Abstract
Objective
To evaluate whether covering the extremities of cats with highly insulating materials with or without active warming slows the rate of temperature decrease during anesthesia.
Methods
Insulating devices were created from the best insulating material—2 layers of down blanket—as determined by an in vitro study. Female cats undergoing ovariohysterectomy in a veterinary student surgical program were block randomized to active warming (insulation with heating element), passive insulation (insulation only), or control (no covering). Core body temperature was recorded every minute from induction through recovery. Multivariable linear regression was used to evaluate the rate of temperature decrease and lowest recorded temperature, controlling for weight, postinduction temperature, ambient temperature, and (for lowest temperature) anesthesia duration.
Results
49 female cats were enrolled. In the first 30 minutes, controls decreased by 0.12 °F/min, passive by 0.11 °F/min, and active by 0.09 °F/min. After 30 minutes, temperature decline slowed, with rates of 0.05 °F/min for controls, 0.03 °F/min for passive, and 0.01 °F/min for active. The lowest recorded temperatures were 1.2 and 1.9 °F, higher in the passive and active groups, respectively.
Conclusions
Covering the extremities of cats undergoing anesthesia with highly insulating materials slows core temperature decrease.
Clinical Relevance
Covering the extremities of cats resulted in a lowest temperature between 1 and 2 °F greater than controls. While active warming has a greater effect than passive insulation, the absolute difference in lowest temperature, 0.7 °F, may not justify the additional challenges of adding a heating source.
Introduction
Following anesthetic induction, a rapid decline in core body temperature is commonly observed within the first hour, primarily due to the redistribution of blood from the warmer core to the cooler periphery. Heat loss to the environment through radiation, conduction, evaporation, and convection contributes to this effect.1 In the subsequent hour of anesthesia, a slower decline in temperature occurs as heat loss to the environment surpasses metabolic heat production.2 It is estimated that 71% of feline patients experience hypothermia during anesthesia,3 a condition significantly impacting postoperative recovery.4 Hypothermia impairs the metabolism of anesthetic drugs, increasing their effects and the risk of overdose.5 Additionally, it may result in bradycardia and other cardiac arrhythmias.6 Recovery times may be delayed,4 and there is a greater risk of surgical site infection.7,8
Veterinary clinics commonly cover feline extremities during anesthesia to decrease the rate of heat loss, with cotton toddler socks used most frequently.9 In a previous study9 of 164 cats, passive insulation was tested with cotton toddler socks, active warming with cotton toddler socks containing a heating element, and a control group with uncovered extremities. The final temperature of the cotton toddler socks compared to the control group showed no significant differences, while active warming of feline extremities modestly reduced the rate of temperature decrease. The difference between active warming and the control was 0.01 °F/min of anesthesia, resulting in an average difference in recovery of 0.5 °F. While this difference was statistically significant, it lacked clinical significance. Infrared photography of the cotton toddler socks revealed that they were poor insulators. Other materials have been studied for their insulating properties during anesthesia, including bubble wrap, reflective blankets, and absorbent pads.10–13 However, many of them were not compared directly.
The aim of this study was to determine whether highly insulating materials, particularly in combination with active warming, would significantly decrease the rate of heat loss during anesthesia. It was hypothesized that covering the extremities with highly insulating materials would decrease the overall heat loss in both passive insulation and active warming groups compared to a control group with uncovered extremities.
Methods
In vitro study and preparation of the covers
Three trials were run to determine the best insulating material for the extremity covers. For each trial, four 500-mL lactated Ringer fluid bags were heated to 125 °F in a warming cabinet (P-2110; Pedigo Products Inc) and manipulated to evenly distribute the heat. The bags were placed on a counter covered with a terry cloth towel, and 3 bags were wrapped in different combinations of materials that included 0.75-lb wool socks, 650- to 700-fill power down camping blankets, a mylar blanket, and bubble wrap. The selection of materials to trial was based on prior studies.10–13 The fourth bag was left unwrapped as a control. A grill thermometer (TP25; ThermoPro) with 4 temperature sensors was inserted through the injection port, the insulation visualized with a Flir One thermal imaging camera (Teledyne FLIR LLC; Supplementary Figure S1), and the temperature monitored once per minute for 1 hour via the app. Materials were evaluated for their insulating properties through multivariable mixed-effects linear regression with trial as random effect that controlled for starting temperature and ambient temperature. The best insulating material—2 layers of down blanket—was used to create the extremity coverings in the shape of elongated socks. The active warming coverings had a graphene heating element (S.Fine) sewn into the covering that was powered via low-voltage USB to 104 °F.
