Topical Humidified Carbon Dioxide to Keep the Open Surgical Wound Warm: The Greenhouse Effect Revisited

The surface temperature in an open surgical wound model was measured at steady state during five different conditions: without gas insufflation (control) and during insufflation of dry and humidified air and carbon dioxide, respectively. We also studied the heating effect of the operation lamps.

The wound model (fig. 1) was placed on the operating table of a fully ventilated operating room (laminar airflow of approximately 2500 m³/h from the ceiling downward). The wound model consisted of a heat-regulated water container (Tempette TE-8A, 50 × 30 × 20 cm³(30 l), 1000W; Techne, Princeton, NJ) with a central elliptical opening in its top cover. To represent a wound cavity we positioned a removable container in the opening, a length, width, and depth of 20, 12, and 5 cm, respectively. The wound cavity was made of 1-mm thick aluminum to maximize heat conduction. During the experiment the cavity was kept surrounded by water. The water temperature was kept at 37°C, controlled with a penetration probe (FK 1020, Pt-100; Heraeus Sensor-Nite GmbH, Kleinostheim, Germany) to represent body core temperature. To simulate the surface of fresh wound tissue, we positioned two intact layers of blood agar directly at the bottom of the model (fig. 1). Each blood agar layer was taken out in one piece from a standard blood agar plate (9-cm Petri dishes, 5% horse blood). A thin-film temperature sensor (M-FK 1020, Pt-100, 18 mm²; Heraeus Sensor-Nite GmbH) was positioned on the center of each agar layer to measure the surface temperature. The temperature sensors were connected to a digital temperature instrument (2802A Thermometer, 34740A Display; Hewlett Packard, Palo Alto, CA), which has a resolution of 0.01°C. For optimal accuracy the instrument was calibrated and each temperature sensor was connected to the instrument via  a 4-wire connection, according to the manual of the manufacturer.

Gas was insufflated into the wound model via  a low-velocity outlet device, a gas diffuser (Cardia Innovation AB, Stockholm, Sweden) that consists of a thin fixable tube with a diffuser of polyurethane foam at the end.11,12The insufflation device was positioned at the acute end of the wound cavity model, with the diffuser positioned at half the depth of the cavity. Medical air or carbon dioxide was delivered at a flow of 5 l/min, controlled with back-pressure compensated flowmeters (AGA Gas AB, Stockholm, Sweden). The flowmeters and the used calibration procedure have earlier been described in detail.11,12Air was taken from the central anesthetic gas system in the operating room, and carbon dioxide was delivered from a pressurized gas cylinder (AGA Gas AB). The gases were humidified via  a disposable bubble humidifier containing sterile water (Aquapak 340 ml; Hudson Respiratory Care Inc., Temecula, California). A pilot study showed that for the humidity of the administered gas to remain constant, the humidifier had to be kept at a constant temperature. Thus, during the experiment the humidifier was kept at room temperature in another heat-regulated water bath (Haake D8/L; Haake Mess-Technik GmbH u. Co., Karlsruhe, Germany). We did not exceed room temperature to prevent condensation in the gas delivery system, which might have interfered with our measurements. Two operation lamps (Angénieux, Type AX 14, 300W; ALM SA, Orléans, France) were focused towards the wound cavity at a distance of 1 meter. The lamps were positioned at opposite sides of the wound model at an angle of 45° to minimize disturbance of the operating room ventilation.

First, while the operation lamps were kept switched off, temperature and humidity in the air above the operating table were measured with a digital hygrometer (HygroPalm 3; Rotronic AG, Bassersdorf, Switzerland). Thereafter, the surface temperatures were recorded at steady state, i.e. , when values fluctuated around a constant value over 30 s. The operation lamps were then switched on, whereupon the temperatures were followed and again measured at steady state.

The evaporation rate of the wound cavity model was then studied during the above described five conditions with the operation lamps switched on. The wound cavity was first closed with a plastic lid, which had been preheated to more than 37°C to avoid condensation. The cavity was then removed from the water container. Thereafter, the cavity was completely dried with a clean and dry tissue and weighed on a precision scale (Sartorius L420S, resolution 0.001g; Sartorius Gmbh, Göttingen, Germany). The cavity was then put back into the water container, the lid was removed, and gas insufflation was started (if used). Ten min later, the cavity was closed, dried, and weighed again, as described above. The evaporation rate could thus be expressed as water loss in mg per cm²of agar surface and minute.

