ABSTRACT
INTRODUCTION: Hydrogen peroxide ingestions cause significant morbidity and mortality due to oxygen gas emboli and are treatable with hyperbaric oxygen therapy. Recommendations for observation are based on small case series. OBJECTIVES: The aim of this systematic review is to define the time of onset of embolic phenomena after hydrogen peroxide exposure and to describe the proportion of patients who received hyperbaric oxygen therapy. METHODS: Cases from a systematic literature search were combined with those from a prior study that used data derived from the American Association of Poison Control Centers National Poison Data System. Air-gas emboli were defined as embolic phenomena (stroke, myocardial infarction, obstructive shock) potentially reversed with hyperbaric oxygen therapy. Simple counts, mean, and interquartile range were used for description and comparisons. RESULTS: A total of 766 records were identified in the literature search. Three-hundred and eighty-three duplicate records were identified and removed. Of the 383 remaining records, 156 met inclusion criteria; 88 were excluded based on predetermined criteria yielding 68 records with 85 unique cases. Forty-one cases were extracted from the 2017 National Poison Data System study resulting in a total of 126 cases for analysis. Case descriptions: We analyzed these 126 cases and documented 213 discrete clinical events, excluding deaths. There were 108 high-concentration exposures, 10 low-concentration exposures, and 8 were unknown. Thirty-five cases were intentional ingestions but not for self-harm, and 84 were unintentional or accidental. Only 4 cases were for self-harm, and there were 23 pediatric cases. There were 99 air-gas emboli reported in 78 patients. Time to onset: The time to onset of air-gas embolic was documented in 70/78. Time to symptom onset ranged from immediate to 72 h after hydrogen peroxide exposure. Over 90% of embolic symptoms occurred within 10 h of ingestion. Hyperbaric oxygen therapy: A total of 54/126 cases received hyperbaric oxygen therapy. Of those 54 cases, 31 had primary portal venous gas while the remaining 23 had air-gas emboli. Of the 23 air-gas emboli cases treated with hyperbaric oxygen therapy, 13 made full recoveries while 10 had residual symptoms or died. Mean time from air-gas emboli symptom onset to hyperbaric oxygen therapy in the full recovery group was 9 h compared to 18.2 h in the partial recovery/death group. Portal venous gas: There were 63 total reported cases of portal venous gas. Forty-nine of these cases were primary portal venous gas, 13 were secondary findings in patients with air-gas emboli and one case was secondary to non-air-gas emboli symptoms. Twenty-seven of 49 patients with portal venous gas (55%) as the primary finding had gastrointestinal bleeding. Thirty of the 63 cases received hyperbaric oxygen therapy for portal venous gas without any documented air-gas emboli. Deaths: Seventeen deaths occurred in the combined cohort. Of these, 13 were associated with high-concentration exposures. All deaths with reported time to symptom onset had symptoms within 1 h of exposure. CONCLUSION: This review of hydrogen peroxide exposure cases suggests that clinically significant embolic phenomena occur within 10 h of exposure, although delayed air-gas emboli do happen and should considered when deciding duration of observation. It remains equivocal whether hyperbaric oxygen therapy is beneficial in cases of primary portal venous gas without systemic involvement.