Hyperbaric Medicine

Optimal sessions of hyperbaric oxygen therapy in patients with carbon monoxide poisoning: A prospective observational study

a b s t r a c t

Introduction: Hyperbaric oxygen therapy may be a useful treatment to prevent the development of de- layed neuropsychiatric sequelae (DNS) in patients with Acute carbon monoxide poisoning. However, there is no clear consensus regarding the optimal number of HBO therapy sessions in patients with CO poisoning. Here, we compared the development of DNS after 3 and > 3 sessions of HBO therapy in patients with Acute CO poisoning.

Methods: This prospective observational study recruited 299 patients with CO poisoning. Demographic and clin- ical information were obtained, including comorbidities, vital signs, and symptoms. Patients were divided into two groups according to whether they received 3 or > 3 sessions of HBO therapy (3 HBO vs. >3 HBO). A propen- sity score-matching process was used to balance potential prognostic factors in both groups.

Results: Of the 299 patients with acute CO poisoning enrolled in this study, 183 (59.0%) were included in the anal- ysis. Patients were excluded for the following reasons: age < 18 years, not underwent HBO therapy, discharged against medical advice, and loss to follow-up. The overall rate of DNS development was 17.5%. The >3 HBO group had a higher incidence of DNS development compared to the 3 HBO group (36.3% vs. 16.3%; p = 0.09). Propensity score-matching analysis revealed similar incidences of DNS (31.3% vs. 28.1%, respectively; p > 0.99).

Conclusions: There is a critical need to determine the optimal number of HBO therapy sessions for patients with acute CO poisoning. This study showed no difference in DNS development after 3 and > 3 sessions of HBO therapy.

(C) 2021

  1. Introduction

Carbon monoxide poisoning is among the most common poi- sonings worldwide. CO is a substance generated through incomplete combustion. When inhaled, CO causes tissue hypoxia because it exhibits greater affinity for hemoglobin compared with oxygen [1]. In addition, CO limits the body’s oxygen supply by binding to myoglobin and mito- chondrial cytochrome oxidase, while causing brain lipid peroxidation [2]. Therefore, CO poisoning causes greater damage to oxygen- dependent organs such as the brain, heart, and kidneys.

Brain damage caused by CO poisoning can take various forms, in- cluding delayed neuropsychiatric sequelae (DNS). DNS comprise neuro- logic symptoms such as headache, difficulty concentrating, lethargy, emotional lability, amnestic syndromes, dementia, psychosis, Parkin- sonism, chorea, apraxia, agnosia, peripheral neuropathy, and urinary in- continence; these symptoms appear after a lucid interval of 2-40 days

* Corresponding author at: Department of Emergency Medicine, Soonchunhyang University Bucheon Hospital, 170 Jomaru-ro, Bucheon 14584, Republic of Korea.

E-mail address: [email protected] (Y.H. Lee).

following CO poisoning [3]. DNS reportedly affects up to 40% of patients with CO poisoning, of which only half recover within 1 year [4,5]. There- fore, various treatment methods have been studied to prevent the onset of DNS.

Hyperbaric oxygen (HBO) therapy is a potential preventive treat- ment for DNS in patients with acute CO poisoning [3]. HBO therapy rap- idly removes CO from the blood and exerts beneficial effects on brain cells traumatized by CO, including reductions in lipid peroxidation and endothelial leukocyte migration, as well as other post-hypoxic events [6]. HBO therapy may be an important part of early treatment in pa- tients with CO poisoning and has been proposed to prevent DNS [7]. However, there is no clear consensus regarding the optimal number of HBO therapy sessions in patients with CO poisoning. A previous nation- wide cohort study showed that patients who received two or more ses- sions of HBO had a better clinical outcome than did those who received HBO therapy only once [8]. Thus, it might be useful to administer more than 2 sessions of HBO therapy in patients with severe CO poisoning. However, there has been no prospective study regarding the effects of more than 2 sessions of HBO therapy.

