Protocol draft

The efficacy of Comex 30 He-O2 vs. US Navy treatment tables for DCI

Jonas Hink, MD, Royal Danish Navy
Erik C. Jansen, MD, Dr.Med.Sci, Director of Hyperbaric medicine, Rigshospitalet, Denmark.


In many centers the US Navy table 6 is the initial and routinely applied treatment for Decompression Illness (DCI). The Comex 30 table is used as a next step, if the initial treatment is not considered sufficient. The scope of the investigation is to support the choice of initial treatment procedure.

Background information

The definite treatment of decompression illness is pressure and oxygen. A wide variety of hyperbaric regimens have been described, but there are to our knowledge no human outcome data obtained in prospective randomized studies comparing the various regimens. However, the US Navy oxygen treatment tables with initial recompression to 18 msw are the treatment tables most widely used and also the "gold standards" against which other treatment tables are compared (Consensus, 45. workshop of UHMS, Treatment of Decompression Illness, 1996, 420-426).

The principles of the US Navy oxygen treatment tables for decompression illness treatment beginning at the surface are initial compression to 18 msw breathing 100 % oxygen (USN table 6). Compression deeper than 18 msw is considered an option available for instances of inadequate clinical response at the initial evaluation at 18 msw (Moon, SPUMS J 1998, 28 (4), 222-231). Further compresion after the initial evaluation could follow USN table 6A (50 msw) using oxygen-nitrogen (nitrox) 40/60 as a treatment gas mixture.

Some investigators have claimed to have good results with initial treatments deeper than 18 msw (Lee et al., J Hyperbaric Med 1991, (6), 11-17; Imbert, Treatment of Decompression Illness, 1996, 389-393; Overlock et al., Treatment of Decompression Illness, 1996, 106-121) and computer simulations have supported this approach (Flook & Brubakk, Undersea Hyperbaric Med 1996, 23 (suppl), 64).

The Comex 30 He-O2 (heliox) table is a recompression procedure for treatment of decompression illness originally designed by Dr. Fructus (Imbert, Treatment of Decompression Illness, 1996, 389-393). The regimen starts with recompression to 30 msw using helium-oxygen (50/50) as a treatment gas mixture. The decompression that follows includes stops at 24, 18 and 12 msw. From 18 msw to the surface the treatment gas is 100% oxygen. During the decompression schedule several episodes of air-breathing for 5 mins (air-breaks) have been included to reduce the risk of oxygen toxicity.

The use of heliox as a treatment gas mixture during recompression have some theoretical advantages as pointed out by Hills and James (Report of proceedings of a symposium on decompression sickness, EUBS, 1981, 127-162). In lipid tissue, the outward flux of nitrogen from bubbles exceeds the inward flux of helium, whether the exchange of gases in the tissue is limited by perfusion or diffusion.

The theoretical benefits of heliox breathing are supported by animal studies (Hyldegaard & Madsen, Undersea Biomed Res 1989, 16 (3), 185-193; Hyldegaard et al., Undersea Biomed Res 1991, 18 (5-6), 361-371; Hyldegaard et al., Undersea Hyperbaric Med 1994, 21 (2), 115-128).

A prospective randomized controlled study of oxygen versus oxygen-helium in the treatment of decompression illness have been underway since 1992 (Drewry & Gorman, Undersea Hyperbaric Med 1994, 21 (suppl), 98) but to our knowledge any final report have never been published.

Non-randomized clinical studies (Shupak et al., Arch Neurol 1997, 54 (3), 305-311) suggests that heliox breathing during recompression might be of benefit in the treatment of neurological decompression illness.

In aqueous surroundings, nitrogen bubbles would be expected to grow during heliox breathing if the gas exchange is limited by diffusion (Hyldegaard & Madsen, Treatment of Decompression Illness, 1996, 313-328). However, neither animal studies nor case-reports are known to us that for safety reasons warrants against the use of heliox in the treatment of decompression illness.

