Abstract
We developed a simple severity index for patients with paraquat poisoning to determine in the initial stages of admission how likely a favorable outcome is. Upper gastrointestinal endoscopic examination was performed in 23 patients with paraquat poisoning over a 9-year period, and the severity of the findings were rated. Using these seventies, we compared survivors and non-survivors of paraquat poisoning. There were no significant differences between the two groups in age, sex, initial management or estimated volume of ingested paraquat. Endoscopic findings were rated on a scale of 0 to 3 as follows: normal esophageal findings 0, esophagitis 1, esophageal ulcer 2; normal gastric and duodenal findings 0, redness, swelling and slight erosion 1, and erosion and ulceration 2. The estimated probability of death (Pr) calculated by logistic regression from these endoscopic scores was 1/{1+exp(17-4.14eso-8.85stom-4.14duod)}. The value of Pr in survivors and non-survivors wan 0.20±0.12 and 0.89±0.05, respectively (P<0.01). At the cutoff value of Pr (0.5), sensitivity and specificity were 100% and 60%, respectively. The 3 patients whose Pr was less than 0.5 survived, and the 14 cases whose Pr was more than 0.5 died. Of the 6 patients whose Pr was 0.5, 2 patients (33.3%) survived and 4 patients (66.7%) had poor outcomes. Those results suggest that the Pr is a reliable and valuable prognostic index in the initial stage of paraquat poisoning.
