Framing the picture: Impact of radiographic imaging on parental vesicoureteral reflux management preferences

Given the variety of treatment options for vesicoureteral reflux (VUR), shared decision making between clinicians and parents is essential. Despite its importance, shared decision making is limited by the framing effect – people process the same information differently depending on how it is presented. Studies have also demonstrated that showing patients their radiology images can impact behaviors. In this pilot study, we sought to determine if showing parents radiographic images could serve as a framing tool that impacts the decision of whether to pursue surgery, endoscopic intervention, or conservative management for VUR. We designed a survey instrument which provided background on VUR and a hypothetical scenario of a 2-year-old child with VUR who had a breakthrough febrile urinary tract infection (UTI). Guideline-concordant management options were presented: (1) change antibiotics, (2) endoscopic management, or (3) open or laparoscopic surgery. All options were similarly presented regarding risks, benefits, and length of stay. Respondents were randomized into a group with no image accompanying the clinical scenario or a group which had a labeled image of a voiding cystourethrogram (VCUG) demonstrating unilateral VUR. Respondents also answered demographic and health experience questions. The instrument was published on Amazon's Mechanical Turk online work interface which provides reliable and validated results in VUR experiments. Parents aged 18–60 years old were eligible. Responses with failed attention questions, duplicate internet addresses, or submission times 30 min were disqualified. Data were analyzed using t-test, chi-square, and multinomial logistic regression. Sensitivity analyses were performed after excluding all responses submitted under 2, 3, and 5 min. There were a total of 914 responses, 426 met inclusion criteria. The presence or absence of a VCUG image did not result in a statically significant difference in the management decision (p = 0.081). Multinomial logistic regression demonstrated that prior UTI experience influenced the management decision (p = 0.027). Sensitivity analyses revealed a significant difference in the management decision when excluding responses