A fresh pair of blood agar layers was used for every new measurement. The five conditions were repeated 10 times in random order, resulting in 50 randomized measurements.

Surface temperatures and evaporation rates in the wound model are presented as medians with 25th and 75th percentiles (n = 10), whereas room temperature and room humidity are presented as means and standard deviations (n = 50). The average surface temperature of both agar layers combined was used for comparison of surface temperatures between the five groups. The nonparametric Mann–Whitney U  test and Wilcoxon test were used when appropriate. Differences were considered statistically significant if P < 0.05.

During the experiment the temperature was 19.4 ± 0.2°C and the relative humidity was 21.6 ± 2.6% (mean ± SD) in the operating room. When the operation lamps were switched off, the surface temperature in the wound cavity in the control group was 31.8°C (fig. 2A). Insufflation of dry and humidified air did not have any statistically significant effect on the surface temperature. With dry carbon dioxide, the surface temperature was higher in comparison with the control and the air insufflation groups (P < 0.001). With humidified carbon dioxide, the surface temperature (33.7°C) was even higher than with dry carbon dioxide (P < 0.001).

When the operation lamps were switched on, the surface temperature increased in all five groups (P < 0.001, fig. 2B). The increase in surface temperature was greater both with dry (P = 0.02) and humidified (P = 0.007) carbon dioxide in comparison with the control group. The corresponding increases for dry and humidified air were not statistically significantly different from the control. With the lamps switched on, the surface temperature in the control group was 33.3°C. In comparison with the control, the surface temperature was slightly higher both with dry (P = 0.03) and humidified (P = 0.05) air. With dry carbon dioxide, the surface temperature was higher (35.2°C) than in the control and air insufflation groups (P < 0.001). With humidified carbon dioxide, the surface temperature was significantly higher (35.6°C) than with dry carbon dioxide (P = 0.002).

With the lamps switched on, the evaporation rate did not differ between the control and when dry and humidified air was insufflated (fig. 3). With dry carbon dioxide, the evaporation rate was almost 40% lower than the control (P < 0.001). When the carbon dioxide was humidified, the evaporation rate was even lower (P < 0.001): it was approximately 60% lower than the control group.

Our major finding is that the creation of a local carbon dioxide atmosphere reduced heat loss from an open surgical wound model. The warming effect was enhanced by the radiative heat from operation lamps and by humidifying the insufflated carbon dioxide.

When a new technique is to be tested, the logical first step is to study its effect in a controlled experimental setting, while keeping in mind that, as pointed out by Sessler,3“The contribution of evaporation [to heat loss] from within surgical incisions remains to be determined in humans because of technical difficulties. An additional difficulty is that surgical incisions presumably also substantially increase loss by radiation.”

As part of our efforts to reproduce the conditions of clinical practice, the experiment was carried out on an operating table in a fully ventilated operating room. The wound cavity model11,14was large enough to hold two standard 9-cm blood agar layers. Blood agar provides a wet surface similar to a fresh surgical wound. An open water surface has earlier been used to represent evaporation from an open wound15on the grounds that “a wound after full-thickness excision of skin transmits water vapor at a rate equal to 91% of that for an open water surface.”16In control experiments, we compared the water loss from Petri dishes containing blood agar and water, respectively, and found that the evaporation rates were equal. In contrast to water, in which circulation may occur, blood agar provides a solid but constantly wet surface on which a surface temperature can be measured. The wound model included a core thermal compartment, the water container, which provided a realistic core-to-wound temperature gradient.

Carbon dioxide was insufflated at a constant flow of 5 l/min with the gas diffuser positioned below the brim to create a complete carbon dioxide atmosphere inside the cavity.11,12The gas diffuser was positioned at the acute end of the cavity, which is a suitable position during surgery.10By measuring the surface temperature at two different locations, close to and further away from the insufflation device (fig. 1), we obtained a reliable estimate of the average surface temperature in the wound cavity model during gas insufflation.

