Sepsis typically results from a presumed or proven infection leading to organ dysfunction and/or death and is a leading cause of death in children in developing countries. In 2017, there were an estimated 48·9 million cases of sepsis worldwide. Children accounted for more than half of cases and 85% occurred in low- and middle-income countries. Most deaths from sepsis are preventable by early detection and treatment with simple, highly effective interventions (e.g., antimicrobials and fluid resuscitation). Thus, improved prioritization, coordination, and timely identification of critically ill children is the cornerstone in efforts to improve sepsis outcomes. The 2024 Pediatric Sepsis Data Challenge was developed to provide an opportunity to address this gap.
Brought to you by the Institute for Global Health, the Pediatric Sepsis Data CoLaboratory (Sepsis CoLab), the World Federation of Pediatric Intensive & Critical Care Societies (WFPICCS), Emory University, and PhysioNet, the challenge will open on November 4th, 2024.
The challenge is to design and implement a working, open-source algorithm that can predict in-hospital mortality from routinely collected data at the time of presentation to a health facility in Uganda. We are seeking a parsimonious model that provides the optimal trade-off between effort in data collection and performance in clinical risk prediction. The model should ideally be implementable in a mobile device in an environment where electrical supply and internet connectivity is unreliable.
The 2024 Pediatric Sepsis Challenge aims to support global participants in building skills in model development for clinical risk prediction. The challenge is open to teams of 2-5 members. We encourage the formation of multidisciplinary, international teams, with the hope of catalyzing new collaborations among experienced mentors and novice researchers and mentees. We invite anyone, from those new to data science to veterans of the field, to participate. Participants can include students, faculty, and individuals with various roles in non-profit, academic, government, and industry organizations.
The data for the 2024 Pediatric Sepsis Data Challenge comes from a deidentified, curated research dataset of a study called “Smart discharges to improve post-discharge health outcomes in children: A prospective before-after study with staggered implementation”. Participants will be provided with a synthetically generated training data set to reduce the risk of patient re-identification. The synthetic training set was generated from a random subset of the original data consisting of 2686 records (70% of the total dataset) 6-60-month-old children admitted with suspected sepsis from the community to the paediatric wards of Ugandan hospitals. All participant solutions will be ran against the remaining 1151 (30% of the total dataset) children that will be unseen and unavailable to participants.
There are two phases for the Challenge: an unofficial phase and an official phase. The unofficial phase of the Challenge allows us to introduce and ‘beta test’ the data, scores, and submission system before the official phase of the Challenge. Participation in the unofficial phase is mandatory for participating in the official phase of the Challenge because it helps us to improve the official phase. The unofficial phase also gives teams a chance to get acquainted with the data and variables and begin work on creating preliminary algorithms. Teams may submit up to 5 entries during the unofficial phase.
During the official phase, teams will have 10 opportunities over this period to submit an entry. After each round, teams are given results and can submit again. Only your final entry will count for ranking.
To maintain scientific impact of the challenge, it is important that all Challengers enter truly independent ideas. For this reason, we impose the following rules on team composition/collaboration:
- Multiple teams from a single entity (such as a company, university or even university department) are permissible as long as the teams are truly independent, and do not share team members (at any point), code or any ideas. Multiple teams from the same research group or unit within a company are not allowed because we do not believe true independence between teams can be maintained.
- New team members can join as long as they have not previously been in another team that year or have communicated with a team member from any other team at any point during the competition concerning this Challenge.
- You may use public code from another team if they posted it before the competition.
- You may not post your code publicly during the competition, or use other competitor's code posted during the competition, whether it was intentional or not.
- You may not publicly post information describing your methods (blog, code, preprint, magazine article, vlog, presentation, talk, etc.) or give a talk outside your own research group at any point during the competition that reveals the methods you have employed or will employ in the Challenge.
- You must use the same team name and email address throughout the course of the competition, and the email address should be the same as the one used to register. If your team has used multiple team names and/or email addresses to enter the Challenge, please contact the organizers immediately to avoid potential disqualification.
Registration will open on July 15th, 2024 in advance of the November 4th open date. In the meantime, we invite you to build a team and hone your data science skills! If you do not have a team you can register as a solo challenger. We will offer a "matchmaking" option at registration where we will team you up with other solo challengers.
If you would like to be sent a message when registration is live, please send an email to Charly Huxford (charly.huxford@bcchr.ca).
