Acta anmsth. scand. 1980, 24, 265-271

A Technique for Disinfection of the MDO Oxygen Electrode N. LUND, B. CARDELL, B.-M. TORNELL and S. ODMAN

Departments of Anesthesiology and Biomedical Engineering and the Unit for Hospital Hygiene, University Hospital, Linkoping, Sweden

. A method for disinfection and aseptic assembly of the MDO (Mehrdraht Dortmund Oberflache) oxygen electrode has been evaluated. The method is based on treating each electrode component separately and then assembling the different parts under aseptic surgical conditions. The pcrformance of the disinfected electrode was studied and found to be unchanged ascompared with a non-disinfected electrode. Bacteriological studies on the effectiveness of the disinfection technique described and on the permeability of the electrode membranes to bacteria and bacterial spores were performed. The electrode membranes were penetrated by bacteria in experiments simulating in-use conditions in about 11 % of membranes tested, after contamination of the inner electrode surface with heavy inocula. When studies were performed with the disinfected MDO oxygen electrode on humans, routine cultures from the wound were positive in about 10 W of the cases. No clinical wound infection was seen. The majority of the positive cultures were probably caused by the surgical technique rather than the handling of the electrode. The disinfection method for the MDOoxygen electrode described in this paper makes it possible to use the electrode clinically, except in cases of hepatitis.

Received 2 March, accepted f o r publication 5 April 1979

Tissue oxygen pressure is the end result of respiration, circulation and metabolism. It would therefore be of great interest to follow this parameter in severely ill patients, The MDU (Mehrdraht Dortmund Ober- flache, Prof. M. Kessler) oxygen electrode ( GLEICHMANN & LUBBERS 1960, KESSLER & GRUNEWALD 1969) (Fig. l ) , is a very suitable instrument for this purpose. However, up till now there has been no known tech- nique for disinfection of this electrode without altering the electrode performance.

The purpose of this study was to find a suitable disinfection method which did not affect the character- istics of the electrode, thus permitting the use of the oxygen electrode in humans.

THE MDO ELECTRODE The electrode consists basically of glass (Thermometer glass 16111, Schott, Mainz, Western Germany) with eight platinum wires sealed airtight (LUBBERS et al. 1969) (Fig. l) , Over this is a plastic cap, within which are enclosed the soldering points of the platinum wires and the anode connection to the outer, teflon covered wires leading to the electronic equipment. This unit is then

covered at the recording surface with an inner cello- phane (Cuprophan, Enka AG, Wuppertal, Western Germany) membrane and an outer PFTE (HN 1261, Indupack ApS, Glostrup, Denmark) (“teflon”) mem- brane, both 12 pm thick. Also covered by the mem- branes are the Ag/AgCl-anode, the electrolyte solution (0.2 M potassium chloride) and a lucite ring holding the cellophane. A rubber ring holds the teflon membrane in place.

When non-sterile electrodes with platinum wires are polarized, the electrode current shows four character- istic phases (Fig. 2). Within seconds after applying the voltage the current decreases from its initial value. Next there is a slow polarization phase, which lasts from 2 to 4 h. The third phase is a 5- to 10-h period of stability. Finally, the current increases and the electrode becomes unstable (KESSLER 1973).

DISINFECTION METHODS First, radiation, autoclaving and gas sterilization will be discussed (a), and then the effects of chemical agents on the polarization phases of the electrode will be eluci- dated (b).

0001 -51 72/80/040265-07 $02.50/0 @ 1980 The Scandinavian Society of Anaesthesiologists

266 N. LUND ET AL

I (nA) f

5 c \

Fig. 1. The figure shows the MDO oxygen electrode and its compo- nents. 1. Rubber ring to hold the teflon membrane, 2. Teflon membrane, 3. Lucite ring to hold the cellophane membrane, 4. Cellophane membrane, 5. Ag/AgCl-anode, 6. Electrolyte solution (0.2 M KCI), 7. Glass, 8. Eight platinum wires.

(a) Discussion Radiation. To kill pathogenic bacteria, fungi and molds, the complete electrode must be subjected to a radiation dosage within the range of0.05-2.1 Mrad (STOLTZ 1972). However, radiation affects the mechanical properties of teflon. First elongation, and, second, tensile and shear strengths are affected. With a dosage of lo5 rad, elongation is reduced to half the initial value, and at lo6 rad tensile and shear strengths are only 50 % of the initial values (VOORDE & RESTAT 1972). These effects eliminate radiation as a useful method. Autocluving. This method is not useful with the complete electrode for obvious reasons: the plastic components will melt and the electrolyte solution will evaporate.

hours

Fig. 2. Polarization phases of a non-sterile MDO electrode. The electrode was placed in saline 200 mV (KESSLER, M. (1973) Repro- duced by kind permission of the author and publishers).

