cost-offset argument that has been criticized is that it ignores the profound cost in disability and impaired work performance and the cost incurred in mortality stemming from mental illness. The current thinking today is that a much broader notion of cost-effectiveness needs to be taken into account when considering the impact of psychotherapy.⁵ Direct costs related to the actual expenditures for delivery of the treatment are only one aspect of an economic assessment. The other aspect is indirect costs associated with problems on the job, disability, and loss of productivity related to the illness. There is also a shift away from emphasizing only cost while minimizing effectiveness.^{6, 7} In other words, within this paradigm there is often a small increase in direct treatment costs, but there may be quite a substantial payoff from that small investment because of the much greater improvement in the indirect costs. This conceptual model is in keeping with a movement in the health economics field to measure cost-effectiveness as involving all costs in the broadest frame of reference possible.

Recently, a panel of 13 scientists and scholars made consensus recommendations on cost-effectiveness analysis.⁸ They stressed that the societal perspective on cost-effectiveness must be the standard, not the narrow interests of an insurance company or managed care firm. They argued for a model of cost-effectiveness that “considers everyone affected by the intervention, and all health effects and costs that flow from it are counted, regardless of who had experienced them. Health effects include both benefits and harms, even when these occur in people who are not the intended recipients of the intervention. Resource costs include all resources used, whether or not money changes hands” (p. 1174).⁸

Research into the costs and benefits of mental health treatment continues actively, and this report by Mumford and colleagues paved the way for a variety of other studies that have demonstrated psychotherapy to be efficacious and cost-effective for many patients.⁵