Using a challenge in Cochrane Crowd to involve medical students in evidence production

Tags: Oral
Pérez-Gaxiola G1, Leyva De Los Ríos CD2, Rodríguez Páez IV2, Requejo-Rivera EV2, Alapizco Castro G2
1Cochrane Mexico, Sinaloa's Pediatric Hospital, 2Autonomous University of Sinaloa

Background: Cochrane Crowd was launched in 2016 as a platform to help anyone get involved in evidence production by offering microtasks such as identifying clinical trials. Cochrane Classmate is a tool that permits trainers and teachers to use Cochrane Crowd tasks and set up challenges with students. There have been previous open challenges in past Colloquia and events such as the March for Science day.

Objectives: to use a challenge in Cochrane Crowd to involve medical students in Cochrane and in evidence production.

Methods: a proposal to set up a university-wide Cochrane Crowd challenge was presented to the academic directors of the Autonomous University of Sinaloa, the largest public college in our region, with the co-ordination of the Cochrane Associated Centre at Sinaloa’s Pediatric Hospital. The challenge was presented as an opportunity for students to get involved with Cochrane and to learn about clinical trials, and was set up as a competition. After approval, we presented the plan to professors and medical students' representatives. We sent a university-wide invitation using social media (Facebook and WhatsApp) and registration began 40 days prior to the challenge. We had four hands-on training sessions for teachers and students, nine short classroom presentations with students, and one open webinar. The microtask in Cochrane Crowd used for the challenge was the identification of randomized controlled trials (RCT). There was no funding for this activity.

Results: the challenge, called 'RCT marathon' in Spanish, was held from 5 December 2019 at 9:00 h to 7 December 2019 at 23:59 h. Of 738 people that signed up, 455 participated. 93% were medical students, the rest were physiotherapy students, nutritionists and teachers. Most were third-year medical students. Participants made 319,643 individual assessments equating to 89,692 records being screened; they identified 8244 RCTs. For participants with over 1000 classifications, average sensitivity and specificity to identify RCTs were 92.3% and 98.4, respectively.

Conclusions: this university-wide challenge was successful in involving hundreds of medical students in the identification of thousands of RCTs within the Cochrane Crowd platform.