Materials and Methods

All procedures for the study were approved by the Kansas State University Animal Care and Use Committee.

Animals—Over a 37-month period, all horses (> 1 year of age) that were scheduled to be anesthetized for surgical procedures at the Kansas State University veterinary hospital were assigned via a random numbers table to recover from anesthesia on the floor of a foam-padded recovery stall (200 horses) or on a rapidly inflating-deflating air pillow in an identical stall (209 horses). There were no exclusion factors, and attending clinicians did not influence recovery method choice. The study population included emergency after-hours cases as well as the routine surgical caseload. No significant demographic differences were apparent between the horses that were allowed to recover on the air pillow versus those allowed to recover on the padded recovery stall floor. The mean ages of horses recovering on the pillow and padded recovery stall floor were 7.4 and 7.3 years, respectively; mean weights were 477 kg (1,049 lb) and 473 kg (1,041 lb), respectively. Of the 409 horses, 167 (41%) were mares, 168 (41%) were geldings, and 74 (18%) were sexually intact males. When grouped according to perceived disposition and use, 233 (57%) were Quarter Horses, American Paint Horses, and grade horses; the remaining horses included 99 (24%) Thoroughbreds and Trotters, 71 (17%) Arabians and Appaloosas, 4 (1%) draft horses, and 2 (0.5%) warmbloods.

During preoperative evaluation prior to anesthesia, an American Society of Anesthesiologists status score⁷ was assigned to each horse. Such scores ranged from I (healthy patient) to V (moribund patient unlikely to survive for 24 hours with or without surgery).

As part of an unpublished companion study, interventions and injuries that occurred during recovery from anesthesia were recorded for the same population of horses used in this study. Intervention was recorded as yes or no and defined as anyone entering the stall to assist in the recovery process. Recoveries from anesthesia without injury were recorded, and injuries that occurred during recovery were categorized as minor abrasions; cast or bandage repairs; lacerations of the tongue, lip, or eye; and myositis. A χ² analysis was used to compare injury categories and interventions between the air pillow and the padded recovery stall. A value of P ≤ 0.05 was considered significant for these data.

Anesthesia—Anesthesia protocols were selected by the attending anesthetist, who considered the physical status and medical needs of the horse. Typically, horses were premedicated with either acepromazine maleate (0.02 to 0.05 mg/kg [0.009 to 0.023 mg/lb]) or xylazine (0.2 to 0.4 mg/kg [0.09 to 0.18 mg/lb]) administered IV, and anesthesia was induced with guaifenesin (50 to 100 mg/kg [22.7 to 45.5 mg/lb], IV) to effect followed by a bolus of either thiamylal sodium (4.0 to 6.0 mg/kg [1.8 to 2.7 mg/lb], IV) or ketamine (2.0 to 2.2 mg/kg [0.9 to 1.0 mg/lb], IV). Each horse was intubated with an orotracheal tube, and anesthesia was maintained with either halothane or isoflurane delivered in 100% oxygen via a circle anesthesia breathing system. Oxygen flow was maintained at 5 to 8 L/min. At the discretion of the attending anesthetist, horses were either ventilated mechanically or allowed to breathe spontaneously. Supplemental tranquilizers, sedatives, or analgesics were not commonly administered prior to discontinuation of anesthesia and transfer of the horses to the recovery stall.

Padded recovery stall—The 2 recovery stalls used in the study were identically constructed and measured 4.27 × 4.8 m (14 × 16 feet). Both floors were covered with 3.81-cm (1.5-inch)-thick, high-density foam mats encased within a nonslick nylon-reinforced vinyl covering. The walls were covered with a polyurethane surface^a (1.3 cm [0.5 inch] thick).

Opposite sides of the stalls contained 2 doors (each 2.1 m [7 feet] wide) with small windows through which all recoveries from anesthesia were monitored visually by observers and recorded by video cameras. A shielded incandescent 60-W bulb was positioned 2.4 m (8 feet) above the floor in the stall corner opposite the cameras.