Here, we attempted to determine whether there was a difference in neurologic outcomes between patients who received 3 sessions of HBO

https://doi.org/10.1016/j.ajem.2020.10.062

0735-6757/(C) 2021

Treatment protocols“>propensity score matching“>therapy (3 HBO) and patients who received >3 sessions of HBO therapy (>3 HBO, up to 14 sessions) after CO poisoning.

  1. Materials and methods

This prospective observational study was performed at a single aca- demic tertiary care center with an annual emergency department cen- sus of 65,000 patients per year. The study was approved by the Institutional Review Board of the study institution. The study was con- ducted in accordance with the provisions of the Declaration of Helsinki.

    1. Study population

Patients who visited the emergency department due to CO poi- soning were recruited from August 2016 to October 2019. CO poison- ing was defined as Carboxyhemoglobin level > 5% (10% for smokers) among patients in our emergency department or primary hospital, or when the patient’s history clearly indicated CO poisoning (e.g., attempted suicide using coal briquettes). The exclusion criteria were: 1) age < 18 years, 2) lack of HBO therapy, 3) discharged against medical advice, and 4) loss to follow-up.

    1. Treatment protocols

Each patient received 100% oxygen by a facial mask or mechan- ical ventilator following endotracheal intubation. Patients received HBO therapy if they met >=1 of the following indications: 1) carboxyhemoglobin level > 25%; 2) presence of neurologic deficits such as mental changes, a loss of consciousness, or seizures; and 3) a suspicious cardiac injury (e.g., elevated cardiac enzymes or chest pain symptoms).

The protocol for HBO therapy at our hospital included 3 sessions within 24 h. Patients were treated with HBO therapy using a monoplace Hyperbaric chamber (IBEX Medical Systems, Seoul, Korea). The first ses- sion was performed within 1 h of arrival at the emergency department. The protocol is as follows: For the first session, titrate to 3 atm absolute (ATA) during the first 15 min then maintain 3 ATA for the next 60 min. Drop to 2ATA in the next 5 min and stay in 2 ATA for the rest 60 min and taper out during 10 min. For the second and third session, titrate to 2 ATA during 10 min then maintain the ATA for 100 min, and taper out during the last 10 min [6]. Normobaric oxygen therapy was maintained during preparation for HBO therapy, and brain magnetic resonance im- aging (MRI) was performed immediately after 3 sessions of HBO therapy.

In patients with acute abnormal lesions on brain MRI (as interpreted by a radiology specialist) or persistent complaints of neurologic deficits after 3 sessions of HBO therapy, up to 11 additional sessions of HBO therapy were recommended (maximum of 14 sessions overall). The protocol for additional HBO therapy was identical to that used during the second and third sessions. However, if a patient did not desire addi- tional treatment, only 3 sessions of HBO therapy were performed. If the brain MRI findings were normal, additional sessions were performed if a patient requested additional treatment.

    1. Data collection

Baseline characteristics were obtained, including age, sex, height, weight, vital signs, comorbidities, current smoking, initial Glasgow Coma Scale score, duration of CO exposure, cause of exposure (acciden- tal or intentional), symptoms, laboratory results, and brain MRI find- ings. The data were reviewed by two emergency physicians who were blinded to the purpose and hypothesis of this study.

DNS was considered in patients with mental deterioration, cognitive dysfunction, gait disturbance, mutism, urinary or fecal incontinence, psychosis, depression, or Parkinsonism. All patients were confirmed to have no Neurological deficits through neurologic consultation before

discharge. The patients were informed of the symptoms of DNS before discharge and were asked to return to the hospital whenever such symptoms appeared. In addition, phone calls were made to check whether the symptoms of DNS had occurred. These calls were made at 2 weeks, 1 month, and 3 months after the discharge. If the patient re- ported any symptoms of DNS during phone interview, the patient was asked to revisit the hospital and brain MRI and neurologic consultation were performed and the final diagnosis was made by a neurologist.

