Response rate disparities in Maryland children oral health survey 2005-2006

Objectives: To assess the current oral health status of Maryland school children, a survey was conducted on children in kindergarten and 3rd grades sampled from the State's public schools between 2005 and 2006. Approximately 5,000 students from 50 Maryland schools were sampled using a multi-stage probability design. A questionnaire collected information on demo-social characteristics (region, grade level, gender, race, eligibility for free or reduced meals at school, caregiver's education level, dental insurance status, dental visit in the past 12 months, and dental care availability). An oral screening exam was performed among those who returned a consent form with questionnaire. Information on outcomes (dental caries and sealants) was collected. Methods: Overall response rates and response rates by demographics for questionnaire and screening exam were calculated respectively. Comparisons of rates by each demo-social characteristic were conducted using Fisher's exact test or chi-square test when appropriate. A p-value < 0.05 was considered statistically significant in all tests. All analyses used SAS software. Results: The overall response rate for the questionnaire was 46%. Approximately half (55%) of those who returned the questionnaire participated in the screening exam. Children who were non-Hispanic black, eligible for free/reduced meals, whose caregivers' education were lower, who had Medicaid, had no dental visit in the past 12 months and had no dental care access when needed were more likely to participate in the screening exam. The disparities, however, did not have statistically significant impacts on weighted estimates of caries and sealant prevalence in the study. Conclusions: This study identified response disparities by demo-social characteristics in the screening exam in the Maryland Elementary School Children Oral Health Survey. The disparities potentially introduce selection biases toward outcome estimates. This can be overcome by stratification technique at sample design stage and by post-stratification at analysis stage.