Pattern and predictors of death from aluminum and zinc phosphide poisoning using multi-kernel optimized relevance vector machine.

ABSTRACT: The use of metal phosphides, particularly aluminum phosphide, poses a significant threat to human safety and results in high mortality rates. This study aimed to determine mortality patterns and predictive factors for acute zinc and aluminum phosphide poisoning cases that were admitted to Menoufia University Poison and Dependence Control Center from 2017 to 2021. Statistical analysis revealed that poisoning was more common among females (59.7%), aged between 10 and 20 years, and from rural regions. Most cases were students, and most poisonings were the result of suicidal intentions (78.6%). A new hybrid model named Bayesian Optimization-Relevance Vector Machine (BO-RVM) was proposed to forecast fatal poisoning. The model achieved an overall accuracy of 97%, with high positive predictive va

SUBMITTER: Abdelghafar S 

PROVIDER: S-EPMC10201496 | biostudies-literature | 2023 May

REPOSITORIES: biostudies-literature