Clinical study
Female cats presented to Midwestern University student surgical programs between June 1, 2024, and July 11, 2024, for ovariohysterectomy were eligible to be enrolled in the study. The programs were at 2 locations, with the Trap Neuter Return (TNR) program that provides sterilization for asocial cats on the Midwestern campus and a program for social cats from shelters, rescues, or the community held at a partner location. Criteria for inclusion were any female cat ≥ 2 months of age, body condition score (BCS) > 3 (scale 1 to 9), and no abnormal findings on visual examination. Patients that required any procedures in addition to ovariohysterectomy or that did not have 4 legs available to be covered (for example, because of the presence of catheters or blood pressure cuffs) were excluded. Cats were assigned via a block randomization scheme (Randomizer.org), created by a researcher (REK) who was not involved in patient assignment, to 1 of 3 different treatment groups by 2 researchers (JB and MK), including active warming (insulated covering with heating element), passive insulation (insulated covering alone), or control (no covering). The study was approved by Midwestern University’s IACUC (approval AZ-4243). The study was reported following the standards for reporting trials in pets.14
Patients in the active and passive groups had insulating devices placed on all 4 limbs (Figure 1) for the entire duration of anesthesia, from induction to recovery. Core body temperature was measured each minute via a temperature probe (Vet30; SunTech Medical Inc) placed rectally or esophageally from induction of anesthesia until the animal was placed back in the carrier or trap. The placement of the probe was changed from rectal to esophageal after the first 11 cats for convenience, as it was less prone to dislodging in this location. Core temperatures occurring at major time points (after induction, movement into the operating room, surgery start, surgery end, and recovery) were additionally noted. Ambient room temperature was recorded. The coverings and temperature probes were placed as soon as possible after the cats became unresponsive to stimuli following an IM induction of anesthesia. All cats received truncal warming during surgery, and supplementary warming was available in recovery. Paw temperatures were monitored (TP25; ThermoPro) for the active warming group to ensure the temperature remained below 110 °F to prevent the risk of thermal burns.
Coverings and thermal imaging (Flir One; Teledyne FLIR LLC) for the control (A and D), passive (B and E), and active groups (C and F), as shown on a model patient.
Citation: Journal of the American Veterinary Medical Association 2025; 10.2460/javma.25.02.0095
Veterinary students overseen by veterinary faculty completed a physical examination on each patient and recorded parameters such as temperature, heart and respiratory rates, estimated age, and BCS. Two anesthetic induction protocols were used: TNR cats received an IM injection of dexmedetomidine (7.5 μg/kg), tiletamine/zolazepam (3 mg/kg), and butorphanol (0.15 mg/kg), while social cats received an IM injection of dexmedetomidine (7.5 μg/kg), ketamine (1.0 mg/kg), and butorphanol (0.1 mg/kg). After induction of anesthesia, the cats were shaved from xiphoid to pubis, the bladder was expressed, the eyes were lubricated, and an airway management device, which was typically a V-gel but also included endotracheal tube or, for kittens under 1 kg, a mask, was placed. The airway management device was not recorded. Intravenous fluids were not routinely provided, and any patient with a catheter would have been excluded. Patients were positioned on the surgical table in dorsal recumbency and connected to the nonrebreathing anesthetic system. Oxygen flow was set to 2 L/min and vapor setting at 1.5% isoflurane unless otherwise indicated by the patient’s depth of anesthesia. Truncal warming was provided via a conductive fabric blanket (HotDog; Augustine Surgical Inc) set to 104 °F (40 °C) and placed underneath the towel holding the patient or forced-air warming. A dedicated anesthetist monitored anesthesia via capnography and pulse oximetry.