    1. Propensity score matching

Propensity score (1:2 nearest neighbor) matching, with a caliper width of 0.05, was performed to analyze the effect of >3 HBO therapy sessions on the probability of DNS after acute CO poisoning compared to matched patients who received 3 HBO therapy sessions [9]. Variables that best predicted HBO use were selected using logistic regression to enter into the Propensity score analysis [9]. Multicollinearity was evalu- ated using the Durbin-Watson test. Creatine kinase, creatine kinase- myocardial band (CK-MB) and myoglobin, which all showed p < 0.01 in univariable logistic analysis, had multicollinearity, so we chose crea- tine kinase with the largest the area under the receiver operating char- acteristic curve (AUROC) value. Optimal cut-off levels for continuous variable were determined using Youden’s index, a measure of test per- formance (sensitivity + specificity -1). The selected variables for logis- tic regression were creatine kinase, Glasgow Coma Scale score and abnormal brain MRI findings. And after logistic regression analysis, Glasgow Coma Scale score and abnormal brain MRI findings remained for propensity score analysis. The total average treatment effect after matching was an acceptable (<0.1) standardized difference of propen- sity score predictors. A two-sided p-value <0.05 was considered statis- tically significant. Propensity score-matching analyses were performed with Stata 14.2 (StataCorp LP, College Station, TX, USA).

    1. Statistical analysis

IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, NY, USA), was used in our statistical analyses. Continuous variables are presented as means with standard deviations or medians with inter- quartile ranges, and categorical variables are expressed as numbers with percentages. The normality of the data was initially assessed using the Shapiro-Wilk test. The chi-square test or Fisher’s exact test was used to compare categorical variables, while the two-sample Student’s t-test or Wilcoxon rank-sum test was used to compare contin- uous variables. Statistical significance was defined as p < 0.05.

  1. Results

A total of 299 patients with acute CO poisoning visited our emer- gency department during the study period. Overall, 116 patients were excluded because they met the following criteria: age < 18 years (13 pa- tients), not underwent HBO therapy (22 patients), discharged against medical advice (13 patients), and loss to follow-up (68 patients). The re- maining 183 patients (61.2%) were enrolled in the study (Fig. 1).

Overall, 139 patients (76.0%) had intentional CO exposure and 33 patients (17.5%) exhibited DNS. In total, 166 patients (90.7%) were treated with 3 sessions of HBO therapy, and 17 patients (9.3%) were treated with >3 sessions of HBO therapy. The median number of ses- sions was 14 (13-14) in the >3 HBO group. Sex, age, body mass index, and comorbidities did not significantly differ between the two groups. The initial Glasgow Coma Scale score (15 [11-15] vs. 10 [9-14]) was sig- nificantly greater in the 3 HBO group. The duration of CO exposure (120 min [60-240] vs. 420 min [202.5-495]) was significantly greater in the >3 HBO group. In terms of laboratory findings, creatinine, crea- tine kinase, myoglobin, troponin I, and creatine kinase-myocardial band were higher in the >3 HBO group. Abnormal findings on brain MRI (23 [16.3%] vs. 6 [35.3%]) were significantly more common in the

Image of Fig. 1

Fig. 1. Flow diagram for the study population.

>3 HBO group. The development of DNS was more common in the >3 HBO group, but the difference between groups was not statistically

Table 1

Basic characteristics of the patients in the 3 and > 3 sessions of hyperbaric oxygen groups

3 HBO >3 HBO p-value

(n = 166) (n = 17)

Age, years 41 (31.3-54.8) 44 (42-54) 0.36

Men, n (%) 111 (66.9) 13 (76.5) 0.59

significant (27 [16.3%] in the 3 HBO group and 6 [35.3%] in the >3

HBO group; p = 0.089) (Table 1, Fig. 2).

neurological symptoms and signs in patients who developed DNS were followings: memory impairment (28 patients), cognitive impair- ment (25 patients), Parkinsonism (22 patients), concentration impair- ment (18 patients), personality change (15patients), ataxia (12 patients), incontinence (11 patients), insomnia (9 patients), anxiety (6 patients), and motor deficits (4 patients). MRI finding who developed DNS were followings: periventricular white matter (32 patients), cor- pus callosum (28 patients), subcortical U fibers (21 patients), external capsule (26 patients), and internal capsule lesion (27 patients).

