Background: The St. Jude Global Childhood Cancer Analytics Resource and Epidemiological Surveillance System (SJCARES) Registry is a St. Jude Global led hospital-based cancer registry network designed to promote registration of pediatric cancer patients internationally. Data is collected by participating sites' teams (Data Entry Specialists (DES) and MD Monitors (MD)) using standardized case report forms (CRFs). Patients are enrolled upon suspicion of cancer and followed for up to 5 years. Data is entered by sites’ DESs into a cloud-based platform where MDs check the data for completion and accuracy. St. Jude registry staff then perform systemic high-level quality monitoring (QM) process as described below.
Objective: To describe the implementation and performance of current systemic QM process within SJCARES registry network.
Methods: The QM process is performed retrospectively using four predefined categories: (1) open CRF specific queries (OQ), (2) delinquent or incomplete forms (DE), (3) potential duplicate patient records (PD), and (4) detection of extreme (EV) or invalid date entries, including negative values. The first category (OQ) highlights unresolved communication between sites' members, while the second category (DE) demonstrates Incomplete CRFs. Potential duplicates category (PD) manifest entries highlighted by the platform logic as possible duplicates entries due to similarity in certain datapoints such as the date of birth, Date of diagnosis and diagnosis ICD-O code. The final category examines the difference between specific dates for extreme values. The dates examined were Date of Initial Onset of Cancer-Related Symptoms, Date of First Medical Assessment, Date of diagnosis and Date of treatment start. These measures were used to ensure completeness of data, accuracy, and internal consistency and delivered to sites through monthly reports. The reports are generated based on data entered in the platform up to a specific cut-off date that occurs in the first week of the corresponding month. Office hours are conducted monthly to provide direct clarification for unique cases. Additional rounds of data revision are performed based on special circumstances such as mid study changes and data fields specific feedback.
?Results: 58 sites were onboarded in the SJCARES Registry by Jan 1, 2025, 49 of which were actively entering data with 18,241 patient entries and 4,319 total queries (534 OQ, 1020 DE, 752 PD and 2013 EV). These 51 sites finished the year with 28,449 patients entered and 3931 total queries (284 OQ, 1279 DE, 179 PD, 2189 EV) as of Jan 14, 2026. Two sites with low staffing and high volume of patients contributed to around 40% of total queries (1561). Additional barriers included staff turnover and lack of electronic medical records. The definition of Date of Diagnosis also presented a challenge as different sites interpreted the definition based on their site standard practice which was addressed by developing a next available hierarchy and a supplemental flowchart for additional clarity.
Conclusion: A structured retrospective data QM process is critical to support data completeness and consistency across a rapidly expanding pediatric cancer registry while highlighting the need for targeted support at high-volume, limited resources sites.