Background: Accurate childhood cancer survival estimates are critical for policy-makers and clinicians for priority-setting and planning decisions. However, observed survival estimates are lacking for many countries, and where available, wide variation in outcomes is reported. Understanding the barriers to optimizing survival can help improve childhood cancer outcomes. We aimed to provide estimates of global childhood cancer survival, accounting for the impact of multiple factors that influence cancer outcomes. Methods: We developed a microsimulation model to simulate childhood cancer survival for 200 countries/territories worldwide, taking into account clinical and epidemiologic factors, including country-specific treatment variables, such as availability of chemotherapy/radiation/surgery. To ensure model results were consistent with reported survival data, we calibrated the model to estimates from the CONCORD 2 and 3 studies using an Approximate Bayesian Computation approach. We estimated five-year net survival for diagnosed childhood cancer cases in each country/territory and estimated potential survival gains if seven policy interventions focused on improving treatment availability and delivery were implemented in isolation or as packages. Findings: Our model estimates that global five-year net childhood cancer survival is currently 37•4% (95% uncertainty interval [UI] 34•7%-39•8%), with large variation by region, ranging from 8•1% (95% UI 4•4%-13•7%) in Eastern Africa to 83•0% (95% UI 81•6%-84•4%) in North America. Among the seven policy interventions modeled, each individually provided limited gains, increasing global five-year net survival to between 38•4% and 44•6%. When bundled into packages of interventions that either improved service delivery or expanded treatment access, five-year net survival increased to 50•2% (95% UI 47•3%-53•0%) and 54•1% (95% UI 50•1%-58•5%), respectively. A comprehensive systems approach consisting of all policy interventions yielded super-additive gains with global five-year net survival of 53•6% (95% UI 51•5%-55•6%) at 50% scale-up and 80•8% (95% UI 79•5%-82•1%) at full implementation. Interpretation: Childhood cancer survival varies widely by region, with especially poor survival in Africa. While expanding access to treatment (chemotherapy/radiation/surgery) and addressing financial toxicity are essential, investments that improve the quality of care, at both the health system and facility-level, are needed to improve childhood cancer outcomes globally.
