Homeotherms maintain a constant body temperature despite variations in their thermal environment. This process is critical because decreases in core body temperature influence almost every organ system, including the cardiovascular, nervous, endocrine, digestive, and urinary systems.1 Hypothermia may develop as a consequence of various triggers.2–5 In neonates, hypothermia develops as a consequence of multiple routes of heat loss (ie, evaporative, convective, and conductive).6–9 In trauma patients, factors associated with hypothermia include extreme age, environmental stresses, impaired neurologic state or mobility, and delayed care by health services.10 In anesthetized human patients and dogs, factors associated with development of hypothermia are poor clinical condition, high body surface-to-volume ratio, extent of surgical procedure, and duration of surgery.11,12
The etiology of hypothermia in most situations is not understood.13,14 Nevertheless, in human medicine, efforts have been made to determine the extent to which body temperature affects important outcomes. Nontherapeutic hypothermia, that is, hypothermia not induced by medical actions, has been reported to be associated with increased mortality rates for several categories of human patients, including neonates,15,16 trauma patients,13 patients hospitalized for blunt trauma,17 patients with congestive heart failure,18 patients with burns,19 and patients undergoing ruptured aneurysm repair.20 Therefore, body temperature may provide an insight into the prognosis for these patients. Furthermore, body temperature control may have a therapeutic role, as there is some evidence that prompt correction of hypothermia may decrease mortality risk in human patients.21 This is supported by results of an experimental study22 involving rats, which found that warming during hemorrhage may prevent exacerbation of hypothermia and hypotension and improve survival.
Rabbits are among the most common pet animals in Europe and the United States.23 Clinicians treating rabbits should be familiar with several factors unique to these animals including hind-gut fermentation, unusual calcium metabolism, very small thoracic size compared with body mass, relatively high metabolic rate, and the fact that rabbits are a prey animal that succumbs to stress easily.24 Because of these physiological peculiarities, rabbits are considered more vulnerable, compared with dogs and cats, in several scenarios, including the perioperative setting.25,26
Rectal temperature of rabbits can be affected by multiple factors; because a reference range has not been established, empirical values of 38.0° to 38.5°C (100.4° to 101.3°F) and 38.0° to 40.0°C (100.4° to 104°F) have been considered normal.27,28 Lower body surface area-to-volume ratio is an important protective factor against core hypothermia in humans11 and dogs.12 Because of their high surface area-to-volume ratio, rabbits are predisposed to suffer rapid changes in body temperature.29,30 Therefore, rabbits may be even more predisposed to hypothermia than other mammals. Considering that authors of previous reports have speculated that obtaining rectal temperature in client-owned rabbits may have little clinical value,27,30 there is an urgent need to evaluate the clinical role of body temperature in hospitalized rabbits.
According to recommendations to reduce research waste,31 we conducted a systematic review of the literature to identify any study that associated body temperature with mortality in pet (ie, client-owned) rabbits and retrieved no relevant studies (Online Supplement eTable 1 available at http://avmajournals.avma.org/toc/javma/248/3). Although some authors suggest that sick rabbits have a low body temperature,30 there is currently no published evidence indicating that body temperature is associated with mortality or poorer clinical outcome in client-owned rabbits. Furthermore, we found no published studies reporting the body temperature of pet rabbits admitted to veterinary hospitals.
The purpose of the study reported here was to investigate whether rectal temperature at hospital admission was associated with mortality risk in client-owned rabbits. Our specific hypothesis was that rabbits admitted with hypothermia would have a higher risk of death, compared with risk of death for normothermic rabbits.
