Prevalence of Willis-Ekbom disease in rural coastal Ecuador. A two-phase, door-to-door, population-based survey

Objective To estimate the prevalence of Willis-Ekbom disease (WED) in a racially homogeneous population of adults Amerindians living in rural coastal Ecuador, and to assess the reliability of the International Restless Legs Syndrome Study Group (IRLSSG) questionnaire when used in population-based studies. Methods Two-phase, door-to-door, population-based survey. During Phase I, rural doctors screened all Atahualpa residents aged ≥ 40 years with the IRLSSG questionnaire. In Phase II, neurologists evaluated suspected WED cases and a matched sample of negative individuals. Results The census identified 665 persons aged ≥ 40 years. An affirmative response to the questionnaire was obtained in 94 persons. Medical history and neurological examination confirmed the diagnosis of WED in 40 of them. The evaluation of 188 non-suspected individuals revealed no further cases. The questionnaire had a sensitivity of 100% (95% CI, 89% to 100%), a specificity of 78% (95% CI, 72% to 83%), a positive predictive value of 0.43 (95% CI, 0.33 to 0.53), and a negative predictive value of 1 (95% CI, 0.98 to 1). The prevalence of WED in Atahualpa residents aged ≥ 40 years was 6%. Conclusion The prevalence of WED in this adult Ecuadorian population is higher than that reported from most studies conducted in tropical countries, but similar to that found in Brazil. Specificity and positive predictive value of the IRLSSG questionnaire are poor, which reinforces the need for a second phase that should include an expert interview.