Death registration in West Africa could capture only a fraction of suicides

An analysis of surveillance capacity across West Africa has quantified, for the first time, how little of the region’s suicide mortality its official statistical systems could actually detect. Combining the World Health Organization’s country-level suicide estimates for 2021 with measured death registration completeness, the study found that civil registration systems across the 16 countries of West Africa could capture no more than 1,446 of the 23,637 suicide deaths WHO estimated for the region that year, a ceiling of 6.1 percent. That figure assumes the best possible case: that every death that is registered also receives an accurate cause assignment. The analysis covered all 16 West African countries, spanning the 15 member states of the Economic Community of West African States plus Mauritania.

The study’s primary outcome was a registrable ceiling for each country, calculated as the WHO-estimated number of suicide deaths multiplied by that country’s all-cause death registration completeness, drawn from the International Completeness of Death Registration dataset. Among the 13 countries where completeness could be measured, the median was 14.1 percent, but the range was enormous: from 0.5 percent in Nigeria to 89.0 percent in Cabo Verde. For three countries, Ghana, Senegal, and Mali, registered death counts could not be obtained at all. Nigeria, which accounts for the largest share of the region’s estimated suicide burden at 10,912 deaths, or 46 percent of the 23,637 total, had a registrable ceiling of just 57 deaths. Because the ceiling assumes perfect cause certification for every registered death, the true number of suicides identifiable through registration in these countries is, by construction, no higher and likely lower.

Beneath these numbers lies a more fundamental problem: the quality of the data on which the region’s published suicide statistics rest. WHO classifies the empirical basis of its suicide estimates as very low quality for 15 of the 16 West African countries and low quality for Cabo Verde; no country in the region reaches a higher classification. In practical terms, this means the suicide rates that appear in international tables for most of West Africa are model-based estimates rather than direct measurements derived from registration data. The analysis also examined whether reported suicide rates track registration capacity, using ecological correlations between completeness and age-standardised suicide rates. The relationship proved unstable: the unweighted Pearson correlation was 0.51, but it fell to 0.14 when Cabo Verde was excluded and rose to 0.61 when countries were weighted by population. A statistical association that swings this widely depending on how countries are weighted or whether a single outlier is removed offers little basis for treating published rates as comparable measures of true suicide mortality across the region.

The authors conclude that death registration, as currently constituted, cannot serve as the foundation for suicide surveillance in most of West Africa. They argue that meaningful monitoring requires triangulation across multiple sources: health facility data, verbal autopsy methods, and surveys, rather than reliance on registration systems alone. They further identify priorities for strengthening civil registration and vital statistics infrastructure in the region, and note that other WHO regions have yet to conduct comparable analyses quantifying how much suicide mortality their own registration systems could capture. The cross-sectional design of the study means it captures registration capacity at a single point in time, and the ceiling calculations rest on WHO model estimates whose own empirical basis the study shows to be weak across the region.

The findings carry implications well beyond West Africa’s borders. Suicide mortality statistics are used to set prevention priorities, allocate resources, and evaluate interventions, yet this analysis demonstrates that in a region home to tens of thousands of estimated suicide deaths annually, the systems nominally responsible for counting those deaths would miss roughly 94 percent of them even under ideal assumptions. The gap between estimated burden and detectable mortality, the authors argue, is not a marginal data-quality issue but a structural limitation of surveillance itself, one that will persist until registration coverage expands substantially or complementary methods for ascertaining suicides are integrated into national health information systems.

Subject of Research: Suicide mortality surveillance capacity and death registration completeness in West Africa

Article Title: Counting the uncounted: suicide mortality and the limits of death registration in West Africa, a cross-sectional surveillance capacity analysis

Article References: Okereke, P. U., Alaaya, B. K., Abdulwasiu, K., Umeh, C. V., Okereke, O. W., Ikegwuonu, O., Okeke, A. C., Popoola, Y., Emovon, O. A., Osakue, S. E., & Akinwale, O. D. (2026). Counting the uncounted: suicide mortality and the limits of death registration in West Africa, a cross-sectional surveillance capacity analysis. BMC Public Health. https://doi.org/10.1186/s12889-026-29989-0

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29989-0

Keywords: suicide mortality, death registration, civil registration and vital statistics, West Africa, surveillance capacity, WHO suicide estimates, data quality, verbal autopsy, Nigeria, cross-sectional analysis, Counting, uncounted