0 seconds 5 5 10 15 20

Gas sterilization. Sterilization with ethylene dioxide is theoretically possible, but it would take from 24 h up to about 1 week before all gas had left the electrode (JO- HANSSON 1976). This is too long a time-delay for using the electrode in acute situations. The electrode would no longer be in the stability phase and the electrolyte solution would evaporate during this period.

(b) The ejjficts of some chemical agents on electrode performance Several fluids were tested: hydrogen peroxide, CidexB (glutaraldehyde, sodium bicarbonate and anti-corrosive agents), DiluformB (alcohol, formaldehyde, sodium nitrate, sodium bicarbonate and water) and 70% etha- nol. Halogenated compounds interfere with the teflon membrane (JOHANSSON 1976), and therefore such com- pounds were not considered in these studies.

The complete electrode was exposed to these agents for varying times (15 s - 10 min) and thereafter washed with sterile saline.

Hydrogen peroxide. Hydrogen peroxide was found useful for disinfection of catheter Pop electrodes (CHARLTON et al. 1963). However, comparison of the polarization curves before and after exposure to hydrogen peroxide showed a large increase in electrode current and a pronounced drift. The electrode returned to normal current levels after 2.5 h (Fig. 3). Microscopy of the electrode surface showed crystalline rings around the platinum wires. Hydrogen peroxide probably affects the platinum wires, the reference electrode and/or the electrolyte solution. Hydrogen peroxide may well dis- turb the normal electrode reaction in which hydrogen peroxide is normally produced in small amounts (LONC-

DISINFECTION OF MDO OXYGEN ELECTRODE

I (nA)

t

‘ I

267

I 15 30 45 60 rn1nut.s

Fig. 3. The figure shows the polarization ofan MDO electrode placed in saline saturated with 21 A oxygen (dotted line.) The continuous line shows the erect of treatment with hydrogen peroxide for 15 s. The current was unstable and unacceptably high.

I (nA)

20

15

10

5

15 30 45 60 rnlnutar

Fig. 4. The figure shows the polarization of an M D O electrode placed in saline sarurated with 21 A oxygen (dotted line). After exposure to Cidex@ for 10 min the electrode initially showed normal polarization and normal reactions in pure nitrogen. However, 30 min later, the electrode showed a slow upward drift.

MUIR 1964). The conclusion is that this agent is not useful.

Cide@,. The electrode showed a slow, upward, non- linear drift (Fig. 4), which means that this agent impairs the electrode function.

L h l u f o f l . No changes ofzero current, drift or sensitivity changes to oxygen were found. However, the response curve showed marked overshoots (Fig. 5). This change

Fig. 5. The figure shows the responses of an M D O electrode (dotted line) when exposed to pure nitrogen initially, thereafter to 21 96 ox) sen and finally pure nitrogen. After treatment with DiluformB (continuous line) the electrode showed marked overshoots.

I (nA)

20

15

10

5

60 rn1nut.r

Fig. 6. The figure shows the polarization of an MDO electrode placed in saline saturated with 21 A oxygen. The dotted line shows the normal behaviour of a non-sterile electrode, and the continuous line shows the reactions of an electrode treated with ethanol for 10 min. Initially the electrode was very unstable when polarizing in 21 %oxygen in saline. This instability probably depended on leakage currents. The electrode was later unstable when calibrated in 21 A oxygen, but stable in pure nitrogen.

15 30 45

was permanent and probably depended on a change of the Ag/AgCl-anode, in that the chloride left the silver surface. It may also be that the membranes were changed in such a way that the diffusion constant was altered. Thus, the agent is not useful for disinfection in this case.

Ethanol 70 %. The electrode showed very unstable reac- tions when calibrated in oxygen, but it wasstable in pure nitrogen (Fig. 6) . A change in sensitivity was also found. Although the electrode responded normally towards nitrogen, ethanol is not useful as the reactions in oxygen were unstable.

268 N. LUND ET AL.

by autoclaving. The separate components of the elec- trode were then assembled under normal aseptic surgi- cal conditions.

After the treatment described, the characteristics of the electrode were not altered, as shown in Figure 7. The reactions in both oxygen and in pure nitrogen were unchanged compared with the behaviour of a non- disinfected electrode. The electrode should be used within 6-8 h after it has been assembled.

