Hart's survival curve are Very Useful for Predicting Prognosis in Paraquat Poisoning.
- Author:
Jung Hwa KIM
1
;
Young Hee LEE
;
Sung Hye SHIN
;
Paul KIM
;
Jai Gyu LEE
;
Ji Eun LEE
;
Jin Gyu LEE
;
Kwang Young LEE
Author Information
1. Department of Internal Medicine, Presbyterian Medical Center, Jeonju, Korea. Kwangyou@unitel.co.kr
- Publication Type:Original Article
- Keywords:
Paraguat intoxication;
Hart's survival curve;
Plasma paraguat level;
Logistic regression
- MeSH:
Agriculture;
Dithionite;
Eating;
Humans;
Logistic Models;
Mortality;
Odds Ratio;
Paraquat*;
Plasma;
Poisoning*;
Prognosis*;
Risk Factors
- From:Korean Journal of Nephrology
2004;23(2):270-277
- CountryRepublic of Korea
- Language:Korean
-
Abstract:
BACKGROUND: Paraquat is highly toxic herbicide used in agriculture, and mortality of paraquat poisoning is very high. The predictions of severity of poisoning assessed with the amount of ingestion, results of urine dithionite test and laboratory findings are not accurate. The aims of this study are to investigate whether Hart's probability of survival curves are useful for predicting outcomes of paraquat-poisoned patients and which factors influence the outcomes of the patients. METHODS: We grouped 175 patients with paraquat poisoning into seven groups using curves of probability of survival suggested by Hart, et al. Group A was patients with plasma paraquat concentration below 90% probability of survival curve, group B patients with paraquat levels between 90% and 70% probability of survival curves, group C patients with paraquat levels between 70 and 50% probability of survival curves, group D the ones with paraquat levels between 50% and 30% probability of survival curves, group E the ones with paraquat levels between 30 and 20% probability of survival curves, group F paraquat levels between 20 and 10% probability of survival curves, and group G patients with paraquat levels above 10% probability of survival curve. And we analyzed the survival and mortality rates of each groups. RESULTS: The mortality rates of groups A, B, C, D, E, F and G were 7.4%, 26.7%, 37.5%, 55.6%, 63.6 %, 83.3% and 96.8% respectively (p<0.0005). The most important risk factor for death of patients was the paraquat concentration grouping with odds ratios (95% CI) of 4.4 (0.7-27.5, group B), 7.2 (0.9-54.9, group C), 15.0 (52.1-105.6, group D), 21.0 (3.2-139.7, group E), 60.0 (7.4-487.1, group F), and 359.9 (56.9- 2277.9, group G). CONCIUSION: In assessing prognosis of patients with paraquat poisoning, grouping of the patients using initial plasma paraquat concentrations is the most important, and the policy of treatment can be decided according the results.