Surgeries were performed by veterinary students under veterinary supervision or by attending veterinarians. The typical surgical technique was as follows: a 1-cm incision was made on the ventral midline, midway between the umbilicus and pubis; a spay hook was used to exteriorize the uterus; autoligation was performed on the ovaries and a single miller’s knot was used to ligate the uterine body; and the body wall was closed with a cruciate pattern and the skin with a purse string pattern with absorbable monofilament suture. All patients were tattooed on their ventral abdomen, and those through the TNR program received an ear-tipping (distal 1 cm) to recognize them as sterilized.15
After the surgery, cats were moved into recovery. Those with postoperative temperatures lower than 98 °F were eligible to receive truncal warming (HotDog; Augustine Surgical Inc) at the discretion of the attending veterinarian, who was not part of the research team and was blinded to the provision of heat support as a research outcome, and on the basis of the availability of warming devices. Patients in both programs were routinely reversed with atipamezole (0.04 mg/kg, IM) for a more consistent recovery (extralabel use). Once cats were observed to be responsive to stimuli, they were placed back into their trap or carrier.
Statistical analysis
Data were analyzed for normality with tests of skewness and kurtosis. Normal data were reported as mean and SD and non-normal data as median and IQR expressed as quartile 1 and quartile 3. Mixed-effects multivariable linear regression was used to compare the insulating properties of materials (as an interaction term between material and time) while accounting for ambient and starting temperatures. Mixed-effects multivariable linear regression was used to determine the effect of active warming and passive insulation on the rate of temperature decrease, controlling for weight, temperature at induction, and ambient temperature. Clinic dates, which accounted for interclinic variation, individual temperature-monitoring devices, and individual cats, were considered for random effects. Temperature decrease over the first hour was curvilinear, so temperature decrease (as an interaction term between treatment condition and time) was modeled separately for before and after 30 minutes. Linear regression with robust errors and controlling for anesthesia duration, temperature at induction, and ambient temperature was used to model the lowest recorded temperature. Body condition score, dose of anesthetic agent (milliliters per kilogram), incision size, prep time, temperature probe placement, and individual monitor were additionally considered for both models. Models were built with a combination of backward selection and prior literature and competing models compared with the Akaike information criterion and Bayesian information criterion. Validation of models was by inspection of the residuals for normality and specification link test for the linear regression model. Commercial statistical software (Stata, version 18; StataCorp LLC) was used for all analyses. Sample size calculation was based on an α of 0.05, power of 0.8, difference of 1 °F, and SD of 1 °F. A P value < .05 was considered significant.
Results
In vitro study
Two layers of down blanket and 1 layer of down blanket plus 2 layers of bubble wrap were the most effective insulators (Supplementary Table S1), decreasing the rate of heat loss between 68% and 76% compared to the uncovered control. Two layers of down were selected for the clinical study on the basis of durability and efficiency.
Multi-arm, parallel-group randomized controlled trial
Fifty-one cats were enrolled, with one cat excluded after undergoing an enucleation in addition to ovariohysterectomy and another cat excluded for being a neutered male. For those involved in the study, the median age was 8 months (IQR, 4 to 12 months), mean weight was 2.3 kg (SD, 0.8), and mean postinduction temperature was 100.6 °F (SD, 1.3). Values for these parameters were similar across the treatment groups (Table 1). No adverse events, such as burned paw pads, were observed.