Materials and Methods
A prospective cohort study was planned. Data on each rabbit were collected specifically to answer the study question as part of an ongoing survey of rabbit morbidity and mortality initiated in 2011. The primary outcome of interest was the presence of an association between rectal temperature and all-cause mortality (defined as death before or within 1 week after hospital discharge). Secondary outcomes were the effects of sex, neuter status, weight, gastrointestinal stasis, health status at admission, and anesthesia on the association between hypothermia and mortality, and the association of these factors with mortality. Data gathering was performed in compliance with Directive 2010/63/EU of the European Parliament and of the Council of September 22, 2010, on the Protection of Animals Used for Scientific Purposes. All owners provided informed consent for inclusion of data collected from their animals in the study. The local ethics committee revised and approved the protocol. The manuscript was reported in compliance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guidelines.32 The STROBE checklist is provided (Online Supplement eTable 2 available at http://avmajournals.avma.org/toc/javma/248/3).
Sample size calculation
A sample size was calculated to achieve 90% power to detect a difference at a significance level of 5% for the primary outcome measure, that is, risk of mortality for rabbits with versus without hypothermia. On the basis of a pilot sample of data, we expected a 10% mortality rate in hospitalized rabbits without hypothermia and a 30% mortality rate in hospitalized rabbits with hypothermia. The following formula for comparison of proportions was used:
where u = 1.28, v = 1.96, p1 = first proportion, p0 = second proportion, and p’ = mean of the proportions = (p1 + p0)/2.
An adjustment for unequal sample sizes was performed, assuming a ratio of hypothermic to nonhypothermic rabbits of 1:3.33 Therefore, minimum sample size was calculated to be 55 individuals for the hypothermic group and 164 individuals for the nonhypothermic group.
Recruitment
Client-owned rabbits that were hospitalized at the Clinica per Animali Esotici (Rome, Italy) between January 1, 2013, and January 31, 2014 (excluding periods from February 15 to February 28, May 7 to May 31, and July 1 to July 14), were eligible for enrollment in the study. All client-owned rabbits that were hospitalized during this period (including day-long hospitalization) were included in the study and followed up during their hospitalization and after discharge. Hospital admission for inpatient care was suggested by the treating clinician for the owners of all rabbits initially examined for signs of systemic illness or that required surgical procedures (minor and major). All owners signed an informed consent form prior to hospitalization. Without informed consent, rabbits were not hospitalized and were not included in the study. Rabbits that were not hospitalized (eg, owners refused hospitalization, or rabbits were immediately euthanized after examination) were excluded from the study. Rabbits that were discharged from the hospital or that died before a clinician could perform the physical examination at admission were also excluded.
Data collection
The following variables were retrieved for each patient at admission by means of a study-specific single-patient form: rectal temperature (degrees Celsius), age (months), sex (male or female), neuter status (sexually intact or neutered), body weight (g), month of admission, clinical history, and whether the rabbit had defecated in the previous 12 hours. The following variables were retrieved during the hospitalization period: suspected presence of systemic diseases, whether the rabbit underwent anesthesia, and whether the rabbit died during hospitalization or in the week following discharge. To avoid data dredging and consequent spurious results, all variables recorded for the study were specified a priori because of their presumed or proven association with mortality in rabbits or other species.
Rectal temperature—At the time of the study, rabbits were typically hospitalized for intensive care, surgical procedures, or both. All hospitalized rabbits underwent a physical examination at admission that included, among other things, measurement of rectal temperature. Temperature was measured with a digital thermometera after rabbits were acclimatized to the recovery room. Briefly, an experienced operator held the rabbit cradled on its back with the hind end supported, and a second operator inserted the thermometer into the rectum to a distance of approximately 3 cm.34 The thermometer required 60 to 120 seconds for temperature reading and provided results in degrees Celsius to 1 decimal place, with an acoustic signal at the beginning and the end of the process. Appropriate care was paid to ensure that the edge of the thermometer contacted the rectal mucosa during temperature measurement.
