Improving population-wide collection of stage at diagnosis for childhood cancer: International collaboration and progress

Abstract

Background: There are huge international disparities in childhood cancer survival. The International Agency for Research on Cancer's Global Initiative for Cancer Registry Development has improved cancer registry coverage of the world's population, particularly in low-and middle-income countries (LMICs). However, for virtually all registries around the world, the challenge remains of collecting comparable and population-wide information on stage at diagnosis. This information is essential to understand and address disparities in outcomes. In response to this, a UICC-endorsed set of consensus guidelines for assigning stage for 16 of the most common types of childhood cancer was recently developed (the Toronto Guidelines), for use by population registries in both high and LMICs. Aim: To trial the Toronto Guidelines on a population-basis, and develop a structured protocol, suitable for global implementation, for collecting the required data elements and assigning cancer stage at diagnosis for childhood cancer. Methods: Using an innovative approach, data items as defined in the Toronto Guidelines were gathered from the medical record and entered electronically. Stage at diagnosis was assigned automatically using computer algorithms, thus reducing errors and maximizing consistency. Data collection and assignment of stage were incorporated into an online platform that was then trialed in the national childhood cancer registries of Australia and New Zealand for cases diagnosed between 2006 and 2014. Results: Stage at diagnosis was successfully assigned for 94% of all eligible patients (n51662) across both countries. In contrast, stage as recorded by the treating clinician was located in the medical record for only 39% of cases in Australia. Conclusion: Practical imple-mentationof the Toronto Guidelines has been highly successful todate and further testing is planned in LMICs. This approach has the potential to improve global epidemiologic monitoring ofchildhood cancer and lead tobetter understandingofthe reasons underlying disparities in outcome.L6271910662019-04-17 <br />