Quality and accessibility of liquid biopsy information: An analysis of online content.

Abstract

11010 Background: Circulating tumor DNA assays, also known as liquid biopsies, are increasingly used by oncologists to evaluate residual disease, cancer genomic composition, resistance mechanisms, and guide therapy selection. Since patients often use the internet as a source of medical information, we evaluated the quality, accessibility, and understandability of publicly available online information on liquid biopsies. Methods: We performed a cross-sectional study of 100 videos and 100 webpages that resulted from the search of “liquid biopsy” on a commonly used search engine (Google) and social network (YouTube). Videos over 5 min, scientific articles, and non-English content were excluded. Videos were evaluated for content, intended audience, and perceived sex and race/ethnicity of speakers. Two independent reviewers evaluated understandability and quality of health information using two validated tools for consumer health information (PEMAT, DISCERN). Any disagreements were resolved by a 3rd independent reviewer. Webpages were evaluated for accessibility (use of visual aids, other languages), quality of information (authors, citations, indications, limitations, cost), reading level, and presence of commercial bias. Reading level was evaluated using the validated Flesch-Kincaid (FK), Gunning Fog (GF), and SMOG indices. Data was summarized with descriptive statistics. Results: We reviewed 100 videos, with 190 median views per video (IQR: 75-719). Most videos were intended for a scientific audience (72%) and lacked diversity of speakers. The majority of featured speakers were perceived as male (65%) and non-Hispanic white (72%). Overall, liquid biopsy videos were complex and low quality. Only 9% of videos were understandable and 8% presented moderate to high quality health information. Furthermore, we reviewed 100 webpages about liquid biopsies; key characteristics are summarized in the table. About a third of webpages included indications (36%) and limitations (35%) of liquid biopsies. Webpages had a median text reading level of 14.1 (IQR: 12.1-15.8), 16.4 (IQR: 14.4-18.6), and 15.9 (IQR: 13.9-16.9) based on FK, GF, and SMOG indices, respectively. Only 2 websites had one index reading level of 8th grade or lower as recommended by the NIH and CDC. Conclusions: Online content about liquid biopsies is complex and requires education levels exceeding the reading level of the general US population. Extensive shortcomings in sources of unbiased and understandable information for patients with cancer exist. As novel diagnostic technologies and treatments emerge, efforts should be made to develop understandable, unbiased, and high-quality information resources for patients with cancer. Characteristics of liquid biopsy webpages. [Table: see text]