The first temperature obtained at admission for each rabbit hospitalized in the clinic was recorded as the admission temperature. When multiple temperature measurements were available, only the admission temperature was used in the analysis, as previously reported for a study of the US National Trauma Data Bank.14 The analytic range of the thermometers used was 32.0° to 42.0°C (89.6° to 107.6°F). If the temperature was lower, the low limit of measurement, 32.0°C (89.6°F), was used for the purposes of analysis. For clinical purposes, a rectal temperature ≥ 38.0° and ≤ 39.9°C (≥ 100.4° and ≤ 103.8°F) was considered normal, on the basis of expert recommendations27 and published data.35 Hypothermia was defined as a rectal temperature ≤ 37.9°C (≤ 100.2°F). All rabbits with an initial rectal temperature ≤ 37.9°C were treated on the basis of current published recommendations.36,37 These rabbits were placed in an incubator,36,b and active rewarming was attempted via administration of warmed IV fluids (as clinically indicated) and assisted feeding of warmed ground food.37,c
Other variables—Body weight was measured with an infant digital scale.d To include a predictor that accounted for the severity of the condition of the patient, each rabbit at admission was categorized as “with suspicion of systemic disease” or “without suspicion of systemic disease” by the treating clinician on the basis of clinical findings. Rabbits with dental disease were categorized as “with suspicion of systemic disease” only if they had signs of altered consciousness (eg, obtundation or lethargy) or if dental disease was associated with other disease states.
Rabbit gastrointestinal syndrome is a common complex of clinical signs and concurrent pathological conditions affecting the digestive system of rabbits.38 The most common clinical sign of rabbit gastrointestinal syndrome is gastrointestinal stasis, which is characterized by decreased fecal output, anorexia, and lethargy.39 For the present study, as a proxy for gastrointestinal stasis we recorded whether the owner reported lack of defecation in the 12 hours before admission. This information was obtained by the treating clinician asking a standard question of the owner (“Did you see any feces produced from the rabbit in the last 12 hours?”) during the initial examination.
All rabbits that underwent sedation or anesthesia were classified as undergoing anesthesia. In the present cohort, a small number of rabbits underwent sedation for a diagnostic procedure, whereas the majority underwent sedation for various therapeutic or elective procedures, such as dental treatment, spay, castration, and other soft-tissue and bone surgeries. When hypothermia was present, elective surgery was discouraged.
The owners of rabbits that were alive at discharge were contacted via telephone by 1 author (GT) to determine whether rabbits died in the 7 days following discharge. Usually multiple (2 to 3) telephone numbers were obtained at hospital admission for each rabbit. Owners were contacted several times to obtain the required follow-up information. When it was not possible to contact the owners after 5 attempts at each telephone number or the owner was unwilling to provide follow-up information, the rabbit was excluded from the study.
Statistical analysis
Continuous variables were summarized as either mean and SD or median and range depending on their distribution. The Shapiro-Wilk test was used to evaluate continuous data for normality. For categorical variables, the percentages of patients in each category were calculated.
Rectal temperature at admission was analyzed as a continuous variable to calculate whether there was an increase in odds of death in rabbits depending on the rectal temperature and as a dichotomous variable (hypothermia vs nonhypothermia) for calculation of relative risk and significance testing by means of the χ2 test or the Fisher exact test as appropriate. Univariate binary logistic regression analyses with 1 predictor were used to identify any other relevant predictor of death and to identify an association between month of admission and hypothermia. Predictors tested were signalment factors (age, sex [male or female], neuter status [sexually intact or neutered], and weight), month of admission, presence of gastrointestinal stasis, exposure to anesthesia, and suspected presence of systemic disease. Odds ratios and 95% CIs were used to quantify the strength of these associations. The original calculations were all performed with temperature in degrees Celsius.