Values for the control, passive, and active groups.
| Variable | Control (n = 16) | Passive (n = 17) | Active (n = 16) | P value |
|---|---|---|---|---|
| Estimated age (mo) | 12 (8–24) | 7 (4–12) | 6 (4–12) | .207 |
| Weight (kg) | 2.4 ± 0.8 | 2.4 ± 0.9 | 2.2 ± 0.8 | .441 |
| BCS (1–9) | 4 (4–5) | 5 (4–5) | 4 (4–5) | .393 |
| Social | 2 (13%) | 2 (12%) | 3 (19%) | .363 |
| Esophageal probe | 12 (75%) | 12 (71%) | 14 (88%) | .575 |
| Duration of surgery (min) | 44 (31–48) | 42 (35–52) | 49 (43–64) | .056 |
| Duration of anesthesia (min) | 72 (64–81) | 77 (69–94) | 87 (78–103) | .054 |
| Incision size (cm) | 1.5 (1–2) | 1.5 (1–2) | 1.3 (1–2) | .924 |
| Ambient temperature (°F) | 70.2 (69.7–70.3) | 69.8 (69.3–70.5) | 70.2 (68.7–71.6) | .900 |
| Postinduction temperature (°F) | 100.5 ± 1.5 | 100.4 ± 1.4 | 101.0 ± 0.9 | .377 |
| Lowest temperature (°F) | 95.2 ± 1.5 | 96.1 ± 1.7 | 96.8 ± 2.0 | .031 |
Data are presented as mean ± SD, median (IQR expressed as quartile 1 and quartile 3), or n (%). P values were determined via ANOVA or Kruskal-Wallis test. No P value for the baseline values was significant, while there was a P value < .05 for the lowest temperature.
BCS = Body condition score.
The decrease in temperature was curvilinear (Figure 2), with an inflection point around 30 minutes after the coverings and temperature monitor were placed. Temperature for all treatment conditions decreased rapidly for the first 30 minutes (T < 30), with controls decreasing by 0.12 °F (95% CI, 0.13 to 0.12), passive by 0.11 °F (95% CI, 0.11 to 0.10), and active by 0.09 °F (95% CI, 0.10 to 0.09) per minute (all P < .001). After 30 minutes (T < 30), the rate of decrease slowed for all treatment conditions, with controls decreasing by 0.05 °F (95% CI, 0.06 to 0.05), passive by 0.03 °F (95% CI, 0.04 to 0.03), and active by 0.01 °F (95% CI, 0.01 to 0.00). There was no overlap in 95% CI for any of the treatment conditions. Weight, ambient temperature, and induction temperature were also significant in the T < 30 and T ≥ 30 models (Table 2). Date and individual temperature-monitoring device were considered for random effects; while both had a statistically significant likelihood ratio test, there was no clinically meaningful effect, with both having extremely low intraclass correlations (1.42 X 10–14 and 6.15 X 10–10, respectively) and no effect on the other model coefficients at the level of precision reported here. Only individual cat was retained as a random effect (interclass correlation of .6).
Scatterplot of temperature readings for the control, passive, and active groups over time (minutes) overlaid with Lowess lines for each condition. The vertical dashed line at 30 minutes indicates an approximate inflection point. Lowess was constrained to 60 minutes as there were comparatively few data points after 60 minutes, and minutes were constrained to 100 as only 1 patient exceeded 100 minutes.
Citation: Journal of the American Veterinary Medical Association 2025; 10.2460/javma.25.02.0095
Results from multivariable mixed-effects linear regression with cat as random effect for time < 30 minutes, time ≥ 30 minutes, and all time points.
| Variable | Coefficient | P value | 95% CI |
|---|---|---|---|
| Time < 30 min | |||
| Treatment (min) | |||
| Control | –0.12 | < .001 | –0.13 to –0.12 |
| Passive | –0.11 | < .001 | –0.11 to –0.10 |
| Active | –0.09 | < .001 | –0.10 to –0.09 |
| Weight (kg) | 0.72 | < .001 | 0.43 to 1.01 |
| Induction temperature (°F) | 0.86 | < .001 | 0.66 to 1.05 |
| Ambient temperature (°F) | 0.32 | < .001 | 0.20 to 0.43 |
| Time ≥ 30 min | |||
| Treatment (min) | |||
| Control | –0.05 | < .001 | –0.06 to –0.05 |
| Passive | –0.03 | < .001 | –0.04 to –0.03 |
| Active | –0.01 | < .001 | –0.01 to 0.00 |
| Weight (kg) | 0.54 | .006 | 0.16 to 0.93 |
| Induction temperature (°F) | 0.84 | < .001 | 0.59 to 1.09 |
| Ambient temperature (°F) | 0.29 | < .001 | 0.13 to 0.45 |
| All time points | |||
| Treatment (min) | |||
| Control | –0.08 | < .001 | –0.08 to –0.08 |
| Passive | –0.06 | < .001 | –0.06 to –0.06 |
| Active | –0.04 | < .001 | –0.04 to –0.03 |
| Weight (kg) | 0.61 | < .001 | 0.31 to 0.92 |
| Postinduction temperature (°F) | 0.85 | < .001 | 0.66 to 1.05 |
| Ambient temperature (°F) | 0.31 | < .001 | 0.19 to 0.44 |
Coefficients represent a °F change per unit change in the respective variable. Treatment represents an interaction term between treatment group and minute, so the coefficient represents the change in temperature per minute for each treatment group.
