Incidence of IgA vasculitis in children estimated by four-source capture-recapture analysis: a population-based study.

Abstract

The aim was to describe the epidemiological characteristics of childhood IgA vasculitis (IgAV) defined by the EULAR/PRINTO/Paediatric Rheumatology European Society criteria in a population-based sample from France and ascertain its incidence over 3 years by a four-source capture-recapture analysis. Cases were prospectively collected in Val de Marne county, a suburb of Paris, with 263 874 residents <15 years old. Children with incident IgAV living in this area from 2012 to 2014 were identified by four sources of case notification (emergency departments, paediatrics departments, private-practice paediatricians and general practitioners). Annual incidence was calculated, and a capture-recapture analysis was used with log-linear modelling to estimate case-finding completeness. We identified 147 incident cases [78 boys; mean age 6.5 (s.d.:2.6) years]. The annual incidence (95% CI) was 18.6 (13.6, 24.5)/100 000 children. Although only 10% of children were exclusively identified by non-hospital sources, the completeness of case finding was 62%, with an undercount-corrected annual incidence (95% CI) of 29.9 (23.7, 37.3)/100 000 children. The annual distribution of diagnoses consistently showed a trough in summer months; 72% of children had infectious symptoms (mainly upper respiratory tract) a few days before IgAV onset; and 23% had a North African background. Our study supports secular and geospatial stability in childhood IgAV incidence and adds further indirect evidence for a possible role of a ubiquitous, non-emerging infectious trigger. Incidence studies from understudied areas are needed to disentangle the role of genetic factors better. Capture-recapture analysis suggests that a substantial portion of IgAV cases may remain unrecognized in epidemiological surveys.