Assessing variations in estimates of drowning mortality in Turkey from 2013 to 2019

Abstract

Abstract Introduction Drowning is an under-recognised public health threat and a leading cause of injury-related mortality and morbidity. However, in many countries, including Turkey, limited data impair understanding of drowning burden and Global Burden of Disease (GBD) Study drowning estimates (defined using International Classification of Diseases [ICD] codes W65–74) do not include flood-related deaths (X38) and water transportation related drownings (V90, V92). A lack of accessible and reliable country-level data impacts a country’s ability to develop appropriate drowning prevention interventions and measure efficacy. This retrospective population-based study aimed to explore differences between two datasets in fatal drowning in Turkey between 2013 and 2019. Methods National, all-age data on fatal drownings (restrictive definition: ICD-10 codes W65–74) were sourced from the Turkish Statistical Institute (TurkStat) and the Global Burden of Disease (GBD) study. In addition, a broader definition of drowning including water transport, flood-related deaths and drowning due to undetermined intent (ICD-10 codes W65–74, V90, V92, X38, Y21, T751) were sourced from TurkStat. Numeric and percentage differences in number of drowning deaths were calculated overall and by sex, age group and death year. Chi square (p < 0.05) and relative risk (95% confidence intervals) using crude drowning rates per 100,000 population were also calculated for TurkStat data. Results From 2013 to 2019, TurkStat reported a total of 5004 drowning deaths (coded W65–74) were reported, compared to 5252 (5% difference; n = 248) using the broader definition. A restrictive definition underreported drowning most significantly in females (9.5%; n = 97), 5–9 year-olds (8.9%; n = 31) and in the 2015 calendar year (30.2%; n = 226). Males accounted for 78.8% of drowning in Turkey, with females significantly (p < 0.001) more at risk under 10 years of age (0–4 years X2 = 67.9; 5–9 years X2 = 23.9) and aged 65+ years (X2 = 29.7). GBD data overestimated a restrictive definition of drowning by 3.2% overall (7.6% for females, 52.5% for 0–4 year-olds) and underreported drowning for 65+ year-olds by 17% when compared to TurkStat restrictive definition of drowning. Conclusions Although a restrictive definition of drowning doesn’t greatly impact estimates at a population level in Turkey, there are variations. This highlights the importance of accurate country-level drowning data to guide decision making for prevention.