Body mass index, kg/m2 23.1 (21.0-25.4) 23.99

(22.6-26.7)

Comorbidities (%)

Hypertension

17 (10.2)

1 (5.9)

>0.99

Diabetes

9 (5.4)

0 (0.0)

>0.99

Current smoker (%)

91 (58.0)

10 (58.8)

>0.99

Vital signs

Systolic blood pressure, mmHg

130

120 (110-150)

0.49

(116.3-140)

Diastolic blood pressure, mmHg

80 (70-90)

80 (70-100)

0.90

Heart rate/min

96.5 (80-104)

90 (84-98)

0.43

Respiratory rate/min

20 (19-20)

20 (18-20)

0.95

Oxygen saturation, %

98 (96-99)

98 (97-99)

0.47

Initial GCS score

15 (11-15)

10 (9-14)

0.01

Intentional exposure (%)

124 (74.7)

15 (88.2)

0.37

Symptoms (%)

Headache

19 (11.5)

0 (0.0)

0.22

Loss of consciousness

61 (36.8)

5 (29.4)

0.74

Dizziness

20 (12.1)

1 (5.9)

0.70

Dyspnea

9 (5.4)

1 (5.9)

>0.99

Chest pain

4 (2.4)

1 (5.9)

0.39

Laboratory findings

White blood cells, x103/mm3

12.0 (8.8-15.2)

11.7 (10.1-16.3)

0.51

Hemoglobin, g/dL

14.8 (13.4-16.1)

15.2 (14.4-15.8)

0.71

Creatinine, mg/dL

1.0 (0.9-1.2)

1.2 (1-1.3)

0.06

Total bilirubin, mg/dL

0.57 (0.44-0.81)

0.57 (0.42-0.92)

0.99

Creatine kinase, U/L

122.5

678 (137-5267)

0.01

(82.3-305)

arterial pH

7.41 (7.37-7.44)

7.4 (7.38-7.43)

0.76

Carboxyhemoglobin, %

10.55 (4.3-18.3)

6.5 (3.4-13.8)

0.11

C-reactive protein, mg/dL

0.11 (0.05-0.36)

0.38 (0.08-0.99)

0.04

Lactate, mmol/L

2.4 (1.5-4.2)

3 (2-5.1)

0.53

Myoglobin, ng/mL

45 (25-215.8)

625

<0.01

(146-2774.5)

CK-MB, ng/mL

2.4 (1.35-7.8)

16.8 (3.9-34.7)

0.01

Abnormal findings on brain MRI

23 (25.0)

13 (81.3)

<0.01

(%)

Development of DNS (%)

27 (16.3)

6 (35.3)

0.09

0.24

Logistic regression model used to select variables for propensity score matching is shown in Table 2. In multivariable analysis, Glasgow Coma Scale score (Odds Ratio [OR] 4.234, 95% confidence interval [CI] 1.016-17.652, p = 0.048) and abnormal brain MRI findings (OR 8.478, 95% CI 2.079-34.570, p = 0.003) were significant variables.

The propensity score model developed to balance these differences satisfied the overlap assumption: patients in the 3 HBO and > 3 HBO groups with similar propensity scores were available for matching. We successfully matched 32 patients from the 3 HBO group with 16 patients from the >3 HBO group. The c-statistic of the propensity-score model was 0.82, and all covariates were well-balanced between the two groups (standardized mean difference < 20% after matching). The con- founding factors of the characteristics were balanced and there were no significant differences between the matched groups. After matching, the groups showed similar incidences of DNS (28.1% vs. 31.3%, p > 0.99) (Table 3, Fig. 2).

  1. Discussion

HBO, hyperbaric oxygen; GCS, Glasgow Coma Scale; CK-MB, creatine kinase-myocardial band; MRI, magnetic resonance imaging; DNS, delayed neuropsychiatric sequelae.