Multivariate logistic regression models were subsequently developed to evaluate whether the association between hypothermia and mortality persisted after adjustment for other patient characteristics and potential confounders. Patient characteristics and covariates demonstrated in the univariate analysis to be prognostically significant (at a significance level of P < 0.05) were entered into the final model. Variables in the final model were entered by means of a backward conditional method with 0.5 classification cutoff and 20 maximum iterations. Variables were retained in the model if the score statistic was less than 0.05 and removed if greater than 0.10. To avoid overfitting the model, no more than 1 predictor variable was included for 10 events (ie, deaths).40
The Hosmer-Lemeshow statistic was used to assess goodness of fit of the model. The Nagelkerke R2 was reported as a measure of predictive power of the model. The linearity assumption of the variable temperature was assessed as described elsewhere.41 In brief, temperature was binned in quartiles, the multivariable model was fitted replacing the continuous temperature variable with the 4-level binned temperature variable, log odds of the upper 3 quartiles (the lower quartile was used as indicator) were plotted versus the respective quartile midpoints, the 4 plotted points were connected with straight lines, and the plot was visually inspected for linearity.
Sensitivity analyses—Sensitivity analyses were conducted to support the robustness and generalizability of the findings and to minimize the possibility of residual confounding. First, all rabbits < 1 year of age were excluded from the cohort because it could be argued that the association of hypothermia and death was present only in young rabbits. Second, the analysis was repeated excluding small rabbits (ie, rabbits < 1,000 g) and rabbits of moderate size (< 1,500 g) because it is likely that lighter rabbits lose temperature quicker than heavier ones. Third, the analysis was repeated with different cutoff points to classify rabbits as having hypothermia (ie, < 37.0°C, < 37.5°C, and < 38.5°C). Fourth, the analysis was performed for male and female rabbits separately. Fifth, the association of hypothermia with mortality was examined in rabbits that were not suspected to have systemic disease.
Missing data—Although 6 of 9 variables had no missing data, for 3 signalment variables (body weight, age, and sex) there were 43 missing values (1.5% of the overall data). However, 276 of 316 patients had complete data. To evaluate how missing data impacted the use of the multivariate models, multiple imputation techniques were used.42 All predictors were used to impute missing values on the basis of multivariate normal distributions.
All analyses were performed and all figures were created with commercial software.e,f A 2-sided P value of 0.05 was considered to be significant. Raw data are freely available from the authors upon request.
Results
During the study period, 316 rabbits were hospitalized and the owners were successfully contacted to obtain 7 days of follow-up information, thus meeting the study criteria (Figure 1). Median rectal temperature of the rabbits at admission was 38.2°C (100.8°F; range, 32.0° to 41.6°C [89.6° to 106.9°F]). One hundred sixteen of the 316 (36.7%) rabbits had hypothermia at admission, 193 (61.1%) had normothermia (≥ 38.0° and ≤ 39.9°C [≥ 100.4° and ≤ 103.8°C]), and 7 (2.2%) had hyperthermia. Ninetyfive of the 316 (30.1%) rabbits died during hospitalization or in the 7 days after discharge, whereas 221 (69.9%) survived (Figure 2). Median rectal temperature at admission of the rabbits that died was 37.2°C (99.0°F; range, 32.0° to 40.2°C [89.6° to 104.4°F]). Median rectal temperature at admission of the rabbits that survived was 38.5°C (101.3°F; range, 35.0° to 41.6°C [95.0° to 106.9°F]). Rabbits with hypothermia had a significantly higher mortality risk, compared with normothermic and hyperthermic rabbits (OR, 5.41; 95% CI, 3.22 to 9.10). For each 1°C (1.8°F) decrease in admission rectal temperature, the odds of death were doubled (OR, 2.11; 95% CI, 1.69 to 2.64; Figure 3). Rabbits with hypothermia at admission had a risk of mortality 3 times the risk for rabbits without hypothermia (relative risk, 3.09; 95% CI, 2.17 to 4.39).
Study flow diagram for patients recruited and enrolled in a study investigating whether rectal temperature at hospital admission was associated with risk for mortality in client-owned rabbits. Deceased = Rabbits that died or were euthanized during hospitalization or within 7 days after discharge.
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.288
Distribution of rectal temperatures at hospital admission in a cohort of client-owned rabbits (n = 316).
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.288
Mortality rate in client-owned rabbits as a function of rectal temperature at hospital admission. For each 1°C (1.8°F) decrease in admission rectal temperature, the odds of death were doubled (OR, 2.11; 95% CI, 1.69 to 2.64).
