94.08 Dr. Google: The Readability and Accuracy of Patient Education Websites for Graves’ Disease Treatment

A. Purdy1, A. Idriss1, S. Ahern1, B. Van Blarigan1, J. Bulter1, D. Elfenbein1  1University Of California, Irvine,Orange, CA, USA

Introduction:

National guidelines for the treatment of Graves’ hyperthyroidism emphasize the importance of incorporating patient preferences into treatment recommendations. In order for patients to express a true preference, they must have an understanding of their options. Today, many turn to the internet as a source of information. Most patients are initially treated with anti-thyroid drugs (ATD), and then ultimately may choose between radioactive iodine ablation (RAI), or surgery for definitive treatment. Our primary objective was to compare the readability and accuracy of patient-oriented online resources for Graves’ Disease by treatment modality and website.

Methods:

A systematic internet search for treatment of Graves’ Disease was used to identify the most popular websites that discussed all 3 treatment modalities. Readability was measured by 9 standardized tests, and the median readability scores were compared among treatment modalities and websites. Accuracy was assessed by an expert panel, consisting of two endocrine surgeons and three endocrinologists. Raters were asked to score the accuracy of documents on a scale of 1-5. Finally, percentage of space on the website dedicated to each of the three treatment options was calculated.

Results: We identified 11 websites that ranked highly using every search strategy and search engine used, and included 2 lower ranking but informative websites from professional organizations. Of the 13 sites, 2 were authored by academic institutions, 2 by government agencies, 5 by non-profit, and 4 by private entities. Readability varied between sites from an 8th to a 13th grade level. The websites differed in the amount of space dedicated to each of the 3 treatment modalities, with the most space dedicated to discussing RAI (mean=41%). There was overall fair to moderate agreement among expert reviewers about the accuracy of the information presented, (Intraclass Correlation Coefficient=0.3-0.41). Accuracy as assessed by our expert panel ranged from a mean of 2.75 (out of 5) for the least accurate to 4.5 for the most accurate. The two least accurate websites overall were authored by private entities; two of the top three most accurate were academic institutions and the third is a popular collaboratively-written website.

Conclusion:

Information that patients obtain from the internet may help shape their preferences for certain treatments before they even see a health care provider. Our analysis found that for Graves’ disease treatments, the most inaccurate websites rank high using traditional internet search methods, and some highly accurate patient education websites were only found using strategies the average patient likely will not use, such as searching by professional organization name. As treatment recommendations are constantly evolving, professional organizations and academic centers must take steps to make sure patients have access to the most accurate information for treatment decisions.