Paraquat Peril: A Retrospective Study from the Frontlines of a Quaternary Care Hospital in South India.

Background and aim: Paraquat (PQ) dichloride is a commonly used herbicide across agricultural sectors worldwide. Despite being banned in over 70 countries, PQ is still frequently misused in India for suicidal and homicidal poisoning. The study aims to evaluate the clinical and laboratory predictors of survival and to estimate time-to-event outcomes of patients admitted to a quaternary care center in South India with a history of PQ poisoning. Patients and methods: This retrospective study included patients aged 16 and above who presented with PQ poisoning over 10 years (January 2015 to October 2024). Data were collected from the emergency department (ED) triage registry and hospital electronic database. Statistical analyses were conducted to identify factors predicting survival. Results: During the study period, 166 patients with PQ poisoning were included. The majority were young males, with 79.5% experiencing poor outcomes. Acute kidney injury (AKI) (92.2%) and lung injury (57.6%) were common complications. The overall survival rate was 20.5%, with a median survival time of 5 days. Multivariate Cox regression analysis identified N-acetylcysteine, ascorbic acid, cyclophosphamide, and steroids (NACS) combination therapy, a low acute physiology and chronic health evaluation (APACHE) II score, and activated charcoal as independent predictors of survival. A meta-analysis of 15 Indian studies revealed a pooled bad outcome rate of 79%. Conclusion: High morbidity and mortality associated with PQ poisoning in India have been highlighted in our study. The authors suggest that a combination therapy using NACS may improve survival; however, this needs to be explored further. Despite these interventions, delayed presentation with multiorgan system involvement points toward a poor prognosis. Further research is needed to standardize treatment protocols and guide policy interventions to limit PQ use and enhance outcomes. How to cite this article: Neeraj GPP, Philip AJ, Prakash G, Praneeth CS, Chandiraseharan VK, Zachariah A,