How can PRAMS survey response rates be improved among American Indian mothers? Data from 10 states.

OBJECTIVES: To examine the low response rates to the Pregnancy Risk Assessment and Monitoring System (PRAMS) survey for American Indian (AI) mothers by comparing characteristics of AI participants, AI non-participants, non-Hispanic White (NHW) participants, and NHW non-participants. METHODS: We analyzed 2000-2002 data from states whose population was at least 5% AI or Alaska Native (AN) (n = 10). Mothers who returned a questionnaire (regardless of completion) or who spoke by telephone with PRAMS personnel were defined as contacts. Mothers who completed a PRAMS questionnaire were defined as respondents. We described overall and state-specific maternal characteristics from birth certificates for both those who were contacted and those not contacted and computed contact and response rates. RESULTS: Contact rates varied by state, ranging from 49% to 92% for AI and AN mothers and 82-93% for NHW mothers. However, once contacted, most mothers completed a questionnaire (85-99%). Both AI and NHW mothers were less likely to be contacted if they were <29 years of age, unmarried, multiparous and had