Citation

Xu Y, Li W. Ann. Emerg. Med. 2020; 75(6): 781-782.

Copyright

(Copyright © 2020, American College of Emergency Physicians, Publisher Elsevier Publishing)

DOI

10.1016/j.annemergmed.2020.01.032

PMID

unavailable

Abstract

In a recent issue of Annals, Cho et al investigated the risk of deep venous thrombosis after carbon monoxide poisoning. They found that it was significantly increased by 90 days after carbon monoxide poisoning. The authors used a cohort crossover design in which each patient was used as his or her own control. This study design was able to abolish the confounders between cases and controls, in an assumption that all confounders were similar in a given patient except for the carbon monoxide poisoning. However, such a study design works under the assumption that the risk factors for deep venous thrombosis were constant except for the event of carbon monoxide poisoning for a given patient in the comparative periods, but we propose that this assumption cannot hold true because numerous confounders can coexist in carbon monoxide poisoning. For example, hospitalization and immobility can significantly increase the risk of deep venous thrombosis. Many patients with carbon monoxide poisoning would have been admitted to the ICU and they monoxide poisoning may not have a causal link with deep venous thrombosis. We propose that all patients with critical illness be monitored for deep venous thrombosis...

Language: en