Childhood cancer incidence and survival as population-based indicators in a geographically isolated population: Evidence from Madeira, Portugal

Background Childhood cancers are rare, and evidence on childhood cancer from small and geographically isolated populations remains limited. This study describes long-term patterns of childhood cancer incidence, mortality, and survival in the Autonomous Region of Madeira, Portugal. Methods A retrospective population-based study was conducted using cancer registry data for all residents aged 0–14 years diagnosed between 1998 and 2022. Tumors were classified according to the International Classification of Childhood Cancer, third edition (ICCC-3). Age-standardized incidence (ASIR) and mortality rates (ASMR) were estimated and compared with national, regional, and global reference populations. Temporal trends were evaluated using Joinpoint regression adapted for small numbers. Overall survival was estimated using the Kaplan–Meier method and compared with National and International benchmarks by diagnostic group, sex, and diagnostic period (1998–2010 vs. 2011–2022). Results A total of 183 childhood cancer cases were identified. Overall ASIR was within the range reported for selected high-income reference populations, although these comparisons should be interpreted as contextual benchmarks because registry practices, standardization approaches, and reference periods differed. No statistically significant temporal change in incidence was detected. Childhood cancer mortality was low for the study period. Overall, 5-year survival improved significantly, increasing from 65.3% in 1998–2010 to 78.6% in 2011–2022. Survival differed markedly by diagnostic group, with the highest survival observed for lymphomas and the lowest for central nervous system (CNS) tumors. For the most recent period (2011–2022), 5-year survival in Madeira was broadly consistent with national and European benchmarks, although poorer outcomes were observed for CNS malignancies. Conclusions Childhood cancer incidence and survival patterns in Madeira were broadly consistent with those reported in selected high-income reference populations, but external comparisons should be interpreted as contextual benchmarks rather than evidence of equivalence. Findings should be interpreted considering the statistical instability and measurement uncertainty inherent to childhood cancer surveillance in small populations. These findings demonstrate the challenges and opportunities of measuring cancer outcomes in small populations and support the cautious use of surveillance indicators for benchmarking and service evaluation.