The lowest recorded temperature varied by treatment group (Figure 3), with a median lowest temperature of 94.9 °F (IQR, 93.8 to 96.6 °F), 96.1 °F (IQR, 95.4 to 96.8 °F), and 96.6 °F (IQR, 95.9 to 98.7 °F) for controls, passive, and active, respectively. A multivariable model (Table 3) was used to predict the lowest recorded temperature for the treatment groups as compared with controls, with the passive and active groups predicted to have a final temperature significantly warmer: 1.2 °F (95% CI, 0.5 to 1.9; P = .001) and 1.9 °F (95% CI, 0.8 to 2.9; P = .001), respectively. Induction temperature (β = 0.60; P = .002; 95% CI, 0.24 to 0.96), ambient temperature (β = 0.40; P < .001; 95% CI, 0.25 to 0.55), and duration of anesthesia (β = −0.02; P = .014; 95% CI, −0.04 to 0.01) also contributed to the lowest recorded temperature. While weight (β = 0.49; P = .090; 95% CI, –0.08 to 1.05) was not significant, its inclusion in the model improved model fit as measured by the Akaike information criterion and Bayesian information criterion. Active warming was not significantly different (P = .183) from passive warming when the referent was changed to passive warming.
Box-and-whisker plot of the lowest recorded temperature for the control, passive, and active treatment groups.
Citation: Journal of the American Veterinary Medical Association 2025; 10.2460/javma.25.02.0095
Results from multivariable linear regression predicting the lowest recorded temperature.
| Variable | Coefficient | P value | 95% CI |
|---|---|---|---|
| Treatment (°F) | |||
| Control | Baseline | ||
| Passive | 1.2 | .001 | 0.5 to 1.9 |
| Active | 1.9 | .001 | 0.8 to 2.9 |
| Anesthesia duration (min) | –0.02 | .014 | –0.04 to 0.01 |
| Postinduction temperature (°F) | 0.60 | .002 | 0.24 to 0.96 |
| Ambient temperature (°F) | 0.40 | < .001 | 0.25 to 0.55 |
| Weight (kg) | 0.49 | .090 | –0.08 to 1.05 |
Although weight was not significant, its inclusion improved model fit as determined by both the Akaike information criterion and Bayesian information criterion and was retained. Coefficients represent a °F change per unit change in the respective variable. Coefficient for treatment is the contribution of the treatment toward the lowest temperature compared to the control.
Of the 49 cats, 46 had information regarding heat support in recovery. For these 46 cats, 37 (80%) were provided heat support. This differed by group, with 14 of 14 control (100%), 14 of 16 passive (88%), and 9 of 16 active cats (56%) provided heat support. In 2-sided tests of proportions, fewer cats in the active condition were provided heat support compared to control and passive conditions (P = .005 and P = .049, respectively). The proportion of cats provided heat support in the passive group was not different from controls (P = .171).
