Equine anesthesia is associated with an extraordinarily high incidence of perioperative mortality. One prospective study1 involving 41,824 cases in 142 different clinics in 19 different countries found a 2.1% mortality rate for all horses within 7 days of general anesthesia. If only anesthetics for noncolic procedures were considered, the perianesthetic mortality rate was still nearly 1 in 100 cases. Approximately one-third of these deaths were due to postanesthetic fractures and myopathies, for which recovery quality likely played a contributing role.
Horses may exhibit dysphoria or excitement during recovery from inhalation anesthesia. Poor recoveries are characterized by repetitive head banging on the ground, paddling limb movements, lunging, rolling, flopping, falling, knuckling, uncoordinated attempts to stand, or a combination of these behaviors.2,3 These behaviors resemble emergence delirium in humans in which patients scream, kick, thrash, or exhibit altered mental status during the first 30 minutes of anesthetic recovery.4 Although long-term consequences in human patients are unclear, emergence delirium in horses may predispose to cuts, abrasions, bruises, or, more seriously, fractures, tendon or ligament tears, myopathies, and death.
The speed of anesthetic recovery, types of anesthetic drugs used, presence of postoperative pain, and patient age and temperament may all contribute to emergence delirium in people. In prospective, randomized, blinded trials5–7 in children, α2-adrenoreceptor agonists have been shown to decrease the incidence and severity of postanesthetic emergence delirium, presumably as a consequence of the sedative and analgesic properties of these drugs.5–7 Similarly, postanesthetic sedation using α2-adrenoreceptor agonists in horses prolongs recovery time but improves recovery quality from isoflurane or sevoflurane, compared with saline controls.8–11
However, which α2-adrenoreceptor agonist and what dose should be used to best improve recovery quality have not been established. Compared with xylazine, the α2-adrenoceptor agonist romifidine is approximately 10 times as potent, has a longer duration of clinical effect, and produces less ataxia at doses that similarly reduce responses to tactile stimulation.12 As a result, romifidine might provide analgesia and tranquilization during anesthetic recovery while minimizing α2-adrenoceptor agonist–mediated ataxia that could otherwise negate some of the recovery benefits. Santos et al9 reported that low, equipotent doses of xylazine or romifidine improved recovery quality from isoflurane anesthesia, but the investigators found no difference between agents. However, only 6 horses were studied, parametric statistical tests were used to analyze a multidimensional composite score, and other contributors to anesthetic recovery were not factored in the model to help reduce the error sum-of-squares in the analysis. Hence, analytic methods and sample size may have precluded detection of a drug effect. Bartman, et al10 also reported no difference in recovery quality from postanesthetic administration of an equipotent dose of either xylazine or romifidine. Yet in addition to other recovery effects not being controlled for in the statistical analysis, horses that received xylazine postoperatively received xylazine for premedication, whereas horses that received romifidine postoperatively received romifidine for premedication. Since romifidine has a longer duration of action than xylazine,12 the contribution of preanesthetic α2-adrenoceptor agonist to postanesthetic sedation for short procedures would be greater for romifidine than for xylazine. Hence, the cumulative postanesthetic α2-adrenoceptor agonist plasma concentration-to-potency ratio between agents may not have been equivalent.
The purpose of the study reported here was to determine whether α2-adrenoreceptor agonist dose, agent (xylazine or romifidine), or both improve anesthetic recovery quality in an equine clinical trial that used a standardized anesthetic protocol. To increase the likelihood of detecting a drug effect, we simultaneously determined whether other signalment, anesthesia, or surgical factors may explain some of variability in patient anesthetic recovery quality. For example, preanesthetic temperament of a horse influences anesthetic recovery quality and time to first head lift2; since temperament is influenced by horse breed and age, these objective measurements were simultaneously examined. Anesthetic washout during anesthetic recovery is influenced by sedative administration, which decreases minute alveolar ventilation,13 total anesthetic time, anesthetic depth, and use of an oxygen demand valve in recovery; these factors were also included for study since faster washout could affect recovery times and qualities. Moreover, because a given postanesthetic sedative-analgesic drug and dose may produce variable effects in an individual horse on the basis of temperament, presence or absence of pain, and concurrent doses of maintenance anesthetics, it was necessary that a study design account for these possible cofactors.
We hypothesized that horses sedated with romifidine would have better recovery scores than horses sedated with xylazine. We also hypothesized that increasing the dose of the postanesthetic α2-adrenoreceptor agonist would significantly improve recovery quality, assuming that the drug did not also worsen postanesthetic ataxia.