The purpose of this study was to determine if the number of HBO sessions performed (3 vs. >3) influenced the prognosis of patients with acute CO poisoning, especially regarding the occurrence of DNS. In our study, the >3 HBO group had much more common brain damage on initial brain MRI, which suggested that these patients were likely to have a higher incidence of DNS. However, propensity score-matching analysis revealed no differences in DNS development between the 3 HBO and > 3 HBO groups.

Unlike other East Asian countries, South Korea has (until relatively recently) experienced few charcoal-burning suicides. However, after the September 2008 suicide by Charcoal burning of the South Korean ce- lebrity Ahn Jae-Hwan, charcoal burning as a method of suicide has gained rapid and sustained adoption in South Korea [10]. Therefore, it

Image of Fig. 2

Fig. 2. Comparison of delayed neuropsychiatric sequelae before and after propensity score matching.

has become important to prevent the development of DNS in patients with CO poisoning in South Korea.

Many studies have shown that multiple sessions of HBO therapy are more effective than a single session. A previous study reported that the clinical outcome was better for patients who received >=2 HBO therapy sessions compared to patients who received one-time HBO therapy [11]. In addition, a nationwide study showed that the survival rate was highest when patients received >5 HBO therapy sessions compared to 1 and 2-5 sessions [8]. Here, we used diffusion brain MRI findings as criteria for the selection of patients who received additional therapy sessions because diffusion-weighted imaging during the Acute phase of CO poisoning may help identify patients at risk of DNS [12]. Fortu- nately, the Korean National Health Insurance program covers the med- ical cost for up to 14 sessions of HBO therapy for patients with CO poisoning. Therefore, if high-risk patients agreed, we provided up to 14 sessions of HBO therapy.

With increased CO exposure time or greater CO intake, brain damage is more likely to be caused by hypoxia. In such patients, the develop- ment of DNS is also more likely [2,13]. However, the development of DNS is not simply explained by Hypoxic brain damage; it is considered to be caused by an adaptive immunological response to chemically modified myelin basic protein [14]. The incidence of DNS is reportedly 10-40% [5], and our study showed a slightly lower level of 18.0%. This may be because patients who did not receive HBO therapy or who were discharged against medical advice were excluded from the study.

HBO therapy has been reported to reduce the incidence of DNS in patients with CO poisoning [6]. The mechanism of HBO therapy in- volves prevention of leukocyte-mediated Inflammatory changes in the brain and inhibition of leukocyte B2 integrins, which mediate leukocyte-to-endothelial cell adhesions [15,16]. The prevention of B2 integrin-dependent neutrophil adherence occurs from exposure to 2 atm absolute [17]. However, hyperoxic seizures, aural barotrauma, and oxidative stress can occur due to high concentrations of oxygen and pressure [18]. Therefore, determination of the optimal number of HBO therapy sessions is critical.

In our study, the >3 HBO group consisted of patients who were pre- dicted to have a poor prognosis. This is because we recommend >3 ses- sions of HBO therapy when abnormal MRI findings are present. Jeon et al. showed that the presence of Acute brain lesions was significantly associated with the development of DNS [12]. Here, we used propensity score-matching analysis to eliminate this bias. Our results revealed no difference in the development of DNS between the two groups.

There are several limitations to this study. First, it was conducted at a single emergency medical center, which limits the generalizability of the findings. Second, 68 patients (58.6%) of the excluded patients were lost for follow-up. This is because almost all of them (65 patients) were suicide attempters and these patients were not only uncoopera- tive for the treatment but also showed poor compliance for the follow-up. Third, since there is no established diagnostic criteria of DNS yet, we made the diagnosis via brain MRI and neurologic

Table 2

Logistic regression analysis for selecting variables that best predicted HBO therapy

Univariable Multivariable

Odds ratio (95% CI) p-value Odds ratio (95% CI) p-value Creatine kinase >656.5 U/L 6.317 (2.234-17.868) <0.001 2.831 (0.798-10.044) 0.107

Initial GCS score < 14 7.682 (2.126-27.753) 0.002 4.234 (1.016-17.652) 0.048

Abnormal findings on brain MRI 13.565 (3.551-51.821) <0.001 8.478 (2.079-34.570) 0.003

HBO, hyperbaric oxygen; CI, confidence interval; MRI, magnetic resonance imaging.