Table 1 Methods for disinfection of the M D O electrode parts.

Material Disinfection method

Rubber ring Teflon membrane Lucite ring Cellophane membrane

Basic electrode unit (glass nucleus with platinum wires) Electric wires and contact Potassium chloride

Autoclaving Autoclaving Ethanol 70% or autoclaving Radiation (2.5 Mrad) or auto- claving Ethanol 70% for 5 min and drying at 70°C in an oven for 10 min

Ethanol 70% Autoclaving

A suitable method f o r disinfection of the MDO electrode Our solution to the disinfection problem described above is to treat each electrode component separately, and then assemble and calibrate the electrode under aseptic conditions. The components were treated as shown in Table 1.

The components treated with alcohol were dipped into 70% ethanol for 5 min, wiped clean with sterile compresses and then dried for 10 min at 7OoC in an oven.

As some components are made of rubber, teflon and cellophane, the autoclaving was done at a temperature of 120°C for 21 min. With this procedure these materials did not shrink and retained their elasticity. The elec- trolyte solution was made non-sterile and then treated

15 30 45 60 mlnutar

Fig, 7. The figure shows the polarization ofan MDO electrode placed in saline saturated with 21% oxygen. The continuous line shows the reactions of a n electrode disinfected with our method. The reactions in both oxygen and in pure nitrogen were unchanged as compared with the behaviour of a non-sterile electrode (dotted line).

BACTERIOLOGICAL STUDIES Materials The following media were used: glucose broth (GB), glucose-bromthymol blue broth (GBT), blood agar and spore agar (SA) (recipes for the different media are obtainable from the authors). As test organisms for inoculation on assays, strains of Staphylococcus epider- midis (GH 37), Escherichia coli (HA 2) and Pseu- domonas aeruginosa (CB 26) isolated from clinical specimens were used. A spore preparation from Bacillus stearothermophilus was kindly provided by docent Ingmar Juhlin, Malmo.

Methods and Findings Disinfection of the basic electrode unit with 70 % ethanol. The basic electrode unit was inoculated on the recording surface with a suspension of Staphylococcus epidermidis in GB and an average inoculum of 5 X 1 O8 cfu/electrode surface as judged from results with untreated controls. The suspension was allowed to dry in air for 2 h. Thereafter the basic electrode was dipped into 70 % ethanol for 5 min. After drying, it was then shaken in 0.5 ml of sterile saline at 37OC for 2 h. After these procedures, bacteriological growth tests with samples from the saline and the recording surface of the electrode were carried out as follows: 0.1 ml samples were taken from the saline for bacterial counts on human blood agar incubated aerobically for 24 h at 37" C. The lower limit for detection of organisms in each sample thus was 10/ml. In addition, the electrode surface was firmly pressed to the surface ofblood agar which was incubated in the same way.

Findings. No bacterial growth was found in any of the cultures either from the saline or the electrode recording surface. Control tests performed with cultures taken from untreated basic electrode units all showed bacterial growth.

Studies on the permeability o f cellophane and teflon. To define the permeability to bacteria of the two membranes, a

DISINFECTION OF MDO OXYGEN ELECTRODE 269

suspension of Escherichia coli in GB with an approxi- mate concentration of lo9 cfu/ml was used. Plastic tubes were covered at one end with either cellophane or teflon held in place by a rubber ring. The tube was then hung in a larger tube with GBT broth so that a close connection between the membrane surface and the broth was achieved. Then 0.3 ml of the bacterial suspension was applied inside the smaller tube, and was thus separated from the broth only by the membrane.

The test surface ofeach membrane was 12.5 mm'. The total area of teflon in 863 tests was approximately 1 dm'. The recording surface area of each MDO electrode is approximately 7 mm . 2

FzndirzgJ. When testing the permeability of the cello- phane membrane, positive bacterial growth was found in 55 out of 58 broths. When testing the permeability of the teflon membrane, however, we found 97 positive in 863 tests, i. e. bacteria penetrated the teflon membrane in 11.2% of the tests. We also found that, in the positive cases, most were positive after 30 min. Very few further positive results were observed after prolonged exposition times up to 7% h.

