6.1. Study population

In order to obtain a homogenous study population and avoid chronicity, the inclusion criteria included dermatomal unilateral pain below the knee (dermatomes LIV to SI) with symptom duration from 3 to 28 weeks. Depressed patients and those applying for early retirement were excluded, because it was anticipitated that their treatment responses could be confounded. Patients who had undergone back surgery were excluded for the same reason. The physicians working in the University Hospital catchment area were thoroughly informed about the trial and most of the patients were referred directly from primary care, which means that the findings of these studies can be generalized to an unoperated sciatic population with a limited symptom duration. The final study population was in fact similar to other sciatic populations with respect to age and gender (Jönsson & Strömqvist 1993, Carette et al. 1997). Selection bias was probably insignificant because consecutive eligible patients were enrolled, and of the eligible patients only 8 refused to take part in the study. The recruitment of 160 patients was based on power calculations with an assumption of 15 % clinical effect and 10 % drop-out rate. Only two patients were lost during the follow-up period, which means that the treatment effects could be calculated reliably. However, even this study population was not large enough to reliably characterize the phenotype of the Trp2 allele. The study population was, however, comparatively homogenous. Most of the patients had experienced a disc herniation, and the mean disability and leg pain scores indicated a severe disease. Of the final population, only 29 had sciatica due to a non-herniated disc. Exclusion of these patients would have rendered the study population more homogenous, but then valuable information about intercorrelations between different types of disc displacement and clinical signs and symptoms would not have been obtained.