Screening for and Experiences of Intimate Partner Violence in the United States Before, During, and After Pregnancy, 2016-2019

Abstract

Objectives. To measure rates of intimate partner violence (IPV) screening during the perinatal period among people experiencing physical violence in the United States. Methods. We used 2016-2019 Pregnancy Risk Assessment Monitoring System data (n = 158 338) to describe the incidence of physical IPV before or during pregnancy. We then assessed the prevalence of IPV screening before, during, or after pregnancy and predictors of receiving screening among those reporting violence. Results. Among the 3.5% (n = 6259) of respondents experiencing violence, 58.7%, 26.9%, and 48.3% were not screened before, during, or after pregnancy, respectively. Those reporting Medicaid or no insurance at birth, American Indian/Alaska Native people, and Spanish-speaking Hispanic people faced increased risk of not having a health care visit during which screening might occur. Among those attending a health care visit, privately insured people, rural residents, and non-Hispanic White respondents faced increased risk of not being screened. Conclusions. Among birthing people reporting physical IPV, nearly half were not screened for IPV before or after pregnancy. Public health efforts to improve maternal health must address both access to care and universal screening for IPV. (Am J Public Health. 2023;113(3):297-305. https://doi.org/10.2105/10.2105/AJPH.2022.307195).

FIGURE 1—

Adjusted Differences in Characteristics of Patients Not Receiving Intimate Partner Violence Screening in the Preconception Period Among Patients Reporting Physical Violence Who (a) Were Not Screened Because of No Visit, and (b) Visited but Were Not Screened: United States, Pregnancy Risk Assessment Monitoring System, 2016–2019 Note. CI = confidence interval. The sample size was n = 6259. Model also adjusted for maternal age, education, marital status, parity, and prepregnancy conditions (obesity, diabetes, high blood pressure/hypertension, smoking, or depression). Percentage point difference estimates were based on multivariable logistic regression results. Values for percentage point differences are provided in Table C (available as a supplement to the online version of this article at http://www.ajph.org). Adjusted predicted probability values from which differences were calculated are provided in Table D (available as a supplement to the online version of this article at http://www.ajph.org).

FIGURE 2—

Adjusted Differences in Characteristics of Patients Not Receiving Intimate Partner Violence Screening During Pregnancy Among Patients Reporting Physical Violence Who (a) Were Not Screened Because of No Visit, and (b) Visited but Were Not Screened: United States, Pregnancy Risk Assessment Monitoring System, 2016–2019 Note. CI = confidence interval. The sample size was n = 6259. The model also adjusted for maternal age, education, marital status, parity, and prepregnancy conditions (obesity, diabetes, high blood pressure/hypertension, smoking, or depression). Percentage point difference estimates were based on multivariable logistic regression results. Values for percentage point differences are provided in Table C (available as a supplement to the online version of this article at http://www.ajph.org). Adjusted predicted probability values from which differences were calculated are provided in Table D (available as a supplement to the online version of this article at http://www.ajph.org).

FIGURE 3—

Adjusted Differences in Characteristics of Patients Not Receiving Intimate Partner Violence Screening in the Postpartum Period Among Patients Reporting Physical Violence Who (a) Were Not Screened Because of No Visit, and (b) Visited but Were Not Screened: United States, Pregnancy Risk Assessment Monitoring System, 2016–2019 Note. CI = confidence interval. The sample size was n = 6259. The model also adjusted for maternal age, education, marital status, parity, and prepregnancy conditions (obesity, diabetes, high blood pressure/hypertension, smoking, or depression). Percentage point difference estimates were based on multivariable logistic regression results. Values for percentage point differences are provided in Table C (available as a supplement to the online version of this article at http://www.ajph.org). Adjusted predicted probability values from which differences were calculated are provided in Table D (available as a supplement to the online version of this article at http://www.ajph.org).