Abstract Background Accurate data on the causes of death (CoDs) is the foundation for analyzing the health situation and developing health policy in various countries. The percentage of garbage codes (GCs) indicates the quality of the CoDs data. Within the Western Balkan Strategic Partnership Project (Work Package 4: Use of Population Health Metrics for Improved Surveillance), we evaluated the impact of the four levels of GCs on the quality of mortality data of Belgrade. Methods In a retrospective study, we examined the underlying CoDs of Belgrade residents by age and sex and the change in GCs’ share in the CoDs in 2010, 2015, and 2020. We used anonymous individual mortality data and population age and sex annual data from the electronic databases of the City Institute of Public Health of Belgrade, as well as the list of GS levels 1-4 from the 2019 Global Burden of Disease Study. Results Over the observed years, the percentage of all GCs in underlying CoDs was approximately 20%. The GCs with profound policy implications level 1 accounted for 52% of all GCs in 2020, higher than those recorded in previous years - 46.3% in 2010 and 45.1% in 2015. The total number of GCs per 100,000 inhabitants in Belgrade was the highest in 2020, with a rate of 325.2/100,000 (males: 374.4, females: 301.3). This rate has increased by about one-third since 2010 (both sexes: 245.4, males: 269.2, females: 221.1) and by about two-fifths since 2015 (both sexes: 235.8, males: 262.7, females: 211.8). Conclusions This research has revealed a significant decline in the quality of data related to the causes of death in Belgrade. One-fifth of the recorded causes of death are inaccurate or unreliable. The study highlights the importance of maintaining strict protocols for data collection, analysis, and management to ensure the accuracy and reliability of CoDs data in developing effective public health policies. Key messages • The rate of total garbage codes per 100,000 inhabitants of Belgrade has increased by about two-fifths since 2015. • Failure in quality, accuracy, and reliability of causes of death can lead to suboptimal health policymaking.
Read full abstract