Abstract
Introduction: Drowning is one of the most significant premature deaths in the world, where 372,000 people die due to drowning yearly. In Thailand, around 30.2% of all drownings were accounted for children under 15. In 2006, the Ministry of Public Health implemented a child drowning prevention policy that included survival swimming, education, and fencing around high-risk areas. Due to the implementations, child drowning significantly decreased from 11.5 to 6.8 deaths per 100,000 children between 2005 and 2014. Although the drowning was significantly improved, multiple drowning prevention interventions recommended by the Division of Injury Prevention, Department of Disease Control, were not widely implemented nationwide, and the guidelines were not well established. Thus, in 2015, the Division of Injury Prevention launched a comprehensive child drowning prevention program called "Merit Maker." Healthcare providers who want to establish a drowning prevention program can voluntarily join. This program consisted of 10 activities and was divided into three levels (Copper, Silver, and Gold) based on the coverage and activities the implementers could do. After initiating the Merit Maker program, the child drowning rate decreased from 6.8 to 6.1 per 100,000 children between 2014 and 2017. Furthermore, every province in Thailand has implemented the program. However, due to the nature of the program, where healthcare providers can voluntarily join, the program may not penetrate every area. This study aimed to systematically evaluate the program's efficacy, where two primary outcomes were a decrease in drowning and death due to drowning in children. Methods: This study used three secondary data sources included 1. the Merit Maker program records between 2015 and 2019 from the Division of Injury Prevention, MOPH, 2. Drowning records between 2009 and 2019 from Vital Statistic, MOPH, and 3. Drowning-related health utilization records between 2016 and 2019 from National Health Security Office. The data was cleaned and transformed into a yearly time series record where the district is the unit of analysis. There are two analysis types based on data timeframe. Multilevel mixed-effect negative binomial regression was used to compare the program's efficacy between districts that implemented and did not implement the program using data between 2015 and 2019. The second analysis used the Difference-in-differences method to evaluate the program's efficacy on drowning related death using data between 2009 and 2019. The Merit Maker program could be classified into three independent variables such as 1. The presence of the program (Yes/No) 2. The highest level of the program in the district (None, Copper, Silver, and Gold), and 3. The number of programs in the district (None, 1,2-3, and 4 and more programs). Furthermore, the analysis included subgroup analysis based on child gender and age group (aged less than 2, age 2-5, age 6-14). In addition, sensitivity analysis based on 1. Program's duration, 2. Water transport accidents, and 3. The district's drowning risk level was included. Results: The efficacy of the program analyses by Multilevel mixed-effect negative binomial regression using data between 2015 and 2019 was found as follows. 1. The districts that implemented the program had the same chance of child drowning (IRR: 1.01, p=0.73) and drowning-related death (IRR: 1.07, p=0.11) compared to districts that did not implement the program. 2. The analysis using other programs' configurations, the highest level of the program, and the number of programs in the district, also resulted similarly. 3. Subgroup analyses on drowning prevention's efficacy were similar to the primary analysis. Boys (IRR: 0.99, p=0.76), girls (IRR: 1.10, p=0.06), children aged less than 2 (IRR: 0.97, p=0.71), 2-5 (IRR: 1.07, p=0.20) and 6-14 (IRR: 1.02, p=0.68) living in the districts that implemented the program had the same chance of drowning as their peers in districts without the program. 4. Subgroup analyses on drowning-related death prevention's efficacy were similar to the primary analysis. Boys (IRR: 1.04, p=0.42), children aged less than 2 (IRR: 0.96, p=0.71), 2-5 (IRR: 1.12 , p=0.10 ), and 6-14 (IRR: 1.09 , p=0.16 ) living in the districts that implemented the program had the same chance of drowning-related death compared to their peers who lived in the districts without the program. However, girls who lived in the districts with the program had higher chances of drowning-related death (IRR: 1.19, p=0.03) than their peers who lived without it. 5. Sensitivity analyses on the three domains mentioned above were similar to the primary analysis. The efficacy of the program in drowning-related death prevention using the Difference-indifferences method had the same direction compared to Multilevel mixed-effect negative binomial regression, where the number of children drowning-related death between the district with and without the program were not statistically different (difference = -0.09, p=0.25). The analysis found that if the MOPH implements the program nationwide, child drowning-related death will be reduced by 84 (-55 to 213) per year. Additionally, subgroup and sensitivity analyses also yield similar results. Conclusion: There was insufficient data to support the effectiveness of a child drowning prevention program in preventing drowning and drowning-related fatalities among children under the age of 15.
