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

Suicide prevention policy in West Africa is based on mortality statistics whose empirical basis has not been quantified. Civil registration and vital statistics (CRVS) systems are the standard source of suicide mortality data worldwide, but their capacity in this region is unknown. We quantified the maximum contribution that death registration could make to suicide surveillance across the 16 countries of West Africa. In this cross-sectional surveillance capacity analysis, we combined country-level estimated suicide deaths, rates, and data-quality classifications for 2021 from the WHO report Suicide worldwide in 2021 with all-cause death registration completeness from the International Completeness of Death Registration (ICDR) dataset for the 15 ECOWAS member states and Mauritania. The primary outcome was the completeness-limited registrable ceiling, defined for each country as WHO-estimated suicide deaths multiplied by all-cause registration completeness, assuming that every registered death receives an accurate cause assignment. Five sensitivity analyses tested this ceiling, and ecological correlations were examined with and without population weighting. WHO classifies the data quality underlying suicide estimates as very low for 15 of 16 countries and low for Cabo Verde; no country attains medium or high quality. Registered death counts were unobtainable for Ghana, Senegal, and Mali. Median completeness among the 13 measurable countries was 14.1% (range 0.5 in Nigeria to 89.0 in Cabo Verde). Of 23 637 WHO-estimated suicide deaths in 2021 (Nigeria 10 912, 46%), registration systems could capture no more than 1446 (6.1%) even with perfect cause certification (8.4% if the three countries without data were assigned the median completeness). Nigeria’s ceiling was 57 deaths, about one registrable suicide per 190 estimated. The ecological correlation between completeness and age-standardised rates was unstable (Pearson r = 0.51; 0.14 excluding Cabo Verde; 0.61 weighted by population). Published suicide rates for most West African countries are model-based estimates rather than direct measurements, and death registration cannot currently support suicide surveillance in most of the region. Surveillance requires triangulation of facility data, verbal autopsy, coroner records, and surveys. Compilation reforms, CRVS strengthening, and removal of legal barriers to accurate certification are also needed. Other WHO regions should quantify this gap.