Underestimation of snakebite mortality by hospital statistics in the Monaragala District

Abstract

INTRODUCTION: Estimates of snakebite mortality rely upon deaths recorded in hospital. This study compared the number of recorded snakebite deaths in hospitals with registered deaths from the same district. METHODS: Snakebite mortality data for the period 1999-2003 were obtained for all hospitals in Monaragala District from the Medical Statistics Unit, Colombo. Data on snakebite as a certified cause of death for the district were obtained from the Registrar General's Department. Hospital mortality data were cross-checked at 7 of the 18 hospitals in the district [Base Hospital Monaragala, District Hospitals Bibile, Kataragama, Siyambalanduwa, Medagama and Wellawaya, and Rural Hospital Ethimale], accounting for 66% of recorded deaths. Registrar General's data were cross-checked at 19 of the 33 Divisional Secretariats in the District. RESULTS: We found that the data recorded centrally corresponded to what was documented in hospitals and Divisional Secretariats that we visited. For the 5-year period, there were 27 snakebite deaths recorded in hospitals in the Monaragala District. However, death registrations for the same district identified 72 deaths due to snakebite. The true number of snakebite deaths in Monaragala District was, therefore, 2.7 (95% confidence interval 2.0 - 3.7) times higher than the number of deaths recorded in hospitals. CONCLUSION: Our study convincingly demonstrates that hospital data underestimate the true burden of snakebite mortality; many victims probably die before reaching hospital. Although our study was limited to the Monaragala district, the findings probably reflect the situation in many areas of Sri Lanka where snakebite is a problem and public amenities are limited. We recommend large scale community surveys to more accurately determine the magnitude of snakebite mortality in this country.