Additional file 1.

Figure showing the TELECHIK study participation profile. The TELECHIK cohort sample was issued from the cross-sectional SEROCHIK survey, a population-based seroprevalence study held between 17 August and 20 October 2006, involving a random sample of the Réunion island community (2,442 individuals), selected by the French National Institute for Statistics and Economical studies (Insee) after stratification by age, gender, residence area, municipality size, and housing type [24,25]. The selection procedure of the TELECHIK population considered six exposure strata: true positive (symptomatic CHIKV infection), false negative (asymptomatic infection), not knowing positive (infection without memory of symptoms and serostatus), true negative (asymptomatic CHIKV negative to infection), false positive (symptomatic CHIKV negative to infection), and not knowing negative (absence of infection without memory of symptoms and serostatus), in order to account for declaration bias and the representativeness of the cohort, taking into account a feasibility constraint. Two subsets of the same size of true positives and true negatives were selected after stratification by age, gender and area of residence to control repartition bias, the allocation of participants within the six strata being conducted by applying reasoned sampling fractions (true positive 0.7 and true negative 0.46), or systematic selection (false positive, not knowing positive, false negative, not knowing negative). After elimination of those missing the call or individuals refusing, exclusion of another 54 individuals because of incomplete data or mismatched responders (different from the index person, parents, or legal guardian)*, the population was slightly skewed towards the selection of more women and older participants [23].

Format: JPEG Size: 91KB Download file

Gérardin et al. Arthritis Research & Therapy 2013 15:R9   doi:10.1186/ar4137