Validity of self-reported nickel allergy.

BACKGROUND To estimate the prevalence of nickel allergy, self-reports are sometimes used in epidemiological studies. Self-reports are practical and may facilitate estimation of prevalence provided that the questions are validated. OBJECTIVES To investigate the validity of self-reported nickel allergy. METHODS Three hundred and sixty-nine women, aged 30-40 years, from the general population participated in the study. The participants answered a questionnaire before a clinical examination and patch testing. The two questions being validated were 'Are you sensitive/hypersensitive/allergic to nickel?' and 'Do you get a rash from metal buttons, jewellery or other metal items that come in direct contact with your skin?' RESULTS Patch test showed nickel-positive reaction in 30% of the subjects. Self-reported prevalence of nickel allergy as indicated by the two respective questions was 40% and 35%. Positive predictive values for the two questions were 59% (95% CI 50-67) and 60% (95% CI 51-69). History of childhood eczema was over-represented among women with 'false-positive' self-reported nickel allergy (P = 0.008). Self-reported hand eczema or 'high wet exposure' did not influence the validity. CONCLUSIONS The validity of self-reported nickel allergy is low. The questions regarding nickel allergy overestimate the true prevalence of nickel allergy.