Discussion
Anesthetic-induced hypothermia is a common and clinically significant challenge in feline patients. This study aimed to evaluate whether highly insulating materials, used as passive insulation or combined with active warming, could reduce the rate of core temperature decline in cats undergoing anesthesia. The best insulating material—2 layers of down blanket—was determined by an initial in vitro study. The findings from the randomized controlled trial demonstrated that both passive insulation and active warming with 2 layers of down blanket significantly reduced core temperature decrease (the 95% CIs for the rate of change did not overlap for any treatment condition) compared to controls, with active warming providing the greatest benefit. The lowest recorded temperature for the passive and active groups was estimated to be over 1 °F and nearly 2 °F higher than controls, respectively. While passive insulation alone did not perform as well as active warming, it still offered a clinically meaningful reduction in heat loss and was not statistically different than active warming at the anesthetic durations observed here.
A previous study9 comparing cotton toddler socks with and without active warming found that passive insulation with cotton toddler socks alone was no different than controls, but active warming resulted in a slower rate of temperature decrease and higher temperature in recovery. While the difference for active warming via cotton toddler socks with a heating element was statistically significant, the effect (a difference of 0.54 °F from controls) was deemed not clinically meaningful. Use of better insulating materials in the current study resulted in passive insulation becoming both statistically and clinically significant. In addition, active warming had nearly 4 times the effect (1.9 vs 0.54 °F) and there was a statistically significant decrease in the proportion of cats provided rescue heat support in the active group. The application of heat support was prioritized on the basis of the relative temperatures of cats in recovery at the same time, as there were not always enough resistive mats to provide heat support to all cats.
In the prior study,9 the rate of temperature decrease was modeled linearly for the entire duration of anesthesia despite a modest inflection point being visible in the data. The multivariable model for the rate of temperature decrease was similar, although ambient temperature was significant in this study when it was not in the previous study. This discrepancy may be due to the greater number of data points in the present study, as temperature was measured every 1 minute rather than every 5 minutes, or the slightly higher median ambient temperature observed here, which was 70 °F (IQR, 69 to 71 °F), as compared to 68 °F (IQR, 67 to 69 °F) in the previous study. Ambient temperature has been associated with perioperative hypothermia in high-quality, high-volume environments.16 Neither study found that BCS was significant, unlike a previous study17 comparing temperature differences between different surgical scrub protocols, although the number of cats with a BCS < 4 was very low in this study (6 of 49 [12%]).
The coefficients for the rate of temperature decrease of controls differed between this study and the previous study. Here, the overall rate for control was 0.08 °F/min while previously it was 0.04 °F/min. Coefficients for weight and postinduction temperature were also modestly different, with both variables exerting greater influence on the rate of temperature decrease than previously (0.61 and 0.85 vs 0.30 and 0.39, respectively). This study also examined the lowest temperature rather than the last (recovery) temperature as a secondary outcome. In the previous study,9 last and lowest temperatures were typically identical, as asocial cats must be returned to their trap as soon as they are responsive to stimuli before their temperatures begin to increase, but social cats with low temperatures may be kept in recovery to allow their temperatures to rise.
There were small differences in the multivariable models for lowest as compared to recovery temperature. In the lowest temperature model, postinduction temperature exerted slightly greater influence (0.60 vs 0.42 °F). Ambient temperature was significant, but weight was not, although it was retained in the model as it increased model fit. The coefficient for weight (0.49 vs 0.42 °F) and anesthesia duration (–0.025 vs −0.015 °F) were similar between models. Notable differences in this study population and design beyond the intervention and secondary outcome measure between this study and the previous study9 include longer surgical duration (45 minutes [IQR, 36 to 52 minutes] vs 30 minutes [IQR, 20 to 37 minutes]), 2 °F warmer median ambient temperature, less variation in temperature-monitoring devices, the inclusion of social cats (7 of 49), and a constrained date range of June and July as compared to March through September. However, longer surgical duration should not have modified the rate of change so dramatically, given that the greatest amount of change happens within the first 30 minutes of monitoring, and a higher ambient temperature should have led to a lower temperature gradient, slowing the rate of change.
