Background: Carbon monoxide poisoning (COP) accounts for a large number of emergency department visits worldwide and is fatal in many cases. In surviving patients, neurological sequelae (NS) attributable to cerebral hypoxia are the most devastating outcome, but reliable predictors are limited. Therefore, we conducted a study to identify predictors of NS in patients with COP and evaluate their effects.
Methods: In this retrospective case-control study, we identified patients with COP in a medical center in Southern Taiwan between January 2005 and December 2014. Cases were patients with NS, and controls were patients without NS. We obtained information on potential predictors of NS from medical records and evaluated their association with NS, including demographic characteristics, exposure source, suicide attempts, duration of exposure (by tertile), histories, symptoms, signs, laboratory data, treatment, and the length of hospital stay.
Results: We included 371 patients with COP. Of them, 93 developed NS, and their mean ages (41.4 ± 14.7 years vs. 39.7 ± 14.2 years) and proportions of males (59.1% vs. 58.6%) were similar to those in the 298 controls. Multivariate logistic regression showed that a history of hypertension (adjusted odds ratio = 2.1; 95% confidence interval = 1.0, 4.5) and a longer duration of carbon monoxide exposure (adjusted odds ratio = 1.7; 95% confidence interval = 1.1, 2.8; the longest tertile [>5 hours] vs. the other two tertiles [≤5 hours]) were independent predictors for NS, but not the level of carboxyhemoglobin.
Conclusions: This study identified two independent predictors for NS that may be useful for public healthcare workers and physicians in predicting outcomes and deciding on treatment strategies for COP patients.