The Economics of Mental Illness
In recent years, we have become accustomed to seeing figures that represent cost estimation or the economics of certain disorders, social interventions, or medications, both in the scientific and lay press. People suffering from schizophrenia in Canada are said to generate expenses of $2.3 billion every year in direct costs and an additional $2 billion in indirect costs for social support systems. Cost-saving measures are deemed important; for example, a new atypical antipsychotic is proposed to be cheaper by $493 because of its compliance record. The costs of hospitalization and medication are obvious and usual outcome measures, but what of lost productivity, degradation of quality of life, and burden on family? This is just a very brief list of factors involved that may be but often are not considered by policy-makers. Health maintenance organizations in the United States are greatly preoccupied by hospitalization and the direct cost of interventions and are maybe less concerned with hardship caused to patient or family, or they may think that these concepts are just too hard to measure reliably and so disregard them.
Recently, I was asked to review an analysis, conducted by the Canadian Coordinating Office for Health Technology Assessment, of prescribing practices, including inferred costs of traditional tricyclic antidepressants (TCAs) compared with the newer selective serotonin reuptake inhibitors (SSRIs) (1). Except for side effect profiles, which were different but equally disturbing to the patients, SSRIs were as successful but costlier than TCAs as a strategy to initiate treating a patient suffering from major depression. A government bureaucrat could strongly encourage the use of TCAs as a starting medication because they are as effective and so much cheaper than SSRIs. However, these results had been derived from the evidence of randomized clinical trials, which are scientifically more robust. When the authors considered more naturalistic settings as well as quality of life and other factors (not the least of which is the comfort of physicians in prescribing them and so treating more patients who need them), conclusions favoured the SSRIs. This example illustrates the importance of understanding the methodology of measurement when studying the cost-effectiveness of a treatment and interpreting the literature. It behoves us, as professionals, to be sufficiently acquainted with these techniques and methods to be able to evaluate reports in our journals and comment on them to the decision makers who allocate resources. (This prerogative, unfortunately, is increasingly absent from our sphere of influence).
The 2 articles concerning health economics that lead this issue of the Journal both address methodology in the measurement of mental health costs. Goeree and colleagues demonstrate that the traditional way (human-capital technique) of looking at productivity costs of schizophrenia-related premature mortality tends to overestimate greatly these costs, and they propose an alternative, the friction-cost method. One could then justifiably ask how useful to us as advocates this approach may be if it reduces the cost figures of this severe pathology. On the other hand, of what use are numbers if they are known to inflate estimates? Legislators might then make wrong decisions or, worse, dismiss the issue entirely and direct the money to other problem areas with more valid numbers.
In the second article, Latimer examines the economic impact of a community-based approach, assertive community treatment (ACT), compared with current alternatives. Surprisingly, this approach is not only successful clinically, it seems to reduce costs as well, although it is labour intensive. It also addresses the issue of measurement of outcome and proposes the use of additional variables to hospitalization costs. Evaluating the use of other resources, such as emergency-room visits, outpatient consultations, and housing adds to the complexity but also represents the everyday reality quality of the assessment. It is fascinating to find that fidelity to a program such as ACT also seems to be a key variable when evaluating outcome, much as is purity or adherence to a model when assessing outcome in psychotherapy research (2). It seems that patients require coherence between theory and treatment strategies for improvement, which should, after all, be our ultimate goal.
Finally, it should be noted that there is a dearth of relevant data concerning the economic consequences of emotional problems in other personal and societal terms: added medical service use, absenteeism, reduced work productivity (including accepting a job that is well below one’s potential), family distress, marital dissatisfaction, and divorce. Decision makers will have to encourage and facilitate the collection of these vital pieces of information so as to enlighten the setting of the right mental health priorities.
1. Trindade E, Menon D. Selective serotonin reuptake inhibitors for major depression. Ottawa: Canadian Coordinating Office for Health Technology Assessment; 1997.
2. De Rubeis RJ, Feeley M. Determinants of change in cognitive therapy of depression. Cognitive Therapy and Research 1990;14:472–82.
Gilbert Pinard, MD, FRCPC