In the current study, the average surface temperature in the wound model in the control group with and without the operation lamps on was 3.7°C and 5.2°C lower than the core temperature (37°C) in the water container, respectively (fig. 2). It is worthy of note that Lamke et al.  17found that the surface temperature of bowels fell by 3–5°C during open abdominal surgery. The current study showed that the operation lamps had a heating effect, which increased the surface temperature in all studied groups. During the experiment the two operation lamps were positioned at approximately 1-m distance with the light directed towards the wound model at an angle of approximately 45°. In pilot studies we found that the surface temperature could be increased further by shortening the lamp distance from the wound model and by decreasing the light angle.

The heating effect of the lamps was even greater when an insulating carbon dioxide atmosphere was present in the cavity. Why did not air insufflation have the same effect? First, carbon dioxide may cause a greenhouse effect, i.e. , absorbing and reemitting long wave energy to the surface of the wound model. Second, as it is heavier than air, carbon dioxide gravitates in the wound model and covers it like a protective layer. Thus, carbon dioxide reduces convection in the wound model caused by air currents from the operating room ventilation. This may explain why insufflation of completely dry carbon dioxide resulted in a substantially lower evaporation rate than the control and even humidified air.

Humidification of carbon dioxide had a marginal but statistically significant effect on the surface temperature in the wound model. Why was this effect not greater? As indicated by figure 3, although considerably reduced, evaporation from the wound model was still present when humidified carbon dioxide was used. When the insufflated gas, humidified at room temperature, entered the wound model, it was heated up at the wound model surface. Because warm gas can carry more water than cold, the relative humidity of the insufflated gas decreased as a result.

If evaporation could be eliminated the remaining heat loss from the wound model would be mainly radiant. This is where the operation lamps come in. In combination with a local carbon dioxide atmosphere, the radiant heat from the lamps may not only compensate the heat loss from the wound model but also cause a net influx of heat if the surface temperature is higher than the core temperature.

The current study may have several clinical implications. Hippocrates’ statement that “Wounds love warmth”18still holds true today, as it is current practice to prevent perioperative and postoperative hypothermia to reduce the incidence of surgical complications.3,4,6,19–22A humidified carbon dioxide atmosphere may provide an effective thermal insulation of an open surgical wound and the light emitted from the operation lamps may be used for heating. The two factors should be combined because mere heating with lamps will most likely increase the evaporation and desiccation in the wound. This method to warm an open wound does not substantially interfere with surgery.10,13 

An increased wound temperature may promote wound healing by increasing tissue oxygenation via  locally mediated vasodilatation and increased oxygen deposition in tissue.21,23Accordingly, increased arterial and subcutaneous oxygen tension is found to significantly reduce the incidence of surgical wound infection.7Furthermore, oxygen tension in subcutaneous wound tissue is found to be a strong predictor for the risk of surgical wound infection.23The presence of a carbon dioxide atmosphere in the wound will theoretically provide synergistic effects, as topically applied carbon dioxide alone causes local vasodilatation as well as augmented tissue oxygenation by the Bohr effect,24,25 i.e. , a rightward shift of the oxygen saturation curve.

The combination of humidified carbon dioxide and operation light may also affect body core temperature. Perioperative heat loss mainly occurs via  the anterior body surface26and rapidly reduces core temperature.27A substantial amount of heat is lost through the open surgical wound via  radiation and evaporation.3,5Moreover, it is well known that core hypothermia is more pronounced during large operations than during small operations,28,29with most of the difference presumably resulting from an increased evaporation.3Because anesthesia causes vasodilatation, heat is more easily redistributed from the body core out to the wound surface and vice versa .19,27The lower the surface temperature, the greater is the core-to-peripheral temperature gradient and thus the rate of heat redistribution and heat loss.19Consequently, if the wound surface temperature can be kept warm, the core temperature may also be kept up, decreasing the risk of wound infection. Kurz et al.  4found that a decrease in perioperative core temperature of only 1.9°C (from 36.6°C to 34.7°C) tripled the incidence of surgical wound infection and prolonged hospitalization after colon resection. Maintaining normothermia has other beneficial effects, such as less blood loss,30decreased incidence of morbid cardiac events,31and improved immune functions.20,32,33Clinical trials are needed to evaluate the clinical effects of intraoperative insufflation of humidified carbon dioxide.