Abstract Body: Introduction
Limited multi-language access in healthcare is a public health concern as language barriers can contribute to lower quality care, miscommunication, and adverse health effects. Language barriers also reduce engagement with digital health technologies, such as remote blood pressure monitoring (RBPM), often developed for English speakers.
Hypothesis
There is lower engagement in postpartum RBPM and higher occurrence of postpartum readmission among non-English speakers than English speakers.
Methods
In 2020, a large safety-net hospital implemented a postpartum RBPM program for patients with hypertensive disorders of pregnancy, including chronic or gestational hypertension, preeclampsia, and hypertension during delivery hospitalization. Program materials and messaging are available in several languages including Spanish, Haitian Creole, and English. Patients are instructed to take their BP daily up to 6 weeks postpartum using a cell-enabled cuff that automatically uploads readings to a provider-facing portal. Using health records, we abstracted data including race, ethnicity, nativity, and language as well as healthcare utilization through 12 months postpartum. Program engagement was examined by preferred language during the six-week postpartum period using two measures: frequency, defined as the number of days with ≥1 BP reading, and duration, defined as the last day with a recorded reading. We also examined the proportion of patients readmitted to the hospital by preferred language within 12 months postpartum.
Results
Between January 2021 and June 2023, 2,200 patients were enrolled in the RBPM program of which 94% (n=2,069) provided ≥2 BP measures. Frequency (days with BP measures) and duration (last day with a BP measure) of program engagement were higher among Spanish (mean: 15.7; 31.7) and Haitian Creole (mean: 17.4; 34.1) speakers compared to English speakers (mean: 13.6; 29.9). Approximately 7% (n=153) of patients were readmitted within 12 months postpartum. Readmission rates were highest among English (8.3%) and Haitian Creole (6.8%) speakers and lowest among Spanish speakers (3.8%), though differences between language groups may be partially explained by variations in race, ethnicity, and nativity.
Conclusions
These findings suggest that when digital health programs are designed with intentional language inclusivity, non-English speakers can engage as effectively as English speakers which may translate to reductions in postpartum readmission rates.