Influence of proxy respondents in children's health interview surveys.

STUDY OBJECTIVES To study the influence of the proxy respondent on health interview surveys in children. DESIGN Cross sectional study. SETTING Children under the age of 15 years drawn from the general population of Catalonia, Spain. PARTICIPANTS The Catalan Health Interview Survey consisted of a multistage probability sample representative to the population of Catalonia. The sample size was 2433 children younger than 15 years of age. The interviews were answered by proxy respondents (the mother, father, or other carer), with the questionnaire adapted for the proxy respondent. Logistic regression models were used to analyse the relation between the proxy respondent's characteristics and health status and health care utilisation, controlling for the effect of sociodemographic factors. MAIN RESULTS Proxy respondent's characteristics influenced the reports of chronic conditions and accidents within the last year. Proxy respondents over 55 years (OR = 0.47; 95% CI = 0.26, 0.82), men (OR = 0.69; 95% CI = 0.53, 0.89), the father (OR = 0.66; 95% CI = 0.50, 0.89), and the grandparents (OR = 0.49; 95% CI = 0.26, 0.89), reported a lower rate of chronic conditions. Age of the proxy 55 years or greater (OR = 0.41; 95% CI = 0.20, 0.82), men (OR = 0.70; 95% CI = 0.52, 0.94), fathers (OR = 0.68; 95% CI = 0.49, 0.92), and grandparents (OR = 0.40; 95% CI = 0.18, 0.85) showed a lower probability to report accidents. No variables related to the proxy were associated with physician visits or hospitalisation in the previous year. CONCLUSIONS Selected characteristics of the proxy respondent can influence responses to health surveys involving children. A minimum set of basic data should be collected from the proxy respondent to evaluate different patterns of response.