Preliminary ex vivo experiment
Vacuum was achieved upon first application in all cadaveric limbs, as confirmed by the green light and wrinkling of the dressing pad occurring within 30s. The vacuum was successfully maintained on each limb for 24 h.
Preliminary in vivo experiment
Vacuum was achieved upon first application in both mares, as confirmed by the green light and wrinkling of the dressing pad occurring within 30s. The vacuum was successfully maintained, with the mares under observation, for a pilot period of one hour. The mares showed no signs of pain or discomfort (as measured by increased heart and respiratory rates, and observation of behavioral patterns). Voluntary movement and manipulation of the mares did not preclude maintenance of vacuum during the experiment.
Experimental wound study
Since no difficulties were encountered with the application or maintenance of the PICO® system during the ex vivo experiment or during the preliminary experiment conducted on intact skin in live horses, it was decided to proceed with the experimental wound study.
Two groups of two mares started the experiment at a 24 h interval from one another. A slight erythema and subcutaneous swelling was noted on the thoracic wall of one mare following application of the hair removal cream prior to wounding. This complication resolved in 48 h and was not observed in the other mares. Although no significant hemorrhage occurred during the first 24 h after wounding, subcutaneous edema developed around thoracic wounds, particularly along the distal border where tunneling was performed.
Difficulties were encountered during the first application of the PICO® dressing, at both anatomic locations, in the first group of two mares. In spite of a thorough initial cleansing with Chlorhexidine gluconate 4% soap (HibiScrub, Mölnlycke Health Care, Oakville, Canada) followed by careful drying of the skin surrounding the wound margins, the silicone adhesive layer of the PICO® dressing failed to adhere to the skin. The addition of an adhesive polyurethane film (OPSITE®, Smith & Nephew) over the dressing’s edges did not improve adhesion as this film also failed to stick to the skin (Fig. 1). As a consequence of the ineffective seal, the vacuum was not achieved at this stage of the experiment. In an effort to preclude the possibly negative effect of the subcutaneous edema having developed around the thoracic wounds, the PICO® dressing was applied immediately following wound creation, at both anatomic locations, in the second group of two horses. In this case, an effective vacuum was achieved in the thoracic area but successfully maintained for only 15 to 20 min. A satisfactory vacuum was not obtained on the limb wounds in any of the four mares. The PICO® hardware was well tolerated by all the mares, for a 24 h period in their stalls.
In view of these unexpected difficulties, an additional set of tests was then undertaken on a separate, unwounded, mare to determine the best configuration to successfully apply the PICO® dressing. Various elements and conditions were tested in an effort to improve the dressing adhesion and vacuum pump efficacy. The initial skin preparation was revised since, during the experimental wound study, two mares were observed to sweat profusely following sedation, which may have reduced the adhesive properties of the PICO® dressing. Consequently, an antiperspirant spray (MEN REXONA Adventure®, Unilever Canada Inc., Toronto, Canada) was applied locally following skin preparation, though this did not improve adhesion of the dressing. A liquid dressing (SKIN-PREP®, Smith & Nephew) was also used to prepare the skin following cleansing and before dressing application, but this also failed to improve adhesion of the PICO® dressing to the horse’s skin. The adhesion of the dressing on the distal limb was tested with and without prior treatment with hair removal cream to assess the cream’s and the hair’s potential effect on the adhesive properties of the PICO® dressing. Although good dressing adhesion was obtained in both cases, the pump was unable to maintain an effective vacuum. At the suggestion of the manufacturer, the vertical depressions between the second and fourth metacarpal bone and the suspensory ligament were filled with 4 × 4 woven gauze to eliminate dead space, however, this did not help to achieve an effective vacuum.
Various adjunctive adhesives were also tested at the periphery of the thoracic wounds of the four mares in which application of the PICO® device failed: cyanoacrylate glue (Krazy Glue®, Krazy glue Inc., High Point, NC), spray adhesive (Mueller Tuffner Pre-tape spray, Mueller sport medicine Inc., Prairie du Sac WI), gel adhesive (RENASYS® Adhesive gel patch, Smith & Nephew) and silicone gel patches (CicaCare®, Smith & Nephew) all failed to improve adhesion sufficiently for vacuum to be maintained. Other elements were superimposed on the edges of the PICO® dressing: a spray dressing (OPSITE SPRAY®, Smith & Nephew), a liquid dressing (NewSkin Liquid bandage, Moberg Pharma North America LLC, Cedar Knolls, NJ) and an adhesive elastic band (TENSOPLAST®, BSN Medical Canada, Laval, Canada); these also failed to enhance adhesion of the PICO® dressing.
The application of a stoma paste (Stomahesive®, ConvaTec, Dorval, Canada) to the edges of the PICO® dressing generated variable results (Fig. 2). Adhesion of the dressing was sufficient to allow an effective and stable vacuum to be achieved and maintained for 24 to 48 h on the thoracic wounds of three mares and on the thoracic skin of the unwounded mare. The mares received a 7-day NPWT with frequent dressing change; indeed, focal liquefaction of the stoma paste was consistently observed 24 h after application and precluded vacuum maintenance, obliging dressing change 3 to 4 times a day. Moderate cutaneous irritations and subcutaneous edema were observed on all mares at the first dressing change (Fig. 3). In these mares, the PICO® NPWT system was well tolerated by three individuals for the week-long trial, whereas one individual required the installation of a shoulder protection (usually used to prevent blanket rubbing over the shoulder) to shield the vacuum pump. No behavioral patterns indicating pain or discomfort, nor changes in heart rate or respiratory rates associated with use of the device were observed during the daily clinical examination for the week-long trial.
A special dressing designed for complex anatomical areas (PICO Multi-site®, Smith & Nephew) was tested, with the stoma paste, on two distal forelimbs: one on a mare of the experimental wound group and one on the mare without wounds. Moreover, a breathable bandage consisting of cotton undercast padding (Webril®, Covidien-Medtronic), stretch bandage (Curity®, Covidien-Medtronic) and bandaging tape (VetRap®, 3 M Canada, London, Canada) was applied over the PICO® dressing. This design enabled the achievement and maintenance of a vacuum for 4 consecutive days. However, cutaneous irritations were noted in both cases after dressing removal at 4 days: the mare with experimental wounds had minor irritations whereas the unwounded mare developed severe cutaneous irritations associated with cellulitis (Fig. 4). These irritations were treated with daily hydrotherapy, bandages and a course of anti-inflammatory drug (phenylbutazone, 2.2 mg/Kg, BID) for 3 days.
Once the NPWT trial was abandoned, 2 weeks after the wound surgery, all wounds were left to heal by second intention (Additional file 1: Figure S1). Limb wounds took longer to heal than did body wounds: average time to full healing for limb wounds was 83 days (SD 2.58 days) and that for body wounds was 62.5 days (SD 2.52 days).