Materials and Methods
This study was designed as a prospective, randomized, blinded clinical trial using client-owned horses that were examined at the Veterinary Medical Teaching Hospital at the University of California-Davis from 2007 to 2009. Drugs and doses used in this study all fell within accepted hospital standards of care for equine postanesthetic sedation. This protocol was approved by the Animal Use and Care Committee at the University of California-Davis.
Horses included in this study had no evidence of cardiovascular, respiratory, hepatic, renal, or neurologic disease on the basis of physical and hematologic evaluation and met the criteria established by the American Society of Anesthesiologists classification system for a physical status of I to II.14 In addition, horses needed to be 2 years of age or older, weigh between 270 and 680 kg (594 and 1,496 lb), and require general anesthesia for an elective procedure lasting 1 to 4 hours. Horses undergoing ocular procedures were excluded since many wear protective eye coverings during recovery that may limit vision and independently affect recovery quality. Horses with major or unstable orthopedic fractures were excluded from study since these animals generally require special recovery techniques to minimize risks of disruption of the repair. Finally, horses undergoing a laparotomy were excluded because of special concerns about analgesia management and differences in postoperative morbidity and mortality rates, compared with those in the general equine anesthetic population.1 To achieve a type I error rate of 5%, power of 80%, and regression effect size (Cohen's f2)15 of 0.25 with a model that could contain up to 16 variables, a minimum of 91 horses would be required for this study. One hundred horses were enrolled to allow for slightly greater variability than initially estimated.
Some horses initially enrolled in this study did not complete it. A priori criteria for censoring included intraoperative reassessment of patient needs with a decision to assist recovery or to administer nonstudy systemic or regional opioid analgesics. Additionally, patients that received any injectable anesthetics within the final 30 minutes of inhaled anesthesia were also removed from the study. Opioids16,17 and ketamine18,19 may independently affect recovery quality or recovery times in horses sedated with α2-adrenoreceptor agonists; thus administration of these drugs potentially could have increased response variability within the xylazine or romifidine treatments.
By means of a random number table, horses were allocated to receive 1 of 4 treatments: xylazinea (100 μg/kg [45 μg/lb], IV), xylazine (200 μg/kg [91 μg/lb], IV), romifidineb (10 μg/kg [4.5 μg/lb], IV), or romifidine (20 μg/kg [9.1 μg/lb], IV). Drugs were prepared by the hospital pharmacy staff with one syringe containing 1 μL/kg (0.45 μL/lb) plus another syringe containing 2 μL/kg (0.91 μL/lb) for each patient; one syringe contained the study drug, and the other contained saline (0.9% NaCl) solution, which allowed the investigators to be blinded to drug dose since both would be simultaneously administered to the patient. The study drug was delivered to the anesthetist by a different pharmacy technician who was also unaware of the assigned treatment. A sealed envelope containing the study drug and dose was dispensed with the drugs so that knowledge of any administered drug would be accessible in the event of an emergency; unopened envelopes were then returned to the pharmacy following recovery from anesthesia so that the investigators and all anesthesiologists would remain blinded to the treatment administered.
The anesthetic protocol was standardized for all patients. All horses were fasted preoperatively for 8 to 12 hours, but water was always freely available. Prior to anesthesia, a 14-gauge catheter was placed percutaneously in a jugular vein by means of aseptic technique, and the mouth was rinsed with water. Horses were sedated with xylazine (0.3 to 0.5 mg/kg [0.14 to 0.23 mg/lb], IV), and anesthesia was induced via guaifenesinc (20 to 50 mg/kg [9.1 to 22.7 mg/lb]) followed by diazepamd (0.05 mg/kg [0.02 mg/lb]) plus ketaminee (2 mg/kg). Horses were orally intubated by use of a cuffed endotracheal tube, and anesthesia was maintained with isoflurane in oxygen delivered in a large animal anesthesia breathing circuit.f
Lactated Ringer's solution was administered IV at 5 to 10 mL/kg/h (2.3 to 4.5 mL/lb/h) throughout anesthesia. Heart rate was monitored with a base-apex ECG, and arterial blood pressure was monitored via a catheter placed directly in a facial, auricular, metatarsal, or tail artery connected to a transducerg by a fluid-filled extension set and displayed on a monitor.h Dobutamine hydrochloridei was administered IV as needed to maintain mean arterial blood pressure ≥ 70 mm Hg. Respiratory rate, end-tidal carbon dioxide, and anesthetic concentrations were monitored with a respiratory gas analyzer,j and once normotension was achieved (mean arterial blood pressure ≥ 70 mm Hg), horses were mechanically ventilated to maintain partial pressure of end-tidal carbon dioxide between 40 and 50 mm Hg. Arterial blood was collected every 30 to 90 minutes to monitor for measurements of Pao2, Paco2, pH, and electrolytes by use of an automated analyzer.k Prior to the end of anesthesia, the final end-tidal isoflurane concentration was recorded.
