Arterial oxygen tension increase 2-3 h after hyperbaric oxygen therapy: a prospective observational study

Inhalation of hyperbaric oxygen (HBO) has been reported to decrease arterial oxygen tension (PaO(2)) in the early period after exposure. The current investigation aimed at evaluating whether and to what extent arterial blood gases were affected in mechanically ventilated intensive care patients within 6 h after HBO treatment.

Arterial blood gases were measured in 11 ventilated subjects [nine males, two females, synchronized intermittent mandatory ventilation (SIMV) mode] undergoing HBO therapy for necrotizing soft tissue infection (seven patients), burn injury (two patients), crush injury (one patient) and major abdominal surgery (one patient). Blood gases were obtained with the patients in the supine position under continuous analgesia and sedation before the hyperbaric session (baseline), during isopression, after decompression, after each transport, and 1, 2, 3 and 6 h after exposure. Heart rates and blood pressures were recorded. Intensive care unit (ICU) ventilator settings remained unchanged. Transport and chamber ventilator settings were adjusted to baseline with maintenance of tidal volumes and positive end-expiratory pressure (PEEP) levels. The hyperbaric protocol consisted of 222.9 kPa (2.2 absolute atmospheres) and a 50-min isopression phase. The paired Wilcoxon's test was used.

Major findings (median values, 25%/75% quartiles) as per cent change of baseline: PaO(2) values decreased by 19.7% (7.0/31.7, P < 0.01) after 1 h and were elevated over baseline by 9.3% (1.5/13.7, P < 0.05) after 3 h. SaO(2), alveolar-arterial oxygen tension difference and PaO(2)/FiO(2) ratio behaved concomitantly. Acid-base status and carbon dioxide tension were unaffected.

Arterial oxygen tension declines transiently after HBO and subsequently improves over baseline in intensive care patients on volume-controlled mechanical ventilation. The effectiveness of other ventilation modes or a standardized recruitment manoeuvre has yet to be evaluated.