PC1088: A Mixed Method Study of Site Level Barriers to Prenatal Ultrasound Screening for Congenital Heart Defects at Community and Tertiary Imaging Sites

Background: Population screening for congenital heart defects (CHD) with fetal ultrasound (US) has an expected efficacy of >80%; however, prenatal detection remains below 50%. Prior work has identified key barriers to effective screening among performing providers. We sought to identify contextual site-specific barriers/facilitators for screening within representative imaging settings at a regional healthcare organization that cares for ~30,000 pregnant women per year.

Methods: We undertook an explanatory, sequential mixed-methods design (qualitative → quantitative) among 3 imaging sites that performed >50 fetal anatomy US per year: a radiology department with a community hospital (Rad), a high-volume obstetric practice (OB) with onsite US, and a maternal fetal medicine (MFM) diagnostic center. Qualitative data included stakeholder interviews and semi-structured site assessments which were coded using a rapid directed content analysis method. This coding procedure aggregated common versus divergent themes across the 3 sites. Quantitative data included site characteristics (volume, number of providers), patient demographics, and image review of 50 randomly selected fetal anatomy scans per site. Cardiac screening images were assessed using a previously validated image score (range: 0-11). Factors associated with image adequacy and scores were analyzed using multivariable regression modeling.

Results: Interviews and site assessments identified 3 main themes for barriers to screening: intrinsic (confidence, experience), extrinsic (feedback, pressure to perform), and organizational (scan-time, volume, interdisciplinary collaboration). Barriers/facilitators varied by clinical setting (Table 2). MFM sonographers reported confidence in their ability to detect CHD by prenatal US, felt well supported, received frequent feedback, but felt pressure to perform. In contrast, sonographers at Rad and OB sites lacked confidence, felt less supported, and were more overwhelmed by volume and time constraints. OB and Rad sonographers described limited ability to advance skills because of lack of feedback and ongoing education. Sites differed by volume of anatomy scans and patient demographics (Table 1). Cardiac image adequacy and image score varied significantly by imaging site, storage mode, and gestational age, but not by sonographer volume or patient BMI on univariable analysis. OB and Rad imaging scores were significantly lower than MFM scores on multivariable analysis (-1.71 and -2.19 points lower, respectively, p < 0.001 for all). Increasing gestational age was associated with higher imaging scores (0.12-point increase per week gestation, p=0.001).

Conclusion: This study reinforces previous reports of differences in cardiac screening quality based on imaging site. However, qualitative assessments revealed new modifiable process characteristics that may be future targets for improvement. Volume of sonographer scans was not related to imaging quality, despite previous reports of volume as a key factor. There are likely other organizational differences by site related to quality, including differences in collaborative feedback and education. The results emphasize the importance of improving sonographer support, education, and feedback in all sites to improve detection of CHD.