Patients
undergoing partial hepatectomy show modified neutrophil activity following
post-operative hyperbaric oxygenation.
Clinical
bottom line:
|
1. Clinical significance
unclear. |
Citation:
1. Ueno S,
Tanabe G, Kihara K, Aoki D, Arikawa K, Dogomori H, Aikou T. Early postoperative
hyperbaric oxygen therapy modifies neutrophile activation. Hepato-Gastroenterology
1999; 46:1798-1799.
Lead author's name and
fax:
Shinichi Ueno, Fax: +81 992 65 7426
Three-part
Clinical Question: In patients undergoing major surgery does hyperbaric oxygen therapy
modify post-operative sinusoidal endothelial damage caused by activated
neutrophils?
Search Terms: hyperbaric
oxygenation, endothelium, neutrophil
The Study:
Non-blinded
randomised controlled trial without intention-to-treat.
Study Patients: Non-cirrhotic
patients undergoing elective hepatectomy for liver cancer.
Control group:
(N = 12;
12 analysed): Standard post-operative care to maintain normal haemodynamic
values.
Experimental group:
(N = 12;
12 analysed): Above, plus 2 sessions breathing 100% oxygen at 2ATA for 60
minutes at 3 hours and 24 hours post-operatively.
The
Evidence:
|
Outcome
|
Time
to Outcome |
Control
group |
HBO
group |
Relative
risk reduction |
Absolute
risk reduction |
NNT
|
|
serum total bilirubin >5mg/dl for 3 d |
3
days |
0.250 |
0 |
100% |
0.25 |
4 |
|
95%
Confidence Intervals: |
|
0.01
to 0.50 |
2
to 200 |
|||
|
Hepatic failure |
3
days |
0 |
0 |
|
0.00 |
|
|
95%
Confidence Intervals: |
|
0.00
to 0.00 |
|
|||
|
Intraperitoneal infection |
3
days |
0.083 |
0 |
|
|
|
|
95%
Confidence Intervals: |
|
|
|
|||
|
Non-Event
Outcomes |
Time
to outcome |
Control
group |
HBO group |
P-value |
|
PMNL elastase
(mcg/L). |
12
hrs |
870 |
500 |
<0.05 |
|
Thrombomodulin
(U/ml) |
12
hrs |
25 |
18 |
<0.01 |
|
Intensity of
fluorescence in CD18 (nm) |
12
hrs |
>10%
increase |
>10%
decrease |
0.003 |
Comments:
1. Difficult to appraise, with minimal information/explanations throughout.
2. Findings poorly explained and discussed.
3. No apparent allowance for multiple testing of dependent data for non-event
outcomes.
Appraised by:
Peter Hodkinson. Edinburgh Medical School. UK. Mike Bennett Sydney, Australia;
Wednesday, October 29, 2003
Email: petehodkinson@hotmail.com
Kill or Update By: November 2006