Analysis of the quality, accuracy, and readability of patient information on polycystic ovarian syndrome (PCOS) on the internet available in English: a cross-sectional study

Abstract

Abstract Background Online information about PCOS lacks reliability for patients seeking information about the disease. Thus, we aimed to perform an updated analysis of the quality, accuracy, and readability of patient information on PCOS available online. Methods We conducted a cross-sectional study using the top five Google Trends search terms in English associated with PCOS, including “symptoms,” “treatment,” “test,” “pregnancy,” and “causes.” Five separate searches in Bing, Yahoo, and Google were performed to obtain the first 10 unique webpages for each term that was categorized as commercial, non-profit organization, scientific resources, or private foundation. We used the 16-item DISCERN with Likert-responses (minimum 1, maximum 5) where the total is 80 and lowest is 16, clarity with the 32-item EQIP, where responses of no = 0 and yes = 1 (minimum 0, maximum 32), and accuracy scores with 1 denoting poor and 5 completely accurate information; low scores of each corresponded to poorly reported information. We assessed readability with Flesch-Kincaid reading ease index, where higher scores correspond to reading ease, and lower grades correspond to easier readability with Flesch-Kincaid grade level, Gunning-Fog, Coleman-Liau index, automated readability index, New Dale-Chall Readability, and simple measure of gobbledygook. We additionally assessed word and sentence characteristics. We used Kruskal-Wallis test to compare scores according to webpage categories. Results Out of 150 webpages, most were commercial (n = 85, 57%), followed by non-profit organizations (n = 44, 29%), scientific resources (n = 13, 9%) and private foundations (n = 6, 4%). Google webpages had higher median DISCERN score ([Md] = 47.0) than Bing ([Md] = 42.0) and Yahoo ([Md] = 43.0) webpages; P = 0.023. No difference in EQIP scores according to search engine was found (P = 0.524). Predominantly, webpages from private foundations had higher DISCERN and EQIP scores, although comparisons were not statistically significant (P = 0.456) and P = 0.653.). Accuracy and readability were similar across search engines and webpage categories (P = 0.915, range 5.0–5.0) and (P = 0.208, range 4.0–5.0). Conclusions Quality and clarity of the data were fair according to search engine and category. Accuracy of information was high, showing that the public may encounter accurate information about PCOS. However, the readability of the information was high, reflecting a need for more readable resources about PCOS.