Air pillow—For air-pillow recoveries, a specially designed mattress-like inflatable cushion^b was positioned in the padded recovery stall. Fully inflated, the pillow was 45 cm (18 inches) deep and completely covered the stall floor (Figure 1). It was protected from cuts or abrasions by a cover that was held in place and tensioned by ropes passed through corner holes in the stall wall and secured around rope cleats. The pillow and cover were constructed of extraheavy vinyl material (0.75 kg/m² [22 oz/yd²]) that was strengthened with nylon and polyester fibers and reinforced with flat nylon webbing (5 cm [1.97 inch] wide). Internal vented vertical baffles, also made of vinyl, were inserted to maintain a relatively constant thickness when the pillow was inflated. Two 75-cm (30-inch) zippers permitted rapid deflation. Corner D rings on the pillow and cover allowed both to be hung up and cleaned between uses as needed.

View Full Size

Photographs of a specially designed rapidly inflating-deflating air pillow for use during recovery from anesthesia in horses. A—Fully inflated air pillow (outside a recovery stall). Inflation is achieved by use of the ducted fan. B—One of 2 zippers in the pillow, which permit rapid deflation of the pillow when opened.

Citation: Journal of the American Veterinary Medical Association 229, 5; 10.2460/javma.229.5.711

• Download Figure
• Download figure as PowerPoint slide

For pillow inflation, air was supplied through a flexible 25-cm (10-inch)-diameter spout that was attached to the air pillow on 1 end and extended through a 20 × 20-cm (8 × 8-inch) hole in the stall door to attach to a shrouded fan located outside the recovery stall. The spout and fan did not interfere with opening and closing the door. Appropriately sized fans are provided by the manufacturer of the pillow.^b The fans are high-volume, direct-drive blowers that produce a flow in excess of 1,500 cubic feet/min. When inflated, the internal pressure of the pillow is 6 cm H₂0. The pillow inflates in less than 30 seconds regardless of the size of the horse.

Procedure for recovery from anesthesia—Each horse was lifted off the surgery table with an electric chain hoist centered in the recovery stall ceiling and placed in lateral recumbency in the center of the stall. If the horse was shod, the shoes were left in place and feet were wrapped with elastic tape. A nasotracheal tube was secured in place unless surgery-associated hemorrhage raised the concern of aspiration, in which case the orotracheal tube was left in place and its cuff was inflated. A rubber tube (outer diameter, 1 cm [3/8 inch]) was attached to an oxygen flowmeter and inserted into the nasotracheal or orotracheal tube. Oxygen was passively insufflated at 15 L/min until the horse was standing.

Horses that were allocated to undergo conventional recovery from anesthesia were placed directly on the foammatted recovery stall floor. Horses that were allocated to undergo air-pillow recovery were lowered onto the deflated pillow, with tension from the top cover decreased by loosening the ropes (Figure 2). Both zippers on the pillow were closed and the shrouded fan was turned on to run continuously during the inflation period. This allowed the horses to sink deeply into the low-pressure, air-filled pillow. The anesthetist dictated when the pillow would be deflated, based on his or her clinical assessment of the horse's ability to stand. In general, after the horse made 3 to 4 aggressive attempts to attain sternal recumbency, both of the zippers were opened and the fan was turned off. The flexible spout was folded beneath the pillow, and a full-thickness plug was immediately and easily replaced in the hole in the door. In approximately 30 to 60 seconds, adequate pillow deflation occurred to allow the horse to bear full weight on the underlying pad. The corner ropes attached to the top cover were pulled taut and secured to their cleats to keep the cover evenly spread over the deflated pillow.

View Full Size

Photograph of an anesthetized horse positioned on the inflated air pillow for monitoring during recovery from anesthesia.

Citation: Journal of the American Veterinary Medical Association 229, 5; 10.2460/javma.229.5.711

• Download Figure
• Download figure as PowerPoint slide

All lights were dimmed after positioning the horses for recovery. Horses in both stalls were allowed to recover to the standing position without stimulation or assistance. Intervention occurred only in instances in which the well-being of the horse was at risk (eg, if the nasotracheal tube, orotracheal tube, or nostrils were partially occluded or compromised).