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.288
In univariate analyses, age, suspected presence of a systemic disease, and presence of gastrointestinal stasis were associated with a significantly increased mortality rate, whereas undergoing anesthesia during hospitalization was associated with a significantly decreased mortality rate (Table 1). Month of admission was not significantly associated with the presence of hypothermia (OR, 1.02; 95% CI, 0.95 to 1.08).
Results of univariate analysis of factors to determine association with mortality (defined as death before or within 1 week after hospital discharge) in a cohort of client-owned rabbits.
| Variable | Alive | Dead | OR (95% CI) | P value |
|---|---|---|---|---|
| Rectal temperature at admission | ||||
| Hypothermic | 55 (47.4) | 61 (52.6) | 5.41 (3.22–9.09) | < 0.001 |
| Normo- or hyperthermic | 166 (83.0) | 34 (17.0) | 1 | |
| Age (mo) | 38.5 ± 32.8 | 51.5 ± 38.6 | 1.01 (1.00–1.01) | 0.003 |
| Weight (g) | 1,673 ± 476 | 1,634 ± 759 | 1.00 (0.99–1.00) | 0.75 |
| Sex | ||||
| Female | 99 (70.7) | 41 (29.3) | 0.96 (0.59–1.57) | 0.88 |
| Male | 121 (71.4) | 52 (28.6) | 1 | |
| Sexual status | ||||
| Intact | 154 (71) | 63 (29) | 0.85 (0.51–1.43) | 0.55 |
| Neutered | 67 (67.7) | 32 (32.3) | 1 | |
| Systemic disease | ||||
| Present | 117 (57.6) | 86 (42.4) | 8.49 (4.07–17.72) | < 0.001 |
| Absent | 104 (92) | 9 (8) | 1 | |
| Anesthesia | ||||
| Performed | 99 (83.2) | 20 (16.8) | 0.33 (0.19–0.57) | < 0.001 |
| Not performed | 122 (61.8) | 75 (38.2) | 1 | |
| Gastrointestinal stasis | ||||
| Present | 68 (61.8) | 42 (38.2) | 1.78 (1.08–2.92) | 0.021 |
| Absent | 153 (74.3) | 53 (25.7) | 1 | |
Continuous data are reported as mean ± SD. Binary data are reported as number of observed events (percentage over the total). A P value < 0.05 indicates a significant univariate association with mortality (2-tailed P value determined with a t test or χ2 test, as appropriate).
The final multivariate logistic regression model included age, presence of gastrointestinal stasis, exposure to anesthesia, and suspected presence of systemic disease. After multivariable adjustment, the association between hypothermia and mortality persisted, with hypothermic rabbits having odds of death 5 times the odds for rabbits without hypothermia (OR, 5.07; 95% CI, 2.81 to 9.13). For each 1°C (1.8°F) decrease in admission rectal temperature, the odds of death were doubled (OR, 1.90; 95% CI, 1.5 to 2.41). The final logistic regression model was well fitted (Hosmer-Lemeshow test; P = 0.88; Nagelkerke R2 = 0.31), accounting for the presence of systemic disease (OR, 10.94; 95% CI, 4.58 to 26.09) and of gastrointestinal stasis (OR, 2.2; 95% CI, 1.16 to 4.17). Because of the lack of inclusion of the other 3 variables in the final multivariate model, the multiple imputation analysis did not provide any benefit to the model.
Sensitivity analyses confirmed the robustness of the finding (Figure 4). Exclusion of young or small rabbits did not alter the study findings, with significantly increased odds of death observed even for rabbits weighing > 2.0 kg (4.4 lb). Increased odds of death were present also in hypothermic rabbits that were classified as not affected by systemic disease. The use of different cutoff points for defining hypothermia did not alter the study results. We did not detect rectal lesions consequent to rectal temperature measurement in any rabbit included in the study.
Sