Adherence to Pregnancy Hepatitis B Care Guidelines in Women and Infants in the United States and Evaluation of Two Interventions to Improve Care

A Multicentre Hospital-based Study

Tatyana Kushner; Elianna Kaplowitz; Rena Mei; Chelsea Xu; Alex Acker; Emma Rosenbluth; Igbagbosanmi Oredein; Monika Sarkar; Norah Terrault; Meena Bansal; Kimberly A. Forde


J Viral Hepat. 2021;28(4):582-591. 

In This Article

Materials and Methods

We performed a retrospective cohort study using EHR chart review data of obstetrics patients at Mount Sinai Health System (MSHS) in New York (the Mount Sinai Hospital (MSH) and Mount Sinai West (MSW)), University of California San Francisco (UCSF), and the University of Pennsylvania Health System (UPHS). Institutional Review Board approval for this study was obtained at each site.

Patients were selected based on the following criteria: (i) positive HBV surface antigen (HBsAg) during pregnancy, (ii) delivery of the pregnancy at the study site, and (iii) diagnosis of pregnancy or delivery between 1 January 2015 and 30 June 2018. For eligible women with more than one pregnancy during this time period, their first reported pregnancy was included for analysis. Infants of mothers with HBV were identified through linkage of maternal and infant charts in the EHR.

Baseline demographic and clinical data collected from the EHR included age, race, ethnicity, country of origin, type of health insurance, parity, gestational age at delivery, select comorbidities, complications of pregnancy and provider type (attending, resident or fellow, NP, PA, midwife). Laboratory data included maximum HBV DNA level during pregnancy, hepatitis B serologic workup (HBeAg/HBeAb, HBcIgM, HBcAb total) and liver biochemical testing. Cirrhosis was determined based on APRI and FIB-4 scores, if laboratory data were available, and chart review for imaging characteristics and provider assessment.

Adherence to the following maternal care measures were recorded (see Table S1 for definitions): referral to an HBV specialist; presentation to HBV specialty appointment if referred; HBV DNA testing,[8] and initiation of AVT when indicated.[8] Adherence to the following infant care measures were recorded at two of the participating institutions (where paediatric and adult hospitals are within the same health system): administration of HBIG and three HBV vaccine doses and whether they were administered at the appropriate time.[16] Adherence to post-vaccine serologic testing (PVST) was evaluated.[17] Factors associated with the completion of all vaccines on time were evaluated.

Two interventions were evaluated to determine impact on adherence to maternal follow-up care.

  1. Best practice alerts (BPAs), automated alerts in the EHR to remind obstetricians to initiate AVT when HBV DNA >200,000 IU/mL, were implemented at UPHS on April 2016. Data on initiation of AVT prior to (2015-March 2016) and after (April 2016-April 2018) BPA implementation were analysed for any change in the proportion of patients initiating AVT when indicated.

  2. A co-located Women's Liver Clinic embedded in the Obstetrics clinic at Mount Sinai was initiated in September 2017, which included clear referral pathway shared with obstetricians. Changes in adherence to care were determined before and after co-Clinic implementation. Similarly, at UPHS, a dual obstetrics/infectious diseases clinic was started in October 2015. Historic data from prior to 2015 were used to determine changes in adherence to care guidelines after clinic implementation.[9]

Demographic and clinical characteristics were described using frequencies and percentages for categorical data and medians with interquartile ranges (IQR) or means with standard deviations for continuous variables. Changes in the proportion of women who adhered to maternal care before and after BPA and Liver Clinic interventions were analysed using chi-square or Fisher's exact tests, as appropriate. Univariable and multivariable logistic regression analyses were performed to identify factors associated with care. Rates of adherence were also plotted over time, and trends were assessed descriptively. All analyses were conducted using SAS version 9.4 (Cary, NC).