Table 3

Comparison of the study population after propensity score matching

3 HBO

>3 HBO

p-value

(n = 32)

(n = 16)

Age, years

45.3 +- 17.3

45.4 +- 11.5

0.98

Men, n (%)

22 (68.8)

12 (75.0)

0.75

Body mass index, kg/m2 23.6 +- 3.7 24.2 +- 3.9 0.63

Comorbidities (%)

Hypertension 1 (3.1) 1 (6.3) >0.99

Diabetes 1 (3.1) 0 (0.0) >0.99

Current smoker (%)

19 (59.4)

9 (56.3)

>0.99

Vital signs

Systolic blood pressure, mmHg

129.6 +- 22.3

122.5 +- 25.4

0.35

Diastolic blood pressure,

80 (74-90)

80 (70-92.5)

0.63

mmHg

Heart rate/min

96.9 +- 14.5

91.8 +- 16.3

0.29

Respiratory rate/min

20 (19-20)

20 (18-20)

0.96

Oxygen saturation, %

98 (95-98)

98 (96.8-98.3)

0.46

Initial GCS score

13.5 (9.8-15)

11 (9-14)

0.18

Intentional exposure (%)

27 (84.4)

14 (87.5)

>0.99

Symptoms (%)

Credit author statement

All authors have read and agreed to the published version of the manuscript.

Declaration of Competing Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgment

This work was supported by the Soonchunhyang University Research Fund.

References

  1. Prockop LD, Chichkova RI. Carbon monoxide intoxication: an updated review. J Neurol Sci. 2007;262:122-30.
  2. Weaver LK. Carbon monoxide poisoning. N Engl J Med. 2009;360:1217-25.

Headache

1 (3.1)

0 (0.0)

>0.99

[3] Thom SR, Taber RL, Mendiguren II, Clark JM, Hardy KR, Fisher AB. Delayed

Loss of consciousness

9 (28.1)

4 (25.0)

>0.99

neuropsychologic sequelae after carbon monoxide poisoning: prevention by treat-

Dizziness

4 (12.5)

1 (6.3)

0.65

Dyspnea

1 (3.1)

1 (6.3)

>0.99

Chest pain

0 (0.0)

1 (6.3)

0.33

Laboratory findings

White blood cells, x103/mm3

14.73 +- 6.49

13.33 +- 4.65

0.40

Hemoglobin, g/dL

14.75 (14.2-16)

15.3 (14.2-15.83)

0.93

Creatinine, mg/dL

1.1 (0.9-1.42)

1.2 (0.98-1.3)

0.91

Total bilirubin, mg/dL

0.7 (0.5-0.92)

0.57 (0.4-1.02)

0.47

Creatine kinase, U/L

204.5

850.5

0.26

(90-4213.5)

(161-6106.25)

Arterial pH

7.41 (7.37-7.44)

7.4 (7.38-7.43)

0.72

Carboxyhemoglobin, %

7.1 (3.75-13.18)

6.9 (2.92-14)

0.80

C-reactive protein, mg/dL

0.22 (0.1-0.82)

0.4 (0.15-1.73)

0.39

Lactate, mmol/L

3 (1.85-4)

3 (2-5.1)

0.96

Myoglobin, ng/mL

119

726 (252.5-3052)

0.06

(35.5-952.5)

CK-MB, ng/mL

5.9 (1.65-27.8)

17.95 (4.12-38.95)

0.15

Abnormal findings on brain MRI

23 (71.9)

13 (81.3)

0.73

(%)

Development of DNS (%)

9 (28.1)

5 (31.3)

>0.99

HBO, hyperbaric oxygen; GCS, Glasgow Coma Scale; CK-MB, creatine kinase-myocardial band; MRI, magnetic resonance imaging; DNS, delayed neuropsychiatric sequelae.

consultation. Fourth, we used propensity score matching to correct se- lection bias, but this could not totally correct the bias. A well-designed randomized controlled trial is needed in the future to overcome these limitations.

