DOP60 Self-reported treatment effectiveness for Crohn's Disease using a novel crowdsourcing web-based platform
Engel, T.(1);Dotan, E.(2);Ben-Horin, S.(1);Kopylov, U.(1);
(1)Tel-HaShomer Sheba Medical Center, Department of Gastroenterology, Ramat Gan, Israel;(2)Stuffthatworks, Company, Tel Aviv, Israel;
Background
The world wide web and social media platforms have become an unprecedented source for sharing self-experience, potentially allowing the collection and integration of health data with patient experience
Methods
StuffThatWorks (STW) is an online open platform that applies machine learning and the power of crowdsourcing where patients with chronic medical conditions can self-report and compare their individual outcomes using a structured online questionnaire. The present study analyzed de-identified self-reported personalized comparative treatments' effectiveness for CD. The design was a cross-sectional, international, crowdsourcing, questionnaire and AI web-based study of patients with Crohn's self-reporting their outcomes by 06/11/21. A proprietary STW Bayesian inference model was built to measures the level of improvement in condition severity and clinical indicators for each treatment and ranks treatment effectiveness. A linear regression model was used to examine co-variate association with the current condition severity as the outcome. Finally, the effectiveness of first-line biological treatments was analyzed by multiple treatment comparisons model and by calculating odds ratio and 95% confidence intervals for each treatment pair.
Results
A total of 5898 self-reported CD patients were included for the analysis. Most participants were female (76.13%) and from English speaking countries (91%). Overall, anti TNF drugs were the most reported tried treatment (71.97%) followed by steroids (46.22%) and diet (43.8%). Among Biologic therapy (BT) tried by STW CD users, Infliximab (IFX) and Adalimumab (ADA) were ranked most effective by the STW effectiveness model, by change in condition severity on 1-5 scale (mean change of 1.19 points, estimated lower-upper bounds 1.10-1.28), followed by Ustekinumab (UST) (mean 1.07, bounds 0.88-1.26), and Vedolizumab (VDZ) (mean 0.96, bounds 0.74-1.17). Bowel surgery (mean 1.19, bounds 0.98-1.40), and cannabis (mean 1.01, bounds 0.72-1.31) were also among the top 5 most effective treatments together with BT. The odds ratio was calculated for each BT pair, with IFX more effective than ADA, UST and VDZ (OR 2.34 (CI 1.88 – 2.80), 3.08 (CI 2.44 – 3,73), 7.16 (CI 6.53 – 7.80), respectively), ADA more effective than UST and VDZ (OR 1.32 (CI 0.74 - 1.9), 3.06 (CI 2.49 - 3.63), respectively), and UST more effective than VDZ (OR 2.32 (CI 1.6 - 3.05)).
Conclusion
We present the first online crowdsourcing platform-based study of treatment self-reported outcomes in CD. Net-based crowdsourcing patient-reported outcomes' platforms can potentially help both clinicians and patients select the best treatment for their condition.