Any pharmacological effect of helium unrelated to gasdiffusion can not be ruled out (Moon, SPUMS J 1998, 28 (4), 222-231).

Summary of the known and potential risks and benefits

The known and potential benefits of the Comex 30 He-O2 treatment table are:

  1. An increase in environmental pressure (though less than during USN 6A) and thus reduction in bubble size according to Boyles Law.
  1. A decrease in nitrogen partial pressure in arterial blood and thus an increase in the gradient for inert gas removal.
  2. An increase in the oxygen content of arterial blood (though less than during USN 6) and thus an increased tissue oxygenation.
  3. A bubble-elimination effect in lipid tissue due to the faster diffusion of nitrogen out of the bubbles than the diffusion of helium into the bubbles.
  4. Pharmacological effects of oxygen (though a lower oxygen partial pressure than during USN 6 or 6A nitrox 40/60) such as vasoconstriction and antiinflammatory properties.
  5. The absent of nitrogen narcosis.
  6. Unknown (if any) pharmacological effect of helium.

The known and potential risks of the Comex 30 He-O2 treatment table are:

  1. A bubble-enlarging effect in aqueous surroundings due to the faster diffusion of helium into the bubbles than the diffusion of nitrogen out of the bubbles.
  2. Toxicological effects of oxygen (though a lower oxygen partial pressure than during USN 6 or 6A nitrox 40/60).
  3. Unknown (if any) toxicological effects of helium.

In summary, there is a sound rationale for the use of Comex 30 He-O2 table in the treatment of decompression illness (investigatorīs brochure need to be prepared).

Trial objectives and design

The trial is planned to be a parallel group design with only two arms in the form of an active control trial with the objective of showing superiority of the investigational product (Comex 30 He-O2 table) to the active control (USN table(s) 6/6A), and thereby showing efficacy. The trial needs to be randomized, single or double-blinded and should preferably be conducted as a multicentre and multiinvestigator trial using multiple measurements as the primary endpoints and a global assessment variable as a secondary endpoint.

The trial has to be conducted in compliance with the protocol, Good Clinical Practice and the applicable regulatory requirements.

Three different designs are proposed:

1. Diagnosis of decompression illness and then randomized allocation into one of two arms being

    1. Comex 30 He-O2 table.
    2. USN table 6 (extended).

2. Diagnosis of decompression illness and then randomized allocation into one of two arms being

    1. Comex 30 He-O2 table.
    2. USN table 6 (extended).

Once at depth initial evaluation should be performed within 5 mins.

In case of inadequate clinical response at the initial evaluation the cases allocated to USN table 6 is converted to USN table 6A nitrox (40/60), whereas the cases allocated to Comex 30 He-O2 table is to be continued on Comex 30 He-O2 table.

In case of adequate clinical response at the initial evaluation no conversion in treatment table is made.

3. Diagnosis of decompression illness and initial compression using USN table 6. Once at depth initial evaluation should be performed within 5 mins. Hereafter stratification into two groups being

    1. Adequate clinical response.
    2. Inadequate clinical response.

Each group is then randomly allocated into one of two arms being

    1. Comex 30 He-O2 table.
    2. USN table 6 (extended).

The clinical criteria for adequate and inadequate response needs to be established.

The physician determining whether the clinical response is adequate or inadequate will not be involved in any way in the evaluation of primary and secondary variables.

Follow-up treatments should follow USN table 6 for all three design proposals. Criteria for follow-up treatments needs to be established.

Timing of the last clinical evaluation needs to be determined.

Randomization should be performed for each centre (blockrandomisation) and techniques for this needs to be determined. Preferably, centres with an equal number of expected patients should be chosen.

Blinding of the physician performing the evaluation of primary and secondary variables is imperative and procedures for this must be established.

Blinding of the patients is difficult, especially in design proposal 2 and 3. However, techniques should be considered including blinding of gauges and watches, ventilation of recompression chambers creating sounds simulating compression/decompression, minute pressure excursions at fixed time intervals simulating compresion/decompression etc.

