Seizure frequency in more than 180,000 treatment sessions with hyperbaric oxygen therapy - a single centre 20-year analysis.
Level 4 - case-series / case-control
Historically controlled retrospective cohort study evaluating a protocol change
PubMed 31523791 · doi:10.28920/dhm49.3.167-174
What was done
A 20-year (1999–2018) single-centre retrospective analysis evaluated all consecutive hyperbaric oxygen therapy (HBOT) sessions to measure central nervous system oxygen toxicity seizures. Medical records were reviewed for patient demographics, comorbidities, treatment indications, and seizure characteristics. Seizure frequency was compared before versus after incorporating a routine 5-minute air break (5´-AIRBK) into the treatment protocol using Chi-square testing.
What was found
Across 188,335 HBOT sessions (74,255 before vs. 114,080 after introducing the air break), 43 seizures were documented. Seizure frequency decreased from 29 events before the air break (3.9 per 10,000 treatments) to 14 events after its introduction (1.2 per 10,000 treatments; P < 0.0001). Seizures occurred at a median of 57 minutes (range 15–85) following compression and after a median of 21 sessions (range 1–126). Indications among seizure cases were diabetic ulcer (n = 11), acute traumatic peripheral ischaemia (ATPI, n = 6), non-diabetic ulcer (n = 5), sudden sensorineural hearing loss (n = 5), chronic refractory osteomyelitis (n = 5), radionecrosis (n = 3), necrotising fasciitis (NF, n = 2), and haemorrhagic cystitis (n = 1), with higher relative frequency in ATPI and NF.
Why it matters
This large dataset demonstrates that oxygen-induced seizures are rare in HBOT and provides evidence that routine 5-minute air breaks significantly mitigate the risk of central nervous system oxygen toxicity.
Limits
The study is retrospective and single-centre, relying on historical controls across a 20-year span where other clinical variables may have changed. The abstract reports total treatment sessions rather than unique patient counts, and confounding factors such as exact chamber pressures or concurrent medications are not reported in the abstract.