Artificial Intelligence in Pediatric Oncology in Africa: A Survey of Awareness, Use, and Readiness Amongst Healthcare Workers

Abstract

INTRODUCTION: Artificial intelligence (AI) has the potential to enhance oncology diagnostics, treatment planning, and patient monitoring. In pediatric oncology, AI can support both clinical care and research. The roles and awareness of AI in African pediatric oncology units are unknown. We aimed to assess knowledge and utilization of AI in pediatric oncology across Africa. METHODS: A cross-sectional online survey was conducted during July and August 2025, using a structured questionnaire in English, French, and Portuguese distributed via Qualtrics. Eligible respondents included professionals working in pediatric oncology across Africa. Descriptive and comparative analyses were conducted using SPSS, and free-text responses were analyzed thematically. RESULTS: There were 138 respondents from 33 African countries. Most were pediatric oncologists (55.1%) and completed the survey in English (71.7%). Only 5.8% reported hospital use of AI tools, mainly for imaging and risk stratification. In the previous year, 51.4% used AI for educational content and 44.2% for research, while 68.8% had used an AI platform in the preceding week. More than half (54.3%) reported no impact of AI on their clinical practice. Familiarity with AI tools was low, with machine learning the most recognized (8.5%). Barriers included insufficient training (93.2%), budget constraints (87.3%), and lack of infrastructure (80.4%). Higher World Bank Income Group and greater clinical experience correlated with higher AI familiarity (p < 0.01). Interest in learning AI tools did not differ meaningfully between low-, lower middle-, and upper middle-income settings. Respondents identified key opportunities for AI adoption in diagnostics (32.6%) and clinical management (39.1%), particularly to automate functions in settings with limited human resources. CONCLUSION: AI adoption in African pediatric oncology remains limited, but there is strong interest and readiness to learn despite major barriers related to infrastructure, training, and resources. The most immediate opportunities for AI lie in automating clinical and administrative functions where human capacity is constrained.

Ähnliche Arbeiten