Hyperbaric Oxygen for Carbon Monoxide Poisoning
Search Strategy: You conduct a PUBMED search using the term for “hyperbaric oxygen AND carbon monoxide” applying the limits humans, clinical trials, meta-analysis, practice guideline, and randomized control trial yielding 26 citations (see http://tinyurl.com/84pb5de). You select the four below for closer review.
While reviewing the weekly emergency medicine residency social calendar you notice there is a “happy hour” at a hookah (see figure where narghile = hookah) bar on Delmar Ave. As a toxicologist your interest is piqued, until you realize that you are scheduled for a shift that night. A few hours into your overnight shift your pager notifies you that a 30 year-old male is arriving by EMS after a syncopal episode. His blood pressure is 131/78, heart rate 92, respiratory rate18, and he is oxygenating 97% on room air. Since the night is relatively slow you are waiting in the room for the patient and upon arrival are surprised to see that he is accompanied by two of the 3rd year EM residents.
Initially quite concerned you obtain a further history from the patient and the EM residents that were witness to the situation. The patient had no complaints prior to the “syncopal episode”, but currently feels slightly dizzy and has a diffuse headache. He admits to drinking “2 beers” while smoking a hookah pipe. After approximately 2½ hours at the bar, witnesses noted the patient’s head fall to the table with the hookah hose still in his mouth. After about one-minute his friends realized that something might be awry and tried to arouse him. After taking the hose out of his mouth, pouring some beer on his face, and sticking a cocktail straw in his nose he responded. After yelling at him he stood up and then fell to the floor. At this point, the residents realized that there was a problem and called 911before rushing to assist him.
After getting a handle on the situation, the off-service intern is quickly on the case. The vital signs remain normal, his pulse ox is 98% on room air, and EKG is sinus rhythm with normal intervals. The physical exam and an extensive neurological exam are both normal. The astute intern, having grown up in another part of the world where hookah pipe smoking is common, immediately puts the patient on 100% oxygen and orders a venous carboxyhemoglobin (COHb) level which returns at 31%.
Population: ED patients with acute carbon monoxide poisoning
Intervention: Hyperbaric oxgen
Comparison: Normobaric oxygen
Outcome: Neurological sequelae, adverse treatment events (pneumothorax, hyperoxia-related seizures)
Article 1: Delayed neuropsychologic sequelae after carbon monoxide poisoning: Prevention by treatment with hyperbaric oxygen, Ann Emerg Med 1995; 25: 474-480.
The incidence of carbon monoxide poisoning (accidental or intentional suicide attempt) have been on the decline for 40-years due to increasingly stringent emissions laws (Mott 2002, Studdert 2010). Nonetheless, carbon monoxide may be the cause of over half the fatal poisonings reported worldwide (Raub 2000). Carbon monoxide poisoning is the third leading cause of unintentional poisoning death in the United States with estimates ranging from 15,000 to 40,000 events (Hampson 1998, Hampson 2005, MMWR 2005), but these figures likely underestimate the incidence since CO toxicity is often misdiagnosed (Barret 1985, Dolan 1987). Carbon monoxide poisoning can be unintentional so patients headache, nausea, dizziness, or syncope and potential CO-exposures (space heaters in winter months, closed spaces heated with motor vehicle, work-related combustion engine exposures) should at least consider this toxin (Keles 2008).
Carbon monoxide binds to hemoglobin with an affinity 200-fold greater than oxygen causing a leftward shift in the oxygen-hemoglobin dissociation curve and decreasing oxygen delivery to the tissues. However, the pathophysiology of carbon monoxide toxicity is more complicated than simply producing a relative anemia. Goldblum demonstrated that dogs breathing 13% CO died within one-hour after reaching a carboxyhemoglobin level of 54%-90%, but that otherwise healthy dogs transfused with 80% carboxyhemoglobin blood resulted in no toxic effect (Aviation Space Environmental Medicine 1975; 46: 1289-1291). The current understanding of CO toxicity is that clinical symptoms are the combined effects of cellular hypoxia and oxidative/nitrosative stress so tissue damage does not correlate with carboxyhemoglobin levels (Piantadosi 2004).
The elimination half-life of carbon monoxide is 4- to 5-hours. Administration of 100% oxygen via tight-fitting facemask speeds the elimination of carbon monoxide from the body to approximately 1-hour (Peterson 1970). Animal studies suggest that hyperbaric oxygen (HBO) improves neurological outcomes. Animal studies also suggest that HBO may be harmful. Specifically, HBO adversely increases PMN adherence (Thom 1993)
ACEP Guidelines for acute carbon monoxide poisoning offer a lukewarm Level C recommendation stating that “HBO is a therapeutic option for CO-poisoned patients; however, its use cannot be mandated.” Our review of the highest quality literature concurred with the ACEP Guideline with approximately one-third of physicians committed to using HBO in the case above, although very few were convinced that HBO would significantly improve the patient’s outcome based upon the available data. Evidence synopsis: Treatment of non-pregnant adult patients suffering from acute (less than 24-hours since removal from the source) symptomatic CO toxicity with one, two, or three HBO sessions reduces the negative cognitive sequelae at 6-weeks and 12-months with NNT of 5 and 7, respectively. The most sensitive neuropsych subtests to detect neurological sequelae are the Trails A, Trails B, and digit span tests. In comatose carbon monoxide poisoning patients, two sessions of HBO may provide worse outcomes than one session. Compared with NBO, HBO is more expensive, not readily available, and associated with adverse reactions including barotrauma, seizures, pulmonary edema, and claustrophobia.
Sufficient clinical equipoise exists to justify future RCT’s that are triple-blinded (investigator, patient, outcome assessor) using sham dives, using explicit and reproducible primary and secondary outcome measures defined before recruitment ensues. Specifically, future researchers should use a similar definition of delayed neurological sequelae and assess trial power a priori while incorporating costs and transport times (to HBO chamber) into a triple-blinded (patient, clinician, outcome assessor) randomized controlled trial. Unfortunately, a review of clinicaltrials.gov reveals no such registered trials currently underway.