LAUSR

OBJECTIVE Hearing loss is increasingly prevalent among younger adults, impacting health and health care use. Deaf and hard of hearing (DHH) women have a higher risk of chronic diseases, pregnancy complications and adverse birth outcomes compared to hearing women. Health care utilization patterns during the perinatal period, remains not well understood. The objective was to examine differences in antenatal emergency department and inpatient utilization among DHH and non-DHH women. STUDY DESIGN We conducted a retrospective cohort study design to analyze 2002-2013 Massachusetts Pregnancy to Early Life Longitudinal data to compare antenatal inpatient and emergency department use between DHH (N=925) and hearing (N=2895) women with singleton deliveries. Matching was done based on delivery year, age at delivery and birth parity in 1:3 case-control ratio. Demographic, socioeconomic, clinical and hospital characteristics were first compared for DHH mothers and the matched control group using chi-squared tests and t-tests. Multivariable models were adjusted for socio-demographic and clinical characteristics. RESULTS Among DHH women (N=925), 49% had at least one emergency department visit, 19% had an observational stay, and 14% had a non-delivery hospital stay compared to 26%, 14%, and 6%, respectively among hearing women (N=2895) during the antenatal period (all p<0.001). The risk of non-delivery emergency department visits (RR 1.58; p<0.001) and inpatient stays (RR 1.89; p<0.001) remained higher among DHH women compared to hearing women even after adjustment. Having four or more antenatal emergency department visits (7% vs. 2%) and two or more non-delivery hospital stays (4% vs. 0.4%) were more common among pregnant DHH women compared to their controls (all P<0.001). CONCLUSION(S) The findings demonstrate that DHH women use emergency departments and inpatient services at a significantly higher rate than their hearing controls during the antenatal period. A systematic investigation of the mechanisms for these findings are needed.