Outcome assessment of acute methanol poisoning: A risk-prediction nomogram approach for in-hospital mortality

Abstract

Acute methanol poisoning could be associated with high morbidities and fatalities. Stratifying high-risk patients is crucial in improving their prognosis. Hence, this study aimed to identify patients with methanol poisoning at high risk of in-hospital mortality. Also, the risk factors for blindness were assessed. The study included 180 acutely methanol-poisoned patients who received standard medical care. Out of 180 patients, 52 (28.9 %) patients presented with blindness, and 43 (23.9 %) patients died. The predictive model was based on four significant variables, including blindness, mean arterial pressure, serum bicarbonate, and serum creatinine. The presence of blindness and elevated serum creatinine significantly increased the likelihood of mortality by 14.274 and 5.670 times, respectively. Likewise, decreases in mean arterial pressure and serum bicarbonate significantly increased mortality risk by 0.908 and 0.407 times, respectively. The proposed nomogram exhibited excellent discriminatory power (area under the curve (AUC)=0.978, accuracy=93.3 %), which outperforms the AUCs of individual predictors. The provided nomogram is easily applicable with outstanding discrimination, making it clinically helpful in predicting in-hospital mortality in acutely methanol-poisoned patients. Regarding the risk factors for blindness, multivariable regression analysis revealed that delayed time for admission (OR=1.039; 95 % CI=1.010-1.069; <i>p</i>= 0.009) and elevated anion gap (OR=1.053; 95 % CI=1.007-1.101; <i>p</i>= 0.023) were significant risk factors. The current study assists physicians in identifying methanol-poisoned patients with a high probability of mortality or blindness on admission. Future studies are recommended for external validation of the created nomogram, in addition to follow-up for patients with visual impairment.