Data collection and analysis—Recoveries from anesthesia were videotaped for later analysis and scoring. The date, owner's name, and accession number had been recorded on the videotape prior to moving the horse into the allocated recovery stall. A 24-hour time and date generator^c ran continuously, and this chronographic information was visible on each recovery tape. Once the horse was placed within the recovery stall, times to first movement, first attempt to attain sternal recumbency, attainment of sternal recumbency, first attempt to stand, and successful standing were recorded. The number of attempts to achieve sternal recumbency and standing were counted, and a quality score for ability to attain a standing position was assigned to each horse (Appendix 1). Quality of the overall recovery was also scored by use of another system (Appendix 2). Interventions during the recovery phase, defined as entrance of personnel into the recovery stall to correct occlusion of the nasotracheal tube, orotracheal tube, or nostrils or address other life-threatening events, were recorded as well. Evaluation of videotapes was accomplished by a veterinary student who was experienced in observing horses during recovery from anesthesia and trained to use the scoring systems. The evaluator was not present during the actual recovery of the horses. Subsequent to completion of the study, the noise levels were measured with a sound meter^d; measurements were made of the background noise in the surgery room without the fan running and during operation of the fan, sound readings obtained at a position 5 feet from the fan and within the recovery stall.

Statistical analyses included t tests and Wilcoxon rank sum tests to compare measures of time (ie, times to first movement, first attempt to attain sternal recumbency, attainment of sternal recumbency, first attempt to stand, and successful standing) and the number of attempts to achieve sternal recumbency and standing between groups. χ² Analyses were used to compare the quality scores for ability to attain a standing position and overall recovery between groups. Significance was set at a value of P ≤ 0.05. Data are reported as mean values ± SD.

Results

In the air-pillow recovery group, the horses were uniformly supported in lateral recumbency by the inflated pillow, which cushioned any potentially violent movements and strongly inhibited their ability to roll into a sternal position. During use of the fan, there was considerable noise outside of the recovery stall; however, the sound was markedly muted inside the recovery stall. Without the fan running, background noise in the surgery room was 61 decibels (dB). During operation of the fan, sound readings obtained at a position 5 feet from the fan and within the recovery stall were 80 dB and 69 dB, respectively.

Movement and position—Mean ± SD times to first movement, first attempt to attain sternal recumbency, attainment of sternal recumbency, first attempt to stand, and successful standing were all significantly longer for the horses in the air-pillow recovery group than the horses in the conventional padded stall recovery group (Table 1). Horses recovering on the air pillow made significantly more attempts to roll upright before successfully achieving sternal recumbency than did horses recovering on the padded floor (number of attempts, 2.6 ± 1.6 and 1.9 ± 1.8, respectively). However, on pillow deflation, those horses were able to stand after significantly fewer attempts than horses recovering on the padded floor (number of attempts, 2.4 ± 0.9 and 3.2 ± 2.1, respectively). Of the 133 horses that required ≥ 3 attempts to attain sternal recum bency, 92 (70%) were in the air-pillow recovery group; of the 111 horses that stood on the first attempt, 71 (64%) were in the air-pillow recovery group.

Quality scores—On assessment of the videotaped recoveries from anesthesia, mean quality score for ability to attain a standing position was significantly better in the air-pillow recovery group, compared with the score in the conventional padded stall recovery group (3.7 ± 1.5 and 4.3 ± 1.5, respectively). Mean quality score for overall recovery quality did not differ significantly between the 2 groups (2.4 ± 0.9 and 2.5 ± 1.0, respectively).

