Background
To understand cancer impact and healthcare system response in countries, information on stage at diagnosis and patient survival at a population level are key. Population-based cancer survival in Africa relies largely on active follow-up, which is labour intensive and requires careful monitoring and quality control. Furthermore, information on stage at diagnosis has not been available until recent years in population-based registries of Africa.
Objectives
To support registries in sub-Saharan Africa to capture complete and high-quality stage and survival data, African Cancer Registry Network (AFCRN) utilised structured training courses led by a network of cancer registry experts and deployed mentorship on site for research activities.
Methods
Since 2020, AFCRN organised two series of online training courses, supplemented by many onsite mentorships, on abstracting information on stage at diagnosis, and carried out follow-up of registered cancer cases, in both English and French. This was preformed to estimate stage at diagnosis, survival and care pathway in children and adults diagnosed with cancer in 20 registries in 14 countries.
Results
Data from earlier years have shown a large percentage of cases with missing information on stage, as high as 100% during early registration period; and a high proportion of cases lost to follow-up when active follow-up was performed (20-50%, cancer type depending). Overall survival was very low among patients diagnosed with cancer in Africa (for example 5-year survival from 11% for oesophagus to 49% for breast cancers diagnosed between 1993-97 in Harare Zimbabwe). In 2021, AFCRN conducted training courses on staging of major cancers diagnosed in adults and children, and, since 2022, has added stage to its data requirement as a compulsory reportable variable. After these changes, there has been a decrease in the percentage of unknown stage from more than 80% to around 50-60%. A validation study showed that cancer registrars from the region can capture information on staging from clinical files to an approximate 80% accuracy rate. With increasing use and coverage of mobile phones, active follow-up has more success than in previous studies. As a result of these initiatives, data on stage-specific survival has become increasingly available. They suggest that the prognosis of cancer patients diagnosed with early stage is lower than in high-income countries (colorectal 5-year survival at 50% compared with 90% in ICBP study), but late-stage cases had much worse outcome in African settings (around 25% only). Studies investigating treatment pathways suggested that multiple factors affected patients’ access to treatment, such as treatment availability and financial difficulties.

Conclusions
With dedicated supervision and training, despite persistent challenges, survival studies with high quality of data in African settings are feasible. The findings are crucial for any rational cancer control planning for the targeted population.