  1. Conclusions

HBO therapy may be performed to prevent the development of DNS in patients with acute CO poisoning. However, there is no clear consen- sus regarding the optimal number of HBO therapy sessions and number of sessions can affect prognosis. Thus, determining the optimal number of HBO therapy sessions is critical for the treatment of patients with CO poisoning. We found no difference in the incidence of DNS after 3 and > 3 sessions of HBO therapy.

ment with hyperbaric oxygen. Ann Emerg Med. 1995;25:474-80.

  1. Sykes OT, Walker E. The neurotoxicology of carbon monoxide-historical perspective and review. Cortex. 2016;74:440-8.
  2. Raphael J-C, Elkharrat D, Jars-Guincestre M-C, Chastang C, Vercken J-B, Chasles V, et al. Trial of normobaric and hyperbaric oxygen for acute carbon monoxide intoxi- cation. Lancet. 1989;334:414-9.
  3. Weaver LK, Hopkins RO, Chan KJ, Churchill S, Elliott CG, Clemmer TP, et al. Hyper- baric oxygen for acute carbon monoxide poisoning. N Engl J Med. 2002;347: 1057-67.
  4. Wang W, Cheng J, Zhang J, Wang K. Effect of hyperbaric oxygen on neurologic se- quelae and all-cause mortality in patients with carbon monoxide poisoning: a meta-analysis of randomized controlled trials. Med Sci Monit. 2019;25:7684-93.
  5. Huang C-C, Ho C-H, Chen Y-C, Lin H-J, Hsu C-C, Wang J-J, et al. Hyperbaric oxygen therapy is associated with lower short-and long-term mortality in patients with car- bon monoxide poisoning. Chest. 2017;152:943-53.
  6. Ali MS, Groenwold RH, Belitser SV, Pestman WR, Hoes AW, Roes KC, et al. Reporting of covariate selection and balance assessment in propensity score analysis is subop- timal: a systematic review. J Clin Epidemiol. 2015;68:122-31.
  7. Chen Y-Y, Yip PS, Chan CH, Fu K-W, Chang S-S, Lee WJ, et al. The impact of a celebrity’s suicide on the introduction and establishment of a new method of suicide in South Korea. Arch Suicide Res. 2014;18:221-6.
  8. Gorman D, Clayton D, Gilligan J, Webb R. A longitudinal study of 100 consecutive ad- missions for carbon monoxide poisoning to the Royal Adelaide Hospital. Anaesth In- tensive Care. 1992;20:311-6.
  9. Jeon S-B, Sohn CH, Seo D-W, Oh BJ, Lim KS, Kang D-W, et al. Acute brain lesions on magnetic resonance imaging and Delayed neurological sequelae in carbon monoxide poisoning. JAMA Neurol. 2018;75:436-43.
  10. Pepe G, Castelli M, Nazerian P, Vanni S, Del Panta M, Gambassi F, et al. Delayed neu- ropsychological sequelae after carbon monoxide poisoning: predictive risk factors in the emergency department. A retrospective study. Scand J Trauma Resusc Emerg Med. 2011;19:16.
  11. Thom SR, Bhopale VM, Fisher D, Zhang J, Gimotty P. Delayed neuropathology after carbon monoxide poisoning is immune-mediated. Proc Natl Acad Sci U S A. 2004; 101:13660-5.
  12. Tomaszewski C, Rudy J, Wathen J, Brent J, Rosenberg N, Kulig K. Prevention of neu- rologic sequelae from carbon monoxide by hyperbaric oxygen in rats. Ann Emerg Med. 1992;21:631-2.
  13. Thom SR. Functional inhibition of leukocyte B2 integrins by hyperbaric oxygen in carbon monoxide-mediated brain injury in rats. Toxicol Appl Pharmacol. 1993; 123:248-56.
  14. Thom SR. Leukocytes in carbon monoxide-mediated brain oxidative injury. Toxicol Appl Pharmacol. 1993;123:234-47.
  15. Tibbles PM, Edelsberg JS. Hyperbaric-oxygen therapy. N Engl J Med. 1996;334: 1642-8.