Interventions and complications—The number of interventions for horses recovering on the padded floor and on the air pillow were not significantly different (12 and 13, respectively). The research reported here did not include documentation of injuries during recovery. However, the records for 277 of the 409 horses in the present study were examined for an unpublished companion study. Those data indicated that there were no significant differences in interventions or injuries between horses in the 2 recovery groups. For 148 horses in the airpillow recovery group, 8 had minor abrasions; 4 had cast or bandage repair; 3 had laceration of the tongue, lip, or eye; and 2 developed myositis. Among these 148 horses, 118 (80%) recovered from anesthesia without injury. For 129 horses in the conventional padded stall recovery group, 5 had minor abrasions; 1 had cast or bandage repair; 6 had laceration of the tongue, lip, or eye; and none developed myositis. Among these 129 horses, 105 (81%) recovered from anesthesia without injury. Data also indicated that the preoperative American Society of Anesthesiologists' scores of the 2 groups did not differ, as would be expected from the random allocation of horses to 1 of the 2 recovery substrates.

Discussion

In nature, survival of a fallen injured horse may depend on its ability to rapidly roll into a sternal position and regain its footing. However, for a horse recovering from anesthesia, slower and safer standing is preferable because severe, self-inflicted trauma can occur when a horse prematurely rolls to a sternal position and when it attempts to stand too early during the recovery phase.

Various recovery systems have been developed to minimize the potential complications associated with recovery from anesthesia in horses. Padded recovery stalls are most commonly used because they require a low level of maintenance and minimal manpower during use. Thick foam floor mats have also been used in an attempt to protect horses from development of myopathies or neuropathies, but they can hinder the horses' ability to attain a standing position.³ Furthermore, if the thick mats do not completely cover the recovery stall floor, horses can stumble over them when attempting to rise.

Water-based systems have also been devised. A system developed at the University of Pennsylvania places an anesthetized horse within a US Navy 6-man life raft in a large pool; once the horse is alert, it is re-sedated and transferred via a sling to a recovery stall. Although this system provides a relatively safe recovery, disadvantages include size limitations of the raft and the expense and manpower required for construction, operation, and maintenance of the pool.⁴ Another water-based recovery system,^e which is similar in concept but even more complex, consists of a water-filled recovery tank floored with a stainless steel hydraulic lift-table. As the adjustable floor is lowered, the horse is submerged with a body sling in place and its head supported above water with crossties and an air-inflated rubber tube. Once the horse is able to support its weight, the hydraulic floor is raised until it is level with the surrounding hospital floor. This system lowers the risks associated with musculoskeletal injury, but its disadvantages include high cost, considerable manpower requirements, and an associated risk of development of pulmonary edema as a result of adverse effects on cardiopulmonary function during submersion of horses under water.^5,7

A third category of assisted recovery techniques involves various types of slings or vertical-lift systems⁸ that are placed while the horse is still anesthetized or sedated. Comparatively low in cost and manpower requirements, sling systems are often used to decrease weight loading on an affected limb when horses have sustained orthopedic injury.

The goals for use of the rapidly inflating-deflating air pillow during recovery from anesthesia were to provide an alternative low-cost recovery system with minimal manpower requirements that would safely prolong recumbency in horses, thereby improving recovery quality and minimizing injuries during the process of standing. Results of the present study suggest that these goals were realized. Once the horses were placed in the recovery stall, the time to successful attainment of a standing position was approximately 9 minutes longer for horses that were recovering on the air pillow than for horses that were recovering on the padded recovery stall floor (a significant difference between groups). Although it must be noted that time measures were influenced by the anesthetist's subjective decision when to deflate the air pillow, it was also evident that the air pillow promoted a longer rest period before horses began to struggle. This is illustrated by the fact that elapsed intervals to first movement and first attempt to attain sternal recumbency (intervals that were unaffected by the anesthetist) were both significantly longer for horses that were recovering on the air pillow than for horses recovering on the padded stall floor. We postulate that the sound of the shrouded fan minimized external noises and the supportive air cushion helped eliminate painful pressure points, thus ameliorating factors that might stimulate early efforts to rise. When in operation, the fan was noisy outside of the recovery stall; however, the sound was markedly muted inside the recovery stall. After completion of this study, the noise levels were measured with a sound meter.^d Background noise in the surgery room was 61 dB without the fan running. During operation of the fan, sound readings obtained at a position 5 feet from the fan and within the recovery stall were 80 dB and 69 dB, respectively.