In a study12 that looked at 2 layers of bubble wrap and an absorbent pad on the extremities and thorax of cats, there was an approximate decrease of 0.12 °F/min for controls and 0.07 °F/min for cats passively insulated with bubble wrap during the first hour after induction, equating to a relative difference of 52% between control and passive rates of change. For this study, there was a decrease of 0.09 °F/min for controls and 0.07 °F/min for the passive group when considering just the first hour, equating to a relative difference in the rate of change of 25%. Our in vitro study noted that 2 layers of bubble wrap and 1 layer of down blanket performed similarly to 2 layers of down blanket, but it is unclear how 1 layer of absorbent pad compares to 1 layer of down. The major differences between these 2 studies, besides the insulating material being bubble wrap, were that this study did not cover any portion of the thorax and the first hour of anesthesia included a large portion of surgical time.
The safety of the active warming device was not tested for anesthetic times exceeding 90 minutes. Patients in this study had all 4 limbs covered, which may not be possible when catheters or blood pressure cuffs are in place or when patients are undergoing orthopedic procedures affecting the limbs. The benefit of covering fewer than 4 legs was not tested. The benefit of covering the thorax in addition to the extremities was also not tested. Neither the researchers nor the nonresearcher veterinarian in recovery could be blinded to treatment condition. However, the primary outcome measure—temperature—was objective and the veterinarian in recovery was not aware of the provision of heat support as a secondary outcome measure. Patients were induced with a drug combination containing dexmedetomidine, which causes peripheral vasoconstriction.18 This vasoconstriction may alter the performance of the devices, potentially decreasing their performance if the vasoconstriction prevents transfer of external heat to the extremities. The devices may have different effects if used on patients receiving other drug combinations. The coverings and temperature-monitoring devices could only be placed after cats became unresponsive to stimuli, which was a median of 14 minutes (IQR, 12 to 23 minutes) after IM induction of anesthesia. This accounts for 20% to 30% of the period of phase 1 hypothermia, decreasing the utility of warming the extremities. While prewarming would be ideal, cats are unlikely to tolerate the coverings when conscious. Only phase 1 and phase 2 hypothermia could be investigated here because anesthesia did not exceed 2 hours. The rate of temperature decrease was significantly higher in this study than in a similarly designed study.9 This may be because different brands of thermometers and different probe placement locations (rectal vs a mix of rectal and esophageal) were used and thermometers were not calibrated in either study. However, there was no systematic bias in the assignment of thermometer brand in the first study or probe placement in this study, so the relative difference from controls should be reliable. Probe placement location was not significant in the regression models. The type of airway management was not recorded, although there should not have been systematic bias in the assignment.
This study demonstrated that the use of highly insulating materials effectively reduces the rate of core temperature decline in anesthetized cats, with both passive insulation and active warming offering significant benefits compared to uncovered controls. While active warming provided the greatest reduction in the rate of temperature decrease, passive insulation alone was also highly effective and represents a practical, economical option for veterinary clinics. These findings demonstrate that peripheral passive insulation or active warming can be a relevant adjunctive strategy for combating anesthetic-induced hypothermia if highly insulating materials are used.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
None reported.
Disclosures
The authors have nothing to disclose. No AI-assisted technologies were used in the composition of this manuscript.
Funding
The authors have nothing to disclose.
ORCID
Rachael Kreisler https://orcid.org/0000-0002-5562-5521
References
- 1.↑
Matsukawa T, Sessler DI, Sessler AM, et al. Heat flow and distribution during induction of general anesthesia. Anesthesiology. 1995;82(3):662-673. doi:10.1097/00000542-199503000-00008
- 2.↑
Kurz A, Sessler DI, Christensen R, Dechert M. Heat balance and distribution during the core-temperature plateau in anesthetized humans. Anesthesiology. 1995;83(3):491-499. doi:10.1097/00000542-199509000-00007
- 3.↑
Redondo JI, Suesta P, Gil L, Soler G, Serra I, Soler C. Retrospective study of the prevalence of postanaesthetic hypothermia in cats. Vet Rec. 2012;170(8):206. doi:10.1136/vr.100184
- 4.↑
Pottie RG, Dart CM, Perkins NR, Hodgson DR. Effect of hypothermia on recovery from general anaesthesia in the dog. Aust Vet J. 2007;85(4):158-162. doi:10.1111/j.1751-0813.2007.00128.x
- 5.↑
Satas S, Haaland K, Thoresen M, Steen PA. MAC for halothane and isoflurane during normothermia and hypothermia in the newborn piglet. Acta Anaesthesiol Scand. 1996;40(4):452-456. doi:10.1111/j.1399-6576.1996.tb04468.x
- 6.↑
Frank SM, Fleisher LA, Breslow MJ, et al. Perioperative maintenance of normothermia reduces the incidence of morbid cardiac events: a randomized clinical trial. JAMA. 1997;277(14):1127-1134.
