ESPE Abstracts

Background: The internet provides a multitude of health information. Understanding disorders/differences of sex development (DSD) can be difficult for families partly due to their complexity and relatively low prevalence. Consequently, families may use the internet to gain understanding of their child's condition, however the quality of this information has not been formally assessed.

Aims: To assess the quality, validity and accuracy of website health information concerning commonly searched terms related to DSD.

Methods: Our multi-professional DSD Team and parents of current patients at Bristol Royal Hospital for Children were consulted to understand commonly searched terms. Five search terms were selected "Disorders of Sex Development OR Differences of Sex Development", "Congenital Adrenal Hyperplasia" (CAH), "Ambiguous Genitalia OR Atypical Genitalia", "Cliteromegaly OR Clitoromegaly" and "Micropenis". The top 20 Google searches were scored by two independent reviewers using the validated QUality Evaluation Scoring Tool (QUEST). The tool scores 6 domains (Authorship, Attribution, Conflict of Interest, Currency, Complementarity and Tone) with 28 the maximum score. Only websites meeting the inclusion criteria: article/information-like leaflet format, in English, no payment/login required, and articles considering aetiology/diagnosis/treatment of disorder were included.

Results: Inter-rater reliability demonstrated substantial agreement across all domains (Cohen's kappa=0.71-0.83) except 'Tone' (Cohen's kappa=0.55, moderate agreement). Thirty-two percent of searches were excluded due to the website either being a tabloid article (6%, only with micropenis searches), in an inappropriate format (10%) or the site had a paywall (16%). There were no significant differences in scores across all 5 search terms; DSD (n=14, mean=18.7 SD=5.6), CAH (n=14, mean=16.7, SD=3.4), Ambiguous Genitalia (n=14, mean=14.1, SD=4.9), Cliteromegaly (n=12, mean=19.6, SD=6.0) and Micropenis (n=12, mean=17.2, SD=7.4). There was no evidence that overall QUEST score was related to search rank. There was evidence that average scores were related to website category with hospital websites scoring the lowest. Compared to hospital websites (n=21, mean score=11.8) scores from health information e.g. Healthline (n=17, mean score=18, P=0.001), general information e.g. Wikipedia (n=4, mean score=21, P=0.003), peer reviewed publication (mean score=22.1, P<0.001) websites were significantly higher.

Conclusion: Our study demonstrates further validation of the QUEST Tool with good interindividual agreement. A high proportion of articles searched are either not accessible or from tabloid sources. The more colloquial term 'Micropenis' produced variable information quality. The lowest quality information came from hospital websites and we would recommend professionals considering the quality criteria in the QUEST tool when designing health information websites.