Equally important, once the pillow was deflated, the horses in the air-pillow recovery group were able to stand after significantly fewer attempts and the quality of standing was improved, compared with horses in the padded stall floor recovery group. Moreover, the air pillow and padded stall floor appeared to be equally safe as a substrate for recovery from anesthesia; the subset analysis revealed no significant differences in injuries or interventions between the 2 recovery groups. Most of the horses that recovered on the padded stall floor and air pillow had no recovery-associated injuries whatsoever. Two horses in the air-pillow recovery group developed myositis; it is the authors' opinion that the myositis was most likely attributable to predisposing factors, duration of surgery, and positioning rather than to the recovery method.

It was not unexpected that horses that recovered from anesthesia on the air pillow made significantly more attempts to roll into a sternal position before succeeding (and thus spent more time in the attempt, which also increased the interval to standing), compared with the horses that recovered on the padded stall floor. When the pillow is inflated, its compliant padding makes it extremely difficult for a horse to roll. The decision to deflate the pillow so that the horse could successfully roll into a sternal position was made on the basis of the anesthetist's clinical assessment of the situation, rather than being standardized for the study. However, although leaving this decision to the anesthetist was medically sound, it affected practical aspects of the data analysis. For example, the independent evaluator who reviewed the videotaped recoveries and scored the horses' overall recovery quality used a scoring system that penalized multiple attempts to attain sternal recumbency. Because the air pillow was deflated only after a horse made 3 or 4 aggressive attempts to attain a sternal position, horses in this recovery group received an artificially higher (ie, more unacceptable) overall recovery quality score. Thus, it was not surprising that the mean overall recovery quality scores between the 2 recovery groups were not significantly different.

Further investigation involving horses will be necessary before direct comparisons can be made between the rapidly inflating-deflating air-pillow recovery system and systems other than a conventional padded stall. Future research design should include standardized anesthetic protocols, deflation of the pillow after a standardized number of attempts to attain sternal recumbency, and direct comparisons of different types of surgery and complications. Nonetheless, it is already apparent that the rapidly inflating-deflating pillow provides a safe and adaptable system for recovery from anesthesia in horses. Compared with several other recovery systems, the air-pillow system requires relatively little manpower and maintenance. During routine recovery from anesthesia, once the horse is positioned on the air pillow, only 1 person is needed to open the zippers and turn the fan off when deflation is desired. The air pillow can be retrofitted to almost any existing recovery stall to help minimize the complications associated with recovery from anesthesia in horses. If warranted, the air pillow can be deflated or easily removed from the recovery stall. Finally, although data regarding assisted recoveries were not formally included in the present study, it is our experience that this air-pillow system can readily be used in conjunction with head and tail ropes for any high-risk patient, such as a horse with a fracture repair. After the horse is in the recovery stall, the head and tail ropes are placed in a routine fashion and the pillow is inflated while 2 to 3 people in the recovery stall handle the ropes. Once the horse has made several aggressive attempts to stand, the pillow is deflated and the horse is assisted with the ropes during the standing phase. Should the pillow become soiled with water or urine, the traction remains satisfactory. The top cover and the pillow are easily cleaned following recovery. If excessive debris is present, the top cover can be readily removed and sprayed with water and scrubbed with disinfectant. The pillow is also easily removed from the stall to be cleaned or folded up for storage.

In 2004, Kansas State Veterinary Medical Teaching Hospital outfitted 2 recovery stalls with rapidly inflatingdeflating air pillows of the design described in this report. Occasionally, the pillows also are used to provide a supportive and well-cushioned environment for recumbent horses. The current price for a 16 × 14-foot pillow, top cover, and ducted fan is approximately $3,500; additional cost would be incurred to drill 4 corner holes in the stall, install 4 rope cleats, and fabricate the hole and plug in the door to accommodate the pillow spout. In 2001, a new, waterproof floor-pad system^f was installed under the air pillows in both aforementioned recovery stalls. Since 1995, the typical cost of repairs per pillow (done inhouse) has been < $50/y.