Following coincident discontinuation of mechanical ventilation and disconnection from the anesthesia circuit, horses were conveyed to a 3.5 × 2.75-m padded recovery stall with thick rubberized flooring. The endotracheal tube was secured to the mouth with tape through which oxygen was insufflated at 15 L/min. A padded helmet was placed on the head, which was secured to a rope intended to prevent injury from head banging, but the rope was not used to actively stimulate or direct the horse during attempts to stand. If the horse was apneic for 3 minutes, an oxygen demand valve,l was used to deliver 1 to 2 breaths/min until the horse began spontaneous ventilation. After the first or second spontaneous breath, the test syringes containing drug and saline solution were administered as an IV bolus. The recovery stall doors were then closed, and recovery was video recorded and observed from outside the stall.
Objective measures of recovery such as total anesthesia time (time from anesthetic induction until circuit disconnection), use of the oxygen demand valve, time between circuit disconnection until first spontaneous movement, number of repetitive head bangs on the floor, time from circuit disconnection until first successful sternal recumbency, time from disconnection until final standing, and number of attempts to stand were recorded by the attending anesthesiologist. Subjective recovery measures including the VAS score, evaluation of ataxia, and limb-paddling severity (none, mild, moderate, or severe) were always made by the same veterinary anesthesiologist either during anesthetic recovery or during review of the video recording. The VAS was scored by marking a 100-mm scale, with 0 indicating a poor recovery quality and 100 indicating an excellent or ideal recovery; the VAS equaled the distance between the mark and 0 in millimeters. Treatment codes for study horses were not revealed until after all patients had been studied and all subjective measurements were completed.
Continuous data were summarized as mean ± SD, median, and range. Binary data, such as anesthesia for a painful procedure and demand valve use, were coded as 1 if the event occurred and 0 if it did not. Binary variables were used to code N-1 categorical variables with the remaining category serving as the reference group. To examine the effect of patient sex, separate variables for gelding and stallion were created, with mare serving as the reference group. Breed effects were examined by creating separate variables for Arabian, Quarter Horse, and Thoroughbred horses; all other breeds were combined as the reference group that consisted largely of various warmblood breeds. For the recovery drugs, binary variables for 200 μg/kg of xylazine, 10 μg/kg of romifidine, and 20 μg/kg of romifidine were created, and the 100 μg/kg xylazine dose was used for reference.
A multivariate stepwise least squares linear regression model was created with statistical softwarem to test for correlation between the VAS score and the following variables: age, weight, sex, breed, anesthesia for a painful procedure, total anesthetic time, end-tidal isoflurane concentration at the time of circuit disconnection, use of an oxygen demand valve after circuit disconnection, time between circuit disconnection and first spontaneous movement, time between circuit disconnection and standing recovery (total recovery time), and postanesthetic recovery drug and dose. Continuous data that were not normally distributed, as evidenced by Wilk-Shapiro tests and visual inspection of probability plots, were mathematically transformed to a variable with a normal distribution for use in the regression analysis. A value of P ≤ 0.05 was used as the criterion for variable inclusion in the forward stepwise model and for variable exclusion in the backward stepwise model.
To understand factors that influence an overall recovery VAS score, Spearman correlation coefficients were used to measure separately the association between VAS and limb paddling severity, ataxia severity, number of repetitive head slaps on the floor, the number of attempts to sternal recumbency, and the number of attempts to stand. For all tests, a value of P ≤ 0.05 was considered significant.
Results
Complete data were available for 101 horses. Summary data by drug treatment group were summarized (Table 1). Fifty-nine of the horses enrolled were of Arabian, Quarter Horse, or Thoroughbred breeds. Of the remaining horses classified as other, there were 25 various warmblood breeds, 4 Morgans, 3 Paints, 2 Quarter Horse/Thoroughbred crosses, and 1 each of Andalusian, Appaloosa, Mustang, Palamino, Paruvian Paso, pony, Standardbred, and Trakehner breeds. Most elective procedures (44%) involved endoscopy of joints, bursae, or tendon sheath, although orchidectomies (20%) and nonpainful diagnostic imaging procedures (11%) were also common (Table 2).
