Evaluation of data quality at the Gambia national cancer registry
Résumé
The Gambia National Cancer Registry (GNCR) is one of the few nationwide population-based cancer registries in sub-Saharan Africa. Most registries in sub-Saharan Africa are limited to cities; therefore, the GNCR is important in providing estimates of cancer incidence in rural Africa. Our study assesses the quality of its data. The methods proposed by Bray and Parkin, and Parkin and Bray (Eur J Cancer 2009;45:747-64) were applied to the registry data from 1990 to 2009 to assess comparability, validity and completeness. The system used for classification and coding of neoplasms followed international standards. The percentage of cases morphologically verified was 18.1% for men and 33.1% for women, and that of death certificate only cases was 6.6 and 3.6%, respectively. Incidence rates in rural regions were lower than in the urban part of the country, except amongst young male adults. Comparison with other West African registries showed that the incidences of liver and uterine cervical cancer were comparable, but those of prostate and breast in The Gambia were relatively low. The overall completeness was estimated at 50.3% using the capture-recapture method. The GNCR applies international standard practices to data collection and handling, providing valuable data on cancer incidence in sub-Saharan Africa. However, the data are incomplete in the rural and elderly populations probably because of health care access and use.