Permeability tests on the complete electrode with bacteria and bacterial spores (1) sPO7'8S. The basic electrode unit was inoculated with 10 pl of suspension of spores from Bacillus stearothermo- philus in sterile HnO. The suspension was allowed to dry in air for 2 h. Controls showed approximately 2X lo3 spores/electrode. The basic electrode unit was then built with sterile electrolyte solution and sterile membranes. The complete electrode was thereafter placed in 0.5 ml of sterile saline at 37°C for 2 h. Growth tests were performed with 0.1 ml samples taken from the saline for counts on SA, incubated as above but at 60°C. In addition, the teflon membrane was cultured on spore agar at 60°C. In 48 tests the following results were obtained: (a) spore growth was found in two of 48 saline solutions; (b) in one case, spore growth was found on a teflon surface. (2) Bacteria. Staphylococcus epidermidis and Pseudo- monas aeruginosa, respectively, were used for inocula- tion in the same way but with inocula about 5 X 10 cfu/electrode, as shown with untreated controls. These procedures gave the following results from a total of 40 electrode tests: (a) Bacterial growth was found in two of 40 saline solutions. (b) In 3 cases, bacterial growth was found on teflon surfaces: In each experimental proce- dure, controls were carried out without covering the basic electrode unit with electrolyte solution and mem- branes. All controls showed growth of bacteria.

HUMAN STUDIES Material. In a limited number of healthy volunteers (n= 10) and patients (n=20), the muscle surface oxygen pressure was studied with the MDO oxygen electrode disinfected as described above.

Method. Askin incision was made over the brachio-radial muscle to expose the muscle surface. This was done with normal surgical aseptic technique. Bacteriological swab tests were taken from the skin incision before the study but after the surgical preparation of the skin, and from the skin incision and the muscle surface after the study. The average exposure time, i. e. from making the skin incision until the wound was closed, was 137 min (range 95-180 min). Routine cultures were performed without any close quantitation. The results were roughly esti- mated as: no growth, moderate growthor heavy growth.

Findings. (a) Among the healthy volunteers, moderate growth was seen in one case with the same type of bacteria as cultivated from the skin incision both before and after the study. In another case moderate growth was observed with the same bacteria as from the skin incision and the muscle surface after the study. (b) In the 20 patients studied, three showed moderate growth in cultures taken from the skin incision before the studies. Of these, two showed moderate growth from the skin incision afterwards and one showed moderate growth also from the muscle surface.

One patient showed moderate growth from the skin incision only before the study, and another one showed moderate growth only from the muscle surface after the study. No case showed any clinically visible wound infections, and all wounds healed normally.

DISCUSSION T o disinfect equipment for use in surgery, many meth- ods and agents are available. A simple method is always preferred to a more complicated one if that is possible. In this particular case only a few methods could be used because of the complex nature of the M D O oxygen elect rode.

Our studies have led to the conclusion that, in order not to affect the electrode characteristics, each compo- nent of the electrode must be sterilized or disinfected individually in accordance with the properties of each part. This also means that the components must be assembled under careful, surgically aseptic conditions immediately before use. Although this is a relatively complicated procedure with a certain risk of infection, no other useful alternative has been found.

270 N. LUND ET AL.

The treatment described above did not affect the electrode characteristics, as seen in an unchanged polarization curve and normal electrode reactions in oxygen and pure nitrogen.

In order to control the bacteriological effectiveness of our method of treating and assembling the electrode parts, laboratory tests with heavily contaminated equip- ment, but otherwise simulating in-use conditions, as well as in uivo studies on patients were carried out. 70% Ethanol was reasonably effective against bacteria, with reference to the surface and design of the basic electrode unit. Under similar conditions the permeability of the membranes to various bacteria and bacterial spores was evaluated. As expected, a rather high frequency of bacterial penetration of the cellophane membrane (55 out of 58 tests) was noted. The cellophane membrane was identical to that used in many slate dialysers.

The teflon membrane was also found to be permeable to bacterial spores to a certain degree. Bacteria pene- trated the teflon membrane in 11.2% of 863 electrode tests. MILLER et al. (1972) studied the permeability of surgical rubber gloves and found that the permeability varied from 24.1 % to 61.1 % in different situations. These results illustrate the risk of perforation of surgical gloves and contamination from the hands of the surgeons. The electrodes are always ground before use to remove organic material. A teflon membrane permeability of 11.2% thus seems acceptably low, particularly with regard to the high inoculation doses used in these studies in contrast to in-use conditions.

The study performed on humans showed that bacteria could be cultivated with conventional methods in slightly more than 10% of cases. An important point is that most of the positive findings were with the same bacteria as those found on the skin before the study. This seems to indicate that the actual infection rate depends not only on the electrode treatment, but also on the disinfection of the operative area and, perhaps, also the wound exposure time.