- 7.↑
Beal MW, Brown DC, Shofer FS. The effects of perioperative hypothermia and the duration of anesthesia on postoperative wound infection rate in clean wounds: a retrospective study. Vet Surg. 2000;29(2):123-127. doi:10.1111/j.1532-950X.2000.00123.x
- 8.↑
Öner Cengiz H, Uçar S, Yilmaz M. The role of perioperative hypothermia in the development of surgical site infection: a systematic review. AORN J. 2021;113(3):265-275. doi:10.1002/aorn.13327
- 9.↑
Kreisler RE, MacDonald LJ, Mensing RN, et al. Effects of peripheral active warming and passive insulation on core body temperature during feline ovariohysterectomy: a multi-arm randomized clinical trial. J Feline Med Surg. 2023;25(3):X231157585. doi:10.1177/1098612X231157585
- 10.↑
Onozawa E, Azakami D, Seki S, Hamamoto Y, Ishioka K. Effect of an insulation device in preventing hypothermia during magnetic resonance imaging examinations for dogs and cats under general anesthesia. Animals (Basel). 2021;11(8):2378. doi:10.3390/ani11082378
- 11.
Tünsmeyer J, Bojarski I, Nolte I, Kramer S. Intraoperative use of a reflective blanket (Sirius rescue sheet) for temperature management in dogs less than 10 kg. J Small Anim Pract. 2009;50(7):350-355. doi:10.1111/j.1748-5827.2009.00758.x
- 12.↑
Sakata H, Walsh V, Chambers JP, Bridges J, Sano H. Effect of insulation with bubble wrap and an absorbent pad on heat loss in anaesthetised cats. N Z Vet J. 2020;68(6):324-330. doi:10.1080/00480169.2020.1776649
- 13.↑
O’Neil BA, Linklater AKJ. Supplemental reflective blankets and wool socks help maintain body temperature in dogs undergoing celiotomy procedures: a prospective randomized controlled clinical trial. J Am Vet Med Assoc. 2022;260(11):1316-1323. doi:10.2460/javma.22.01.0001
- 14.↑
Sargeant JM, Ruple A, Selmic LE, O’Connor AM. The standards of reporting trials in pets (PetSORT): explanation and elaboration. Front Vet Sci. 2023;10:1137781. doi:10.3389/fvets.2023.1137781
- 15.↑
Dalrymple AM, MacDonald LJ, Kreisler RE. Ear-tipping practices for identification of cats sterilized in trap-neuter-return programs in the USA. J Feline Med Surg. 2022;24(10):e302-e309. doi:10.1177/1098612X221105843
- 16.↑
Rodriguez-Diaz J, Hayes G, Appel L, Buote N, Moyal M. Incidence of perioperative hypothermia in a high-quality high-volume spay/neuter setting and association with environmental temperature. J Shelter Med Community Anim Health. Published online June 16, 2023. doi:10.56771/jsmcah.v2.27
- 17.↑
Kreisler RE, Douglas ML, Harder KN. Comparison of the effect of isopropyl alcohol and chlorhexidine solution rinses on body temperature of female cats undergoing sterilization surgery. J Feline Med Surg. 2021;23(10):875-882. doi:10.1177/1098612X20979565
- 18.↑
Pypendop BH, Honkavaara J, Ilkiw JE. Pharmacokinetics of dexmedetomidine, MK-467 and their combination following intramuscular administration in male cats. Vet Anaesth Analg. 2017;44(4):823-831. doi:10.1016/j.vaa.2017.02.008