Sample size determination and a statistical analysis plan will be prepared after final determination of endpoints and trial designs.

Outcome measures

Three primary variables are proposed:

1. Measurements of maximum voluntary isometric contraction (MVIC) in selected muscle groups using computer-based, fixed strain-gauges are recommended means of testing for ALS clinical trials and might also be used as means of testing for decompression illness clinical trials.

Advantages includes:

    1. High intra-rater and inter-rater reliability with well trained evaluators.
    2. Sensitivity.
    3. Accurate measurement of muscle strength in weak as well as strong muscles.

Disadvantages includes:

    1. Lack of standardized test procedures.
    2. Vigorous quality assurance procedures needed.
    3. Requires extensive training and equipment.

(Brinkmann, J Neurol Science 1997, 147 (1), 97-111).

Time-consumption and muscle groups to be tested should be considered before implementation.

2. Kinematic tasks (motions) analysed by an optoelectronic three-dimensional motion analyzer based on active infrared light emitting markers.

Tasks such as finger tapping, aiming movements and dual-task conditions have been investigated as measures of performance in high-functioning stroke patients. The method proved to be sensitive with a marked responsiveness to change and a test-retest reliability high enough for follow-up group studies but not high enough for assessment of individuals (Platz et al., Arch Phys Med Rehab 1999, 80 (3), 270-277. Similar tasks could be useful as primary variables in our trial.

3. The sharpened Romberg test conducted in a standardised manner and scored as the best attempt of four has been evaluated in injured divers. The test is considered a valuable examination in divers suspected of having decompression illness, and performed in a standardised manner it might be a useful indicator of illness and recovery (Fitzgerald, SPUMS J 1996, 26 (3), 142-146; Gorman & Fitzgerald, Undersea Hyperbaric Med 1996, 23 (1), 55).

The Romberg test may be measured i quantitative terms by use of a force plate (Jansen 1988).

Prior to the implementation of any of the proposed primary variables there should be sufficient evidence that they can provide a valid and reliable measure of some clinically relevant and important treatment benefit in the patient population described in our trial (ICH Harmonised Tripartite Guideline, 1998).

Proposals for secondary variables:

Residual neurological features can be used as secondary endpoints in the form of global assessment variables. From the investigators overall impression based on the clinical assessment at follow-up the severity of residual neurological features could be categorized into three groups (severe, mild, none) as described by Pitkin et al., Aviat Space Environ Med 1999, 70 (5), 517-521.

Severe: Neurological sequelae with functional impact on normal everyday life including any form of motor weakness, uncoordination, gait or urinary disturbance and objective sensory deficits of functional importance (e.g. in the hands).

Mild: All residual neurological features not classified as severe (typically parasthesiae or subjectively altered sensation).

None: Complete resolution of all neurological symptoms and signs.

Residual neurological features as secondary variables are supportive measurements related to the primary objective. However, their roles in interpretation of trial results needs to be determined and the validity and reliability needs to be evaluated.

Evaluation of both primary and secondary variables should be performed by only one physician for each patient.

Population and criteria

The population to be studied covers recreational scuba divers in the European countries.

Inclusion criteria:

  1. Indication for treatment of decompression illness based on a complete history and a thorough examination performed by a physician trained in diving and hyperbaric medicine as recognised by the trial investigators. Diagnosis on basis of
    1. A history compatible with decompression illness including a hyperbaric exposure followed by decompression.
    2. Symptoms compatible with decompression illness according to the literature.
    3. Time-latency between a and b within recognised "limits" taking into account the possible role of any provocative event such as flying after diving.
  2. The actual diagnosis of decompression illness is related to a scuba diving episode using air or nitrox as a breathing gas.

Exclusion criteria:

  1. Unconscious patients and patients not able to give informed consent.
  2. Patients participating in other clinical trials.
  3. Patients given hyperbaric treatment for the actual indication before recruitment in the trial.



Study protocol NOT finalised !!!
Send comments or expressions of interest to: Christian Risby Mortesnsen