Controlled trials of therapy in fibromyalgia syndrome.
Many different interventions have been studied in the therapy of fibromyalgia syndrome (Tables 1 and 2). While most have been effective, in general these trials have been short term. Furthermore, important or substantial improvement, when it has been assessed, occurs in only small proportions of patients. Long-term, comparative trials of both efficacy and toxicity are necessary. Trials such as these require large numbers of patients (compared with placebo-controlled trials, which are generally impractical in long-duration trials due to the large numbers of dropouts in the placebo arm) and therefore are expensive and difficult to accomplish. Two other approaches offer potential solutions to the problem of adequate long-term comparative trials: (a) N-of-1 trials and (b) meta-analysis. N-of-1 trials have the advantage of random assignment, double-blinding and multiple potential comparisons in the same patient. Meta-analysis involves combining the results of studies, which individually may have conflicting results and lack adequate statistical power, to reach an overall result with sufficient statistical power to make meaningful conclusions, especially with respect to comparative efficacy. Peluso and colleagues (1993) have performed a recent meta-analysis of available therapies in fibromyalgia syndrome and found that the effect-size (a standardized measure of the efficacy of a given therapy) of several non-medication therapies such as electroacupuncture exceeded that of traditional medication therapies. Unfortunately, lack of uniformity in the use of outcome measures across included trials and the small numbers of comparable non-medication trials makes definitive conclusions regarding relative efficacy of therapies difficult. Nevertheless, application of meta-analytic methods such as these should facilitate future comparisons of different interventions. Ideally, future clinical trials in fibromyalgia syndrome should employ the same outcome measures to permit application of these methods. Few trials have assessed improvement in functional status. Functional status measures such as the HAQ (Fries et al., 1980), the Fibromyalgia Impact Questionnaire (Burckhardt et al, 1991) or similar instruments should be employed in future studies of therapy in fibromyalgia. Given that individual modalities appear to confer relatively modest benefit on average. Combination approaches are reasonable, although randomized, blinded trials to assess these approaches are methodologically complex. Several preliminary studies which have addressed this approach appear promising (see Chapter 12; Goldenberg et al, 1993). Finally, no studies have yet assessed the comparative cost-efficacy of available treatments. Controlled trials which address the cost-efficacy of commonly employed, but unproven treatments such as physiotherapy chiropractic manipulation and injection techniques are urgently needed.