Potential Risk Estimation Drowning Index for Children (PREDIC): A pilot study from Matlab, Bangladesh

Abstract

Childhood drowning is a major public health problem that has been neglected in many low- and middle-income countries. In Matlab, rural Bangladesh, more than 40% of child deaths aged 1-4 years are due to drowning. Aim: The main objective of this paper was to develop and evaluate a childhood drowning risk prediction index. Methodology: A literature review was carried out to document risk factors identified for childhood drowning in Bangladesh. The Newacheck model for special health care needs for children was adapted and applied to construct a childhood drowning risk index called "Potential Risk Estimation Drowning Index for Children" (PREDIC). Finally, the proposed PREDIC Index was applied to childhood drowning deaths and compared with the comparison group from children living in Matlab, Bangladesh. This pilot study used t-tests and Receiver Operating Characteristic (ROC) curve to analyze the results. Results: The PREDIC index was applied to 302 drowning deaths and 624 children 0-4 years old living in Matlab. The results of t-test indicate that the drowned children had a statistically (t = -8.58, p = 0.0001) significant higher mean PREDIC score (6.01) than those in comparison group (5.26). Drowning cases had a PREDIC score of 6 or more for 68% of the children however, the comparison group had 43% of the children with score of 6 or more which was statistically significant (t = -7.36, p < 0.001). The area under the curve for the Receiver Operating Characteristic curve was 0.662. Conclusion: Index score construction was scientifically plausible; and the index is relatively complete, fairly accurate, and practical. The risk index can help identify and target high risk children with drowning prevention programs. PREDIC index needs to be further tested for its accuracy, feasibility and effectiveness in drowning risk reduction in Bangladesh and other countries. © 2011 Elsevier Ltd.