Understanding and preferences regarding risk communication during pregnancy: a survey to facilitate provider communication with patients

BACKGROUND: Clear communication of medical risk helps to ensure proper patient understanding of healthcare options and supports informed decision-making. Communication involving visual and written risk typically conveys risk more effectively than conversations alone between a patient and a clinician. However, perception of risk is context-dependent, and the efficacy of and preferences for commonly-used risk communication for-mats are not well-understood during pregnancy, which is a time of com-plex decision-making. We sought to address this knowledge gap. OBJECTIVE: This study aimed to assess pregnant and recently preg-nant people's understanding and preferences for different risk communi-cation formats. STUDY DESIGN: We conducted an open online REDCap survey of pregnant and recently pregnant people over a 1-month period in 2022. Study participants were aged 16 to 49 years, pregnant or recently pregnant, and able to provide informed consent in English. Data collected included demographics, measurements of accuracy of understanding including both gist accuracy (general understanding) and verbatim accuracy (numeric quan-tification), and preferences for risk communication formats including icon arrays, pie charts, bar graphs, and text. Descriptive analyses of the propor-tion of correctly answered questions were calculated. RESULTS: A total of 247 participants completed >= 1 item on accuracy and risk communication preferences, and 230 provided complete responses. Gist (general) understanding was accurate between 74% and 89% of the time for most graphical formats. Verbatim understanding (exact numeric quantification) was approximately 90% accurate for most formats. Respondents preferred that figures be used over circles to display risk in icon arrays, both for themselves and for infants, although figures generated more worry. However, participants substantially preferred pie charts over bar graphs (59%-70% vs 19%-25%). Respondents pre-ferred risk to be expressed with a lower denominator of 200 rather than a higher denominator of 1000 (79% vs 13%, although the lower denomina-tor generated more worry), and in terms of chance of survival rather than chance of death (50% vs 33%). CONCLUSION: In a survey of pregnant and recently pregnant people, most respondents preferred pie charts over other graph formats, and lower rather than higher denominators in text. Presentations of survival rather than death estimates were also preferred. Approximately 75% to 90% of respondents accurately understood risk presented with visual and written communication. For the remaining participants, for whom accurate understanding was challenging, new strategies need to be developed.