A special problem is viruses, especially hepatitis virus. To destroy these viruses it has been suggested that the equipment should be treated e.g. in boiling water for more than 5 min, in a heat sterilizer (160O-180°C), by autoclaving, or with halogenated compounds (SYKES 1965, KRUCMAN et al. 1979). It is not possible to subject the MDO electrode to such treatment.

Our conclusion is thus that the MDO electrode is a useful instrument for human studies, and that it can be used with an acceptably low risk of infection if the following rules are observed strictly:

(1) The method described for treating and assembling the electrode parts must be followed strictly.

(2) A meticulously aseptic surgical technique must be used when preparing the skin and the muscle surface.

(3) Patients with suspected hepatitis infection or posi- tive HB,-antigen tests must not be studied with an electrode which is later to be used in non-hepatitis cases.

We are convinced that the usefulness of the MDO oxygen electrode outweighs the risks of infection, pro- vided that the above-stated rules are followed.

ACKNOWLEDGEMENTS Our sincere thanks to Professor M. Kessler, Erlangen, for supplying us with the MDO oxygen electrodes, and to Asst. Prof. Birgitta Nilthn and Ann-Margret Hallert for valuable advice without which this study would not have been possible.

This study was supported by grants from Trygg-Hansa, and Forenade Liv Mutual Group Life Insurance Company, Stockholm, Sweden, the Swedish Medical Research Council (project Nr 02042), the Swedish Defence Research Council (H564) and the Medical Faculty, Linkoping University.

Enka AG, Wuppertal, Western Germany, kindly supplied us with Cuprophan free of charge.

REFERENCES

CHARLTON, G., READ, D. & READ, J. (1963) Continuous intra-arterial PQ in normal man using a flexible microelectrode. J . appl. Physiol. 18, 1247.

GLEICHMANN, U. & LUBBERS, D. W. (1960) Die Messung des Sauer- stoffdruckes in Gasen und Flussigkeiten mit der Pt-elektrode unter besonderer Beriicksichtigung der Messung im Blut. PJugers Arch. ges. Physiol. 271, 431.

JOHANSSON, G. (1976) Personal communication. KESSLER, M. (1973) Problems with the use of platinum cathodes for

the polarographic measurement of oxygen. Oxygen Supply. Theoretical and Practical Aspects of Oxygen Supply and Microcirculation of Tissue, ed. KESSLER, M., BRULEY, D. F., CLARK, JR., L. C., LUBBERS, D. W., SILVER, I. A. & STRAUSS, J. Urban & Schwarzenberg, Miinchen, Berlin, Wien, pp. 81-85.

KESSLER, M & GR~NEWALD, W. (1969) Possibilities of measuring oxygen pressure fields in tissue by multiwire platinum electrodes. Progr. resp. Res. 3, 147.

KRUGMAN, S., OVERGY, L. R., MUSHAHWAR, I. K., LING, C.-M., FR~SNER, G. G. & DEINHARDT, F. (1979) Viral hepatitis, type B. Studies on natural history and prevention re-examined. New Engl. 3. Med. 300, 105.

LONGMUIR, I. S. (1964) The oxygen electrode. Oxygm in the Animal Organism, ed. DICK~NS, F. & NEIL, E. Pergamon Press, Oxford,

LUBBERS, D. W., BAUMGARTL, H., FABEL, H., HUCH, A., KESSLER, M., KUNZE, K., RIEMANN, H., SEILER, D. & SCHUCHHARDT, S. (1969) Principle of construction and application of various platinum electrodes. Progr. resp. Res. 3 , 136.

MILLER, J. M., COLLIER, C. S. & GRIFFITH, N. M. (1972) Permeability of surgical rubber gloves. Amer. 3. Surg. 124, 57.

pp. 219-237.

DISINFECTION OF MDO OXYGEN ELECTRODE

STOLTZ, W. (1972) Strahlenrterilisation. Grundlagen und Anwendung in Medizin und Phrmacie. J. A. Barth, Leipzig, pp. 90.

SYKES, G. (1965) Disinfectionand Sterilization. Theory and Practice, 2nd Ed. Chapman & Hall, London.

VOORDE, VANDE, M. H. & RESTAT, C. (1972) Selection Guide to Organic Materials for Nuclear Engineering. CERN 72-1, Laboratory I, Inter- secting Storage Rings Division, Geneva.

Address: Niels Lund, M.D. Dept. of Anesthesiology University Hospital S-581 85 Linkoping Sweden

271