Participation in survey research among mothers with a recent live birth: A comparison of mothers with living versus deceased infants - Findings from the Pregnancy Risk Assessment Monitoring System, 2016-2019

Background Despite high infant mortality rates in the United States relative to other developed countries, little is known about survey participation among mothers of deceased infants. Objective To assess differences in survey response, contact and cooperation rates for mothers of deceased versus. living infants at the time of survey mailing (approximately 2-6 months postpartum), overall and by select maternal and infant characteristics. Methods We analysed 2016-2019 data for 50 sites from the Pregnancy Risk Assessment Monitoring System (PRAMS), a site-specific, population-based surveillance system of mothers with a recent live birth. We assessed differences in survey participation between mothers of deceased and living infants. Using American Association for Public Opinion Research (AAPOR) standard definitions and terminology, we calculated proportions of mothers who participated and were successfully contacted among sampled mothers (weighted response and contact rates, respectively), and who participated among contacted mothers (weighted cooperation rate). We then constructed multivariable survey-weighted logistic regression models to examine the adjusted association between infant vital status and weighted response, contact and cooperation rates, within strata of maternal and infant characteristics. Results Among sampled mothers, 0.3% (weighted percentage, n = 2795) of infants had records indicating they were deceased at the time of survey mailing and 99.7% (weighted percentage, n = 344,379) did not. Mothers of deceased infants had lower unadjusted weighted response (48.3% vs. 56.2%), contact (67.9% vs. 74.3%) and cooperation rates (71.1% vs. 75.6%). However, after adjusting for covariates, differences in survey participation by infant vital status were reduced. Conclusions After covariate adjustment, differences in PRAMS participation rates were attenuated. However, participation rates among mothers of deceased infants remain two to four percentage points lower compared with mothers of living infants. Strategies to increase PRAMS participation could inform knowledge about experiences and behaviours before, during and shortly after pregnancy to help reduce infant mortality.