Summary of regression model variables factored by treatment drug and dose in a study of effects of postanesthetic sedation with romifidine or xylazine administered IV on quality of recovery from isoflurane anesthesia in horses.
| Parameter | Xylazine | Romifidine | ||
|---|---|---|---|---|
| 100 μg/kg | 200 μg/kg | 10 μg/kg | 20 μg/kg | |
| Sample size | 25 | 25 | 26 | 25 |
| Signalment | ||||
| Age (y) | 7 ± 4 | 6 ± 4 | 7 ± 4 | 7 ± 3 |
| 7 (2–15) | 5 (2–17) | 5 (2–17) | 7 (3–14) | |
| Sex | ||||
| Mare | 5 | 3 | 6 | 6 |
| Stallion | 9 | 5 | 6 | 10 |
| Gelding | 11 | 17 | 14 | 9 |
| Weight (kg) | 516 ± 87 | 528 ± 66 | 507 ± 77 | 496 ± 70 |
| 525 (272–656) | 525 (400–680) | 514 (375–655) | 494 (403–665) | |
| Breed | ||||
| Arabian | 1 | 1 | 4 | 2 |
| Quarter Horse | 4 | 5 | 5 | 10 |
| Thoroughbred | 6 | 6 | 10 | 3 |
| Other | 14 | 13 | 7 | 10 |
| Anesthesia events | ||||
| Total anesthesia time (min) | 124 ± 45 | 117 ± 39 | 126 ± 44 | 132 ± 34 |
| 110 (60–240) | 110 (65–205) | 120 (64–210) | 124 (75–200) | |
| End-tidal isoflurane (%) | 1.5 ± 0.2 | 1.5 ± 0.1 | 1.5 ± 0.1 | 1.5 ± 0.1 |
| 1.5 (1.2–1.9) | 1.5 (1.2–1.8) | 1.5 (1.3–1.7) | 1.5 (1.2–1.7) | |
| Demand valve use | ||||
| No | 13 | 18 | 16 | 13 |
| Yes | 12 | 7 | 10 | 12 |
| First movement time (min) | 16 ± 7 | 17 ± 8 | 21 ± 9 | 26 ± 9 |
| 15 (4–32) | 15 (0–39) | 19 (6–39) | 24 (6–49) | |
| Total recovery time (min) | 35 ± 16 | 40 ± 21 | 35 ± 9 | 50 ± 15 |
| 32 (16–72) | 34 (11–96) | 36 (17–53) | 49 (17–76) | |
| Painful procedure | ||||
| No | 4 | 2 | 4 | 2 |
| Yes | 21 | 23 | 22 | 23 |
| VAS score (mm) | 54 ± 27 | 65 ± 25 | 68 ± 22 | 69 ± 24 |
| 58 (0–92) | 74 (21–98) | 68 (14–94) | 72 (23–98) | |
Data are number of animals or mean ± SD and median (minimum–maximum).
Frequency of procedures performed (n = 101 total) during anesthetic maintenance and their classification as painful or not painful for the patients in Table 1.
| Procedure | Cases | Painful |
|---|---|---|
| Arthroscopy (1 joint) | 32 | Yes |
| Arthroscopy (2 joints) | 5 | Yes |
| Arthroscopy (3 joints) | 1 | Yes |
| Bone or hoof débridement | 2 | Yes |
| Bursoscopy | 1 | Yes |
| Castration | 9 | Yes |
| Castration + umbilical hernia | 1 | Yes |
| Cryptorchid castration | 10 | Yes |
| CT | 9 | No |
| CT + angiography | 3 | No |
| CT + angiography + débridement | 1 | Yes |
| Desmotomy or fasciotomy | 3 | Yes |
| Fracture (sesamoid, splint) | 6 | Yes |
| Mass resection or excision | 4 | Yes |
| Neurectomy | 7 | Yes |
| Soft tissue wound débridement | 2 | Yes |
| Tenoscopy | 5 | Yes |
There were no anesthetic complications associated with anesthetic induction. All but 1 horse required IV dobutamine administration for blood pressure support during anesthetic maintenance. Alveolar-to-arterial Po2 gradients were increased in many patients during anesthesia, but only 4 had a measured Pao2 between 60 and 80 mm Hg, and no horses experienced more severe hypoxemia. Of the former, 3 horses were assigned the 100 μg/kg xylazine dose for recovery and had VAS scores of 12, 58, and 76. The remaining horse was assigned to the 200 μg/kg xylazine treatment group and had a VAS of 23. The low number of affected patients in this study precluded further inferences pertaining to hypoxemia. No other complications during anesthetic maintenance were reported.
