Improvement in gross motor function measure following a long course of hyperbaric oxygenation in children with cerebral palsy.

1. Some indication of benefit from HBOT with cerebral palsy in respect to gross motor function measures.
2. No clear benefit for either spasticity or speech scores, or SPECT findings.
3. Very small study which has not been peer reviewed.

Citation/s:
1. Mathai SS, Bansali P, Singh Gill B, Nagpal S, John MJ, Aggarwal H, Bhatt V. Effects of hyperbaric oxygen therapy in children with cerebral palsy. Proceedings of the International Conference on Diving and Hyperbaric Medicine, Barcelona 7-10 September 2005:193-197.

Three-part Clinical Question: In children with cerebral palsy, does hyperbaric oxygen improve function?
Search Terms: Cerebral palsy

The Study:
Double-blinded randomised controlled trial with intention-to-treat.
The Study Patients: Children with cerebral palsy aged 1 to 10 years and able to attend chamber over several months.
Control group (N = 9; 9 analysed): Compression to 1.5 ATA breathing air for one hour daily for 5 weeks (30 treatments). Repeated 3 times with one month interval (total 90 treatments).
Experimental group (N = 11; 11 analysed): As above, but breathing 100% oxygen.

The Evidence:

Outcome

Time to Outcome

Sham rate

HBOT rate

Relative risk reduction

Absolute Risk Reduction

NNT

Improvement in upper limb spasticity score

One month

0.333

0.364

9%

0.03

32

95% Confidence Intervals:

135% to -100%

0.45 to -0.39

NNT = 2 to INF; NNH = 3 to INF

Improvement in lower limb spasticity score

One month

0.222

0.455

105%

-0.23

4

95% Confidence Intervals:

285% to -75%

0.63 to -0.17

NNT = 2 to INF; NNH = 6 to INF

Improved SPECT scan

One month

0.222

0.545

145%

0.32

3

95% Confidence Intervals:

326% to -35%

0.72 to -0.08

NNT = 1 to INF; NNH = 13 to INF

 

Measure

Sham Group

HBOT Group

Difference

95% CI

Mean

SD

Mean

SD

Mean change in gross motor function measure

5.8

2.9

14.6

4.2

8.8

12.3 to 5.3

 

Non-Event Outcomes

Time to outcome/s

Control group

Experimental group

P-value

Change in speech score

One month

pre: 2.0  +/- 0.7

post: 2.4  +/- 0.8

pre: 2.2  +/- 0.5

post: 3.1  +/- 0.5

>0.05

<0.05

Comments:
1. No sample size calculation and this study has very low power to detect even quite large differences between groups.
2. Given the method of randomisation by sealed envelopes, it is not clear why the groups are not equal in numbers.
3. Long course of HBOT is unusual - 90 treatments over six months.
4. No plausible mechanism of action and the authors imply that most cases of CP are due to hypoxia in the perinatal or early postnatal period, rather than developmental in aetiology.
5. Many potentially important confounders have not been considered (e.g. Apgar scores, prematurity) and there is a strong possibility of bias given the small numbers.

6. Clinical significance of the improvement in GMFM requires expert interpretation.

Appraised by: Sean Hopson, Mike Bennett, POWH Sydney ; Friday, 11 August 2006
Email: m.bennett@unsw.edu.au
Kill or Update By: August 2007