Traditionally, treatment for manic–depressive disorder has been categorized as acute versus prophylaxis, or maintenance; that is, treatment geared toward resolution of a specific episode versus continued treatment to prevent further symptoms. Treat-ment can also be considered along several other lines. In gen-eral, more structured treatment settings, such as full or partial hospitalization, are indicated if patients are likely to endanger self or others, if manic–depressive disorder is complicated by other medical or psychiatric illnesses that make ambulatory management particularly dangerous, or if more aggressive management is desired than is easily available on an ambula-tory basis (e.g., intensive psychosocial intervention or rapid dosage titration of psychotropic agents). In addition, social fac-tors play an important role in the decision to hospitalize in the real world of clinical psychiatry. Such reasons may include lack of social support to ensure medication compliance during acute illness, social stresses aggravating symptoms and making treat-ment compliance difficult (e.g., manipulative or hostile living situation), or lack of transportation to accommodate frequent ambulatory appointments during acute illness. Unfortunately, it is sometimes the case, although less frequent in this era of managed care, that a person’s insurance plan covers inpatient but not ambulatory mental health treatment, forcing expensive inpatient care when less costly, time-limited, intensive ambula-tory care would suffice.
Finally, treatment can be categorized according to its goals. Treatment can be focused on improving clinical outcome (episodes and symptoms) or functional outcome (social and oc-cupational function and health-related quality of life). Although this categorization appears straightforward, clinical practice reveals many subtleties. For instance, it is erroneous to assume that clinical outcome is the domain of pharmacotherapy and that functional outcome is the domain of psychotherapy. In actual-ity, most psychotherapies by design focus on improving symp-toms. Likewise, pharmacotherapeutic stabilization of symptoms clearly contributes to improved role function. Further, treatments that improve one domain may cause decrements in another. For instance, effective maintenance treatment with lithium may come at the cost of hand tremor, which interfere with work function and causes embarrassment in social situations. Compassionate psych-oeducation and alliance building are integral goals of each form of treatment. In analogy to infectious disease treatment, attention to such host factors can often make the difference between suc-cess and failure of treatment.
Great optimism justifiably accompanied the introduction of lith-ium in the 1960s, with the drug projected to save society millions of dollars in direct and indirect treatment costs. However, there are reasons to be concerned that lithium has made much less of an impact than originally projected, and there is clear evidence that manic–depressive disorder remains a major health concern, even with the addition of anticonvulsants to our armamentarium. For example, readmission rates for manic–depressive disorder may be as high as 90% during 2-year follow-up, with no differ-ence between lithium-treated and nontreated patients (Markar and Mandar, 1989). Overall, the impact of lithium “under ordi-nary clinical conditions” appears to be much less than would be expected from results of randomized clinical trials.
How can these data be reconciled with early estimates projecting dramatic decreases in treatment costs due to the intro-duction of lithium? Presumably, the medications themselves do not differ between controlled clinical trials and general clinical practice. If anything, the diffusion over time of the new phar-macological technology into general clinical practice might be expected to lead to further gains in illness management beyond those initially seen. The use of several anticonvulsants such as carbamazepine, valproic acid and lamotrigine in the treatment of manic–depressive disorder holds promise for further improve-ment in outcome. For instance, these drugs may have efficacy in controlled clinical trials, but concerns regarding the effective-ness of lithium in clinical practice also apply to the use of these anticonvulsants.
What, then, are the sources of this efficacy–effectiveness gap in the treatment of manic–depressive disorder? It is likely that the gap is in part due to the exclusion of “complicated” manic–depressive patients from clinical trials (e.g., those with substance abuse, personality disorders, or medical problems, and those unwilling to risk exposure to placebo. Although such ex-clusions are appropriate for establishing the efficacy of potential treatments, the exclusivity of structured clinical trials limits their relevance in the general clinical setting. Another likely contribu-tor to the efficacy–effectiveness gap is variation in provider at-tributes such as attitudes and capabilities. For instance, it is well established that even at academic medical centers, the intensity of medication treatment for mood disorders is much less than that which experts consider optimal. It is possible, then, that sup-porting providers with specific data regarding treatment options will aid in decreasing the efficacy–effectiveness gap. With this in mind, several organizations have developed clinical practice guideline to assist providers. Finally, the organization and orietn-tation of care giving systemes may not be optimally supportive.