No pain, no gain. Measuring treatment effects in fybromyalgia

Fibromyalgia syndrome (FM) is a condition of chronic pain, fatigue and other symptoms without a known anatomical substrate. Until now, treatment is mostly unsatisfactory; empirical studies on pharmacological and non-pharmacological treatment strategies in FM are needed. Following an overview of FM and its treatment, two such trials are described in this thesis. The first one is a pre-randomised clinical trial of spa treatment in FM. A combination of thalassotherapy, physical exercise and patient education in a Tunisian spa resort improved FM symptoms and health-related quality of life. Differences between spa treatment (n = 58) and treatment-as-usual (n = 76) were statistically significant after 3 months, but not after 6 and 12 months. Cost-effectiveness analysis showed no significantly different costs in the spa and control group, except for the cost of spa treatment (€ 1,526). Quality of life was higher in the spa group: 0.04 Quality Adjusted Life Years (QALYs) based on the visual analogue scale for general health (p< 0.01). The incremental cost-utility ratio after 6 months of follow-up was € 1,311 / 0.04 = € 32,775 per QALY gained (95% confidence interval -375,000 to 273,000). The second study is a 6-week double-blind randomised controlled trial comparing venlafaxine 75 mg/day (n = 45) to placebo (n = 45). The results of the Fibromyalgia Impact Questionnaire (FIQ) suggested a positive effect. Primary outcome measures (McGill Pain Questionnaire and visual analogue painscales) improved by 10%. However, no statistically significant differences between venlafaxine and placebo were seen. The importance of placebo effects in many FM drug trials is discussed separately. The combined data from both trials were used to validate a Dutch translation of the FIQ, which showed sufficient reliability, construct validity and responsiveness to be recommended as an outcome measure in FM. The studies described in this thesis show that empirical studies in FM are feasible and determination of effects and effect sizes is possible.