Baseline and Time-Dependent Laboratory Prognostic Factors for In-Hospital Mortality After Acute Diquat Poisoning: A Retrospective Cohort Study

Abstract Background Acute diquat poisoning can progress rapidly from gastrointestinal injury to renal, neurologic, circulatory, and multiorgan failure. Most prognostic studies rely on measurements obtained at presentation and therefore do not capture how laboratory abnormalities evolve during hospitalization. We evaluated baseline prognostic factors and explored associations between serial laboratory measurements and in-hospital mortality. Methods This single-center retrospective cohort included patients with analytically confirmed acute diquat poisoning treated from January 2019 through June 2024. Admission was defined as time zero; in-hospital death was the event, and survivors were censored at discharge. The baseline Cox model included log10-transformed admission plasma diquat concentration, anuria, central nervous system injury, and Acute Physiology and Chronic Health Evaluation II score. Serial laboratory results were structured in a start-stop data set and assessed individually as internal time-dependent covariates in adjusted Cox models with patient-clustered robust standard errors. False-discovery-rate adjustment was applied across the laboratory analyses. Results Among 201 patients, 60 (29.9%) died in hospital. Admission plasma diquat concentration (hazard ratio [HR] per 10-fold increase, 2.72; 95% CI, 1.63–4.53), anuria (HR, 7.43; 95% CI, 2.52–21.91), central nervous system injury (HR, 4.78; 95% CI, 1.55–14.74), and APACHE II score (HR per 5-point increase, 1.23; 95% CI, 1.07–1.41) were independently associated with mortality. The model C-index was 0.937, and the global proportional-hazards test was not significant (P = 0.067). Ten of 24 time-dependent laboratory variables remained associated with mortality after false-discovery-rate correction. The largest standardized associations were observed for blood urea nitrogen (HR, 2.18; 95% CI, 1.54–3.10), hemoglobin (HR, 1.77; 95% CI, 1.15–2.74), creatinine (HR, 1.76; 95% CI, 1.27–2.43), and creatine kinase (HR, 1.60; 95% CI, 1.29–1.99). Conclusions Admission toxin burden, anuria, neurologic injury, and physiologic severity identified patients at high early risk. Serial abnormalities in coagulation, renal function, tissue injury, inflammation, and hepatic function tracked the evolving mortality hazard and may support repeated bedside reassessment. These exploratory dynamic associations require external validation and should not be interpreted as causal treatment targets.