Visual analog scale scores among all horses ranged from 0 to 98, with a median score of 67 and a mean ± SD of 63 ± 25. Upon standing following a violent recovery, 1 Thoroughbred horse that had received 100 μg/kg of xylazine for recovery sustained a comminuted biarticular pastern fracture and was euthanized. More minor injuries occurred in 13 other horses and were composed primarily of cuts or abrasions on the face. When present, injuries of any kind almost exclusively occurred in horses with a VAS ≤ 33. The VAS negatively correlated with subjective assessments of limb paddling severity during recovery and with subjective assessments of ataxia once the horse was standing. Visual analog scale scores were also negatively correlated with several objective recovery events, namely the number of repetitive head slaps on the floor, the number of attempts to lie in sternal recumbency, and the number of attempts to stand.
Normalized Q-Q plots and Wilk-Shapiro tests indicated that patient age, the first movement time, and the total recovery time deviated significantly from a normal distribution. This was resolved by use of square root transformations of these variables, which were then subsequently used in regression modeling.
A backward and forward stepwise regression analysis identified the same 4 variables as significant predictors of VAS. Total anesthesia time measured in minutes had a slope of −0.19 ± 0.06 (mean ± SEM) and a P = 0.001. The Arabian breed, when compared with the reference of horses classified as other, had a slope of −20 ± 8 and a P = 0.01. Painful procedures (Table 2), when compared with nonpainful procedures as a reference, had a slope of −14 ± 7 and a P = 0.04. Administration of 20 μg of romifidine/kg, compared with the 100 μg/kg reference dose of xylazine, had a slope of 10 ± 5 and P = 0.05. The regression constant also significantly differed from zero, having a mean of 99 ± 8 and P < 0.001. The resulting regression equation predicting VAS was summarized as follows:
where BArab indicates whether a horse is an Arabian breed (1 = true and 0 = false), Tanes is the total anesthesia maintenance time (in minutes), Ppain is the a priori assessment of a procedure as either painful (1 = true) or not painful (0 = false), and R20 indicates whether 20 μg/kg of romifidine was (1 = true) or was not (0 = false) used for postanesthetic sedation. No other variable considered in the model reached the criteria for inclusion at the stated level of significance (P ≤ 0.05). The coefficient of determination (R2) for the model was 0.25.
In both backward and forward stepwise regression procedures, R20 was the last variable to be respectively retained or entered into the model. The mean VAS scores (in millimeters) for each postanesthetic treatment group were 65 ± 14 for 100 μg/kg of xylazine, 62 ± 14 for 200 μg/kg of xylazine, 63 ± 15 for 10 μg/kg of romifidine, and 61 ± 14 for 20 μg/kg of romifidine. Were the residual sum of squares not reduced by inclusion of other significant VAS predictors in the stepwise regression model, the postanesthetic drug treatment effect for R20 in the present study would not have been detectable. This is because other highly significant variable responses were not identically distributed across drug treatments and thereby behaved as confounding factors. For example, despite randomization, horses that received 20 μg/kg of romifidine also had the longest mean anesthetic times and underwent more procedures classified as painful (Table 2), compared with all other drug treatments groups. Although 20 μg/kg romifidine administration was associated with a 10-mm improvement on a 100-mm VAS via a multivariate analysis, longer and potentially painful procedures in this group would have obscured this treatment effect in a univariate analysis. Stepwise regression can control for confounding factors,20 which, in this study, enabled detection of a drug treatment effect for R20.
Discussion
In the present study evaluating healthy adult horses anesthetized with isoflurane for > 1 hour, the results supported the use of 20 μg/kg of romifidine for postanesthetic sedation to improve recovery versus lower romifidine doses or xylazine. Four modifiers of recovery quality from isoflurane anesthesia were identified. Of these, only use of 20 μg/kg of romifidine was associated with an improvement in the VAS score. Painful procedures, longer anesthesia times, and Arabian horse breeds were all associated with poorer recovery quality. Of note, longer anesthetic duration and orthopedic surgeries for fractures, perhaps among the most painful equine noncolic surgical procedures, were also correlated with an increased risk of perioperative death in horses.1 It is not surprising that factors associated with violent anesthetic recoveries that increase risks of catastrophic injuries might likewise be associated with increased postanesthetic death.
The α2-adrenoreceptor agonists may improve recovery quality through at least 2 mechanisms. During recovery, isoflurane remains detectable in the end-tidal breath at concentrations approximately ≥ 10% of the inspired circuit concentration for at least 25 minutes.2 Subanesthetic inhalant concentrations by definition do not prevent voluntary movement, but nonetheless can exert important neurologic effects such as amnesia,21–24 diminished cognitive function,25,26 loss of proprioception,27,28 altered motor neuron function,29,30 and hyperalgesia.31,32 Sedatives administered during recovery delay attempts by the horse to stand.8–10 Thus, there is more time for inhaled anesthetic washout and return of neurologic functions, assuming alveolar ventilation is maintained, which should allow for a more sensible and coordinated recovery. The higher dose and longer duration of action of romifidine may have improved recovery simply because they provided the most time for isoflurane washout.
Another possible mechanism for improved VAS scores in the present study is the contribution of α2-adrenoreceptor agonists to analgesia. In horses, romifidine may produce more intense and longer-lasting antinociception to some nociceptive stimuli when compared with xylazine and detomidine,33 thereby preventing premature arousal and attempts to rise. Painful procedures (Table 2) were also a negative independent predictor of recovery quality, which is in agreement with a previous study34 relating poorer recoveries with increased surgical invasiveness. Aside from its humane and ethical justifications, adequate postoperative analgesia could be important for decreasing the morbidity and mortality associated with poor anesthetic recoveries in horses.
Anesthetic uptake by peripheral tissues continues throughout maintenance until the tissue partial pressure comes to equilibrium with the alveolar anesthetic partial pressure. With the contemporary agents, equilibrium between vessel-rich groups such as brain, heart, and kidney occurs within minutes. However, equilibrium in tissues with lower flow, such as muscle and intestine, may take hours. At the other extreme, isoflurane partial pressure equilibrium between arterial blood and fat is never achieved during most anesthetic periods. Consequently, as anesthetic time increases, total body anesthetic content also increases. During recovery, anesthetic leaving low- and moderate-flow tissue reservoirs can redistribute to the CNS, resulting in delayed alveolar washout and prolonged recovery time.2,35 Presumably, delayed alveolar washout should also prolong neurologic and behavioral effects associated with subanesthetic inhalant concentrations and thereby explain a negative association between anesthetic duration and recovery quality observed in the present study and a previous study.34
When controlled for other significant factors, Arabian horses tended to have a 20-mm lower VAS than other horse breeds in this study. In a survey of compulsive behaviors in stabled horses, Arabians exhibited a much higher incidence of stall walking and stall kicking than most other horse breeds36; it has been postulated that these activities may be related to anxiety or stress. Humans that are anxious, fearful, emotional, or impulsive preoperatively are more likely to exhibit delirium during anesthetic recovery.4 The effect of horse breed on recovery quality might more accurately reflect an effect of horse temperament.37
As with any clinical trial, this study had several limitations. First, the treatment codes were disclosed after VAS scoring by only a single evaluator; subjective recovery scores may have altered slightly if additional evaluators were included. However, the VAS tends to show relatively little intraobserver variability, even among less experienced individuals, and it correlates well with other composite scoring systems.38,39 Hence, additional evaluators probably would not have altered the significant findings. Second, there was no saline control against which to compare the different α2-adrenoceptor agonist drugs and doses. This was a deliberate decision. There is sufficient evidence to indicate that recovery quality in healthy horses following 1 or more hours of inhalant anesthesia tends to be worse when no postanesthetic sedation is administered.9,10,40 Under these circumstances, use of a placebo control in patients that would likely increase the risk of poor recovery outcomes and patient morbidity would have been unethical and may not have met current standards of care for equine anesthetic practice.
The 20 μg/kg IV romifidine dose increased VAS more than any other recovery drug treatment. This raises the question about whether this is the most effective dose of romifidine for equine recovery or whether even higher doses or different administration routes might yield additional benefit. On the other hand, horses recovering from anesthetics such as desflurance, which have a very low blood and tissue solubility, may not realize additional benefits from using longer-acting α2-adrenoreceptor agonists at high doses since the elimination rate for very insoluble inhaled agents is much more rapid.
The clinical trial reported here included only healthy horses anesthetized for 1 to 4 hours. Horses that are metabolically ill or exhausted might not need postanesthetic sedation since preexisting obtundation or weakness might already prolong recovery and allow sufficient time for anesthetic washout. Horses sedated for very brief periods with high doses of sedatives administered for premedication likewise may not need additional sedation for recovery. In these cases, sufficient postanesthetic sedation could be achieved by residual concentrations of premedication drugs.
Finally, the drug treatment, signalment, surgical, and anesthetic factors identified in the present study describe only 25% of the variability in VAS. We are left with the unsettling realization that 75% of patient variability in recovery quality remains unexplained. Future recognition of other determinants of recovery quality and development of preventative treatment for emergence delirium will be essential to reducing perianesthetic mortality risk in horses.
ABBREVIATION
| VAS | Visual analog scale |
Rompun, 100 mg/mL, Bayer, Pittsburgh, Pa.
Sedivet 10 mg/mL, Boehringer Ingelheim, St Louis, Mo.
UCSF Drug Product Services Laboratory, San Francisco, Calif.
Baxter Healthcare, Deerfield, Ill.
KetaVed, Vedco, St Joseph, Mo.
2800C, Mallard Medical, Redding, Calif.
DT-XX, Becton Dickinson, Franklin Lakes, NJ.
Spacelabs, Issaquah, Wash.
Dobutrex, Lilly, Indianapolis, Ind.
Poet IQ, Criticare, Waukesha, Wis.
ABL800 Radiometer America, Westlake, Ohio.
Model 5040, Hudson Oxygen Therapy Sales, Wadsworth, Ohio.
SPSS, version 11, SPSS Inc, Chicago, Ill.
References
1. Johnston GM, Eastment JK, Wood JLN, et al. The confidential enquiry into perioperative equine fatalities (CEPEF): mortality results of phases 1 and 2. Vet Anaesth Analg 2002; 29:159–170.
2. Whitehair KJ, Steffey EP, Willits NH, et al. Recovery of horses from inhalation anesthesia. Am J Vet Res 1993; 54:1693–1702.
3. Donaldson LL, Dunlop GS, Holland MS, et al. The recovery of horses from inhalant anesthesia: a comparison of halothane and isoflurane. Vet Surg 2000; 29:92–101.
4. Vlajkovic GP, Sindjelic RP. Emergence delirium in children: many questions, few answers. Anesth Analg 2007; 104:84–91.
5. Tesoro S, Mezzetti D, Marchesini L, et al. Clonidine treatment for agitation in children after sevoflurane anesthesia. Anesth Analg 2005; 101:1619–1622.
6. Shukry M, Clyde MC, Kalarickal PL, et al. Does dexmedetomidine prevent emergence delirium in children after sevoflurane-based general anesthesia? Paediatr Anaesth 2005; 15:1098–1104.
7. Ibacache ME, Munoz HR, Brandes V, et al. Single-dose dexmedetomidine reduces agitation after sevoflurane anesthesia in children. Anesth Analg 2004; 98:60–63.
8. Matthews NS, Hartsfield SM, Mercer D, et al. Recovery from sevoflurane anesthesia in horses: comparison to isoflurane and effect of postmedication with xylazine. Vet Surg 1998; 27:480–485.
9. Santos M, Fuente M, Garcia-Iturralde R, et al. Effects of alpha-2 adrenoceptor agonists during recovery from isoflurane anaesthesia in horses. Equine Vet J 2003; 35:170–175.
10. Bartmann CP, Schiemann V. Development of an intrauterine distension pressure for hysteroscopy in the horse [in German]. Dtsch Tierarztl Wochenschr 2003; 110:43–48.
11. Bienert A, Bartmann CP, von Oppen T, et al. Standing behavior in horses after inhalation anesthesia with isoflurane (Isoflo) and postanesthetic sedation with romifidine (Sedivet) or xylazine (Rompun) [in German]. Dtsch Tierarztl Wochenschr 2003; 110:244–248.
12. England GC, Clarke KW, Goossens L. A comparison of the sedative effects of three alpha 2-adrenoceptor agonists (romifidine, detomidine and xylazine) in the horse. J Vet Pharmacol Ther 1992; 15:194–201.
13. Steffey EP, Kelly AB, Farver TB, et al. Cardiovascular and respiratory effects of acetylpromazine and xylazine on halothane-anesthetized horses. J Vet Pharmacol Ther 1985; 8:290–302.
14. Dripps RD, Lamont A, Eckenhoff JE. The role of anesthesia in surgical mortality. JAMA 1961; 178:261–266.
15. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates, 1988.
16. Bettschart-Wolfensberger R, Dicht S, Vullo C, et al. A clinical study on the effect in horses during medetomidine-isoflurane anaesthesia, of butorphanol constant rate infusion on isoflurane requirements, on cardiopulmonary function and on recovery characteristics. Vet Anaesth Analg 2011; 38:186–194.
17. Clark L, Clutton RE, Blissitt KJ, et al. The effects of morphine on the recovery of horses from halothane anaesthesia. Vet Anaesth Analg 2008; 35:22–29.
18. Wagner AE, Mama KR, Steffey EP, et al. A comparison of equine recovery characteristics after isoflurane or isoflurane followed by a xylazine-ketamine infusion. Vet Anaesth Analg 2008; 35:154–160.
19. Larenza MP, Ringer SK, Kutter AP, et al. Evaluation of anesthesia recovery quality after low-dose racemic or S-ketamine infusions during anesthesia with isoflurane in horses. Am J Vet Res 2009; 70:710–718.
20. Draper RN, Smith H. Applied regression analysis. 3rd ed. New York: Wiley, 1998.
21. Dutton RC, Maurer AJ, Sonner JM, et al. Isoflurane causes anterograde but not retrograde amnesia for pavlovian fear conditioning. Anesthesiology 2002; 96:1223–1229.
22. Perouansky M, Rau V, Ford T, et al. Slowing of the hippocampal theta rhythm correlates with anesthetic-induced amnesia. Anesthesiology 2010; 113:1299–1309.
23. Alkire MT, Gorski LA. Relative amnesic potency of five inhalational anesthetics follows the Meyer-Overton rule. Anesthesiology 2004; 101:417–429.
24. Dwyer R, Bennett HL, Eger EI II, et al. Effects of isoflurane and nitrous oxide in subanesthetic concentrations on memory and responsiveness in volunteers. Anesthesiology 1992; 77:888–898.
25. Heinke W, Schwarzbauer C. Subanesthetic isoflurane affects task-induced brain activation in a highly specific manner: a functional magnetic resonance imaging study. Anesthesiology 2001; 94:973–981.
26. Galinkin JL, Janiszewski D, Young CJ, et al. Subjective, psychomotor, cognitive, and analgesic effects of subanesthetic concentrations of sevoflurane and nitrous oxide. Anesthesiology 1997; 87:1082–1088.
27. Barter LS, Mark LO, Antognini JF. Proprioceptive function is more sensitive than motor function to desflurane anesthesia. Anesth Analg 2009; 108:867–872.
28. Padoan S, Fransson PA, Magnusson M, et al. Postural control reduced by subanesthetic nitrous oxide narcosis. J Vestib Res 1993; 3:173–180.
29. Brandes IF, Zuperku EJ, Stucke AG, et al. Isoflurane depresses the response of inspiratory hypoglossal motoneurons to serotonin in vivo. Anesthesiology 2007; 106:736–745.
30. Antognini JF, Carstens E, Buzin V. Isoflurane depresses motoneuron excitability by a direct spinal action: an F-wave study. Anesth Analg 1999; 88:681–685.
31. Zhang Y, Eger EI II, Dutton RC, et al. Inhaled anesthetics have hyperalgesic effects at 0.1 minimum alveolar anesthetic concentration. Anesth Analg 2000; 91:462–466.
32. Flood P, Sonner JM, Gong D, et al. Isoflurane hyperalgesia is modulated by nicotinic inhibition. Anesthesiology 2002; 97:192–198.
33. Moens Y, Lanz F, Doherr MG, et al. A comparison of the antinociceptive effects of xylazine, detomidine and romifidine on experimental pain in horses. Vet Anaesth Analg 2003; 30:183–190.
34. Young SS, Taylor PM. Factors influencing the outcome of equine anaesthesia: a review of 1,314 cases. Equine Vet J 1993; 25:147–151.
35. Voulgaris DA, Hofmeister EH. Multivariate analysis of factors associated with postanesthetic times to standing in isoflurane-anesthetized horses: 381 cases. Vet Anaesth Analg 2009; 36:414–420.
36. Luescher UA, McKeown DB, Dean H. A cross-sectional study on compulsive behaviour (stable vices) in horses. Equine Vet J Suppl 1998;(27):14–18.
37. Leece EA, Corletto F, Brearley JC. A comparison of recovery times and characteristics with sevoflurane and isoflurane anaesthesia in horses undergoing magnetic resonance imaging. Vet Anaesth Analg 2008; 35:383–391.
38. Vettorato E, Chase-Topping ME, Clutton RE. A comparison of four systems for scoring recovery quality after general anaesthesia in horses. Equine Vet J 2010; 42:400–406.
39. Portier KG, Sena A, Senior M, et al. A study of the correlation between objective and subjective indices of recovery quality after inhalation anaesthesia in equids. Vet Anaesth Analg 2010; 37:329–336.
40. Steffey EP, Mama KR, Brosnan RJ, et al. Effect of administration of propofol and xylazine hydrochloride on recovery of horses after four hours of anesthesia with desflurane. Am J Vet Res 2009; 70:956–963.
