E2402. Patient’s Perceptions on Sharing Information About Imaging Examinations: A Pilot Study
  1. Joás Gonçalves; State University of Ponta Grossa
  2. Arthur Pinto; State University of Ponta Grossa
  3. Cassiano Ianke; State University of Ponta Grossa
  4. Pedro Joaquim; State University of Ponta Grossa
  5. Pascoal Adura Filho; São Camilo Group - Diagnostic Medicine
  6. Marcelo Schafranski; State University of Ponta Grossa
  7. Marcelo Dreweck; State University of Ponta Grossa
A questionnaire was developed for a cross-sectional pilot study, aiming to examine the process of sharing information about imaging examinations focusing on patient’s perspective. Our purpose was to evaluate the reliability of this instrument while gathering an initial profile of local patients.

Materials and Methods:
Survey included multiple choice questions about demographics and personal history related to imaging examinations, and Likert scale (LS) questions about physician-patient interaction and valued information before the procedure. LS items comprised a range of 0 to 5 representing negative and positive connotation respectively, and descriptive subtitles were exhibited under the extremities. These questions were pooled in five clusters (one about quality of physician-patient interaction and 4 about valued information) and the sum of isolated answers was stratified only for descriptive purposes. Consecutive patients were asked to participate by front-desk staff members while waiting before their imaging examination during 9 days of May 2019. LS items reliability was evaluated by internal consistency analysis, so Cronbach’s alpha (a) and McDonald’s omega (omega) were computed for questions about interaction with the ordering provider (domain A) and valued information before the procedure (domain B). A threshold value of 0.70 was considered to indicate a good internal consistency for both tests. Descriptive statistics were used to summarize the respondents’ profile. Clusters of LS items were treated as continuous variables and compared by using Kruskal-Wallis test. A P value <.05 was considered to establish statistical significance. Statistical assessment was executed using R version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria, 2020).

A total of 188 responses returned and 49 were excluded to accomplish a final sample of 139 (63.3% completely vs. 36.7% partially responded). Respondents mean age was 43.3±14,7(SD), mostly women (57.6%), white people (62.6%), with no more than 8 years of education attained (55.3%). For questions in domain A, a was 0.85 and omega=0.87, indicating a good internal consistency of this items. We also obtained a=0.78 and omega=0.79 for questions in domain B. Patients most often reported a good interaction with the ordering provider (63.3%) and also expressed a greater value to information about examination preparation compared to the other clusters (P<.001). Post hoc analysis revealed that option “5” was the most common answer for all LS items (range, 42.4-89.2% of responses for isolated questions) with a significantly greater frequency (median, 88; Q3-Q1, 109-81; P<.001).

In conclusion, LS items showed a good internal consistency for both domains and acknowledged the instrument reliability. Respondents revealed a greater value to information about examination preparation, which is supported by previously reported data. Response distribution reflected positive skewness inherent to LS items, and it has been possibly reinforced by exhibited subtitles. Adjustments on survey structure may improve its good reliability.