Li2

Some evidence of benefit from the addition of hyperbaric oxygen therapy in the treatment of non-traumatic intracerebral haemorrhage.

Clinical Bottom Line:

1. The addition of hyperbaric oxygen reduced death rate 2. The modified Rankin Score for disability was improved with HBO but not the Barthel Index.

---

Citation/s:1. Li X, Li J, Yang X, Sun Z, Zhang J, Zhao W, Dong S, Li C, Ye Y, Chen J, Li Y, Xiang Y, Mao J, Li G, Guo H, Zhang W, Guo H, Zhang Y, Zhang M, Zhang W, Xu Z, Zhao B, Wei J, Zhao G, Ma R, Shen X, Ge C, Zheng C, Li S, Wang Y. Hyperbaric-Oxygen Therapy Improves Survival and Functional Outcome of Acute Severe Intracerebral Hemorrhage. Arch Med Res. 2017 Oct;48(7):638-652. doi: 10.1016/j.arcmed.2018.03.001. Epub 2018 Mar 13. PMID: 29548729
Lead author's name and fax: Xiaowei Li lxwljz@163.com

Three-part Clinical Question:For patients with non-traumatic intracerebral haemorrhage, does the addition of hyperbaric oxygen to normal care result in any improved outcome?
Search Terms: intracranial haemorrhage; stroke; gastrointestinal haemorrhage

The Study:Double-blinded concealed randomised controlled trial with intention-to-treat.
The Study Patients: Adult patients with acute severe hypertensive basal ganglia haemorrhage who received emergency craniotomy or decompressive craniectomy with simultaneous evacuation of hematoma immediately after admission.
Control group (N = 113; 113 analysed): All standard care including blood pressure control, proton pump inhibitors, mechanical ventilation and urokinase if rebleeding, plus sham hyperbaric from day eight breathing air with a brief excursion to 1.34 ATA at the start of each 100 minute session daily. Patients had three cycles of 30 sessions each with seven day breaks between cycles - total of 90 sessions each.
Experimental group (N = 452; 426 analysed): As above but breathing 100% oxygen at different pressures and number of sessions. Grp B: two cycles of 30 treatments at 2.0 ATA; Grp C: three cycles of 30 sessions at 2.0 ATA; Grp D: two cycles of 30 treatments at 1.5 ATA; Grp E: three cycles at 1.5 ATA.

The Evidence:

Outcome	Time to Outcome	Control group	HBO groups	Relative risk reduction	Absolute risk reduction	NNT
Death	6 months	0.35	0.142	60%	0.21	5
	95% CIs:			33% to 86%	0.12 to 0.31	3 to 8
Gastrointestinal bleeding	6 months	0.43	0.52	-22%	-0.1	-11
	95% CIs:			-46% to 2%	-0.20 to 0.01	NNT = 110 to INF; NNH = 5 to INF

 

Measure	Control Group		HBO Group C		Difference	95% CI
	Mean	SD	Mean	SD
Modified Barthel Index at 6 months (0 total dependence to 100 independent)	37.8	24.5	56.47	28.0	18.7	-137.2 to 174.6
Modified Rankin Score at 6 months (0 best to 10 dead)	4.5	1.4	3.4	1.8	1.17	0.55 to 1.79

Comments:
1. 30 authors have made an intellectual contribution to this study.
2. 5 arms to this complex study - four of which received HBOT and we have compared the sham control to all groups for event outcomes.
3. 32 patients were withdrawn due to intolerance of the treatment at some stage but included with worst case assumption in the ITT used here.
4. Per-protocol analyses produced similar results.
5. Disability scores in the tables above compare control to Group C. In general, groups B and C did better than D and E suggesting 2.0 ATA HBO may be superior to 1.5 ATA.

Appraised by:Mike Bennett m.bennett@unsw.edu.au ; Friday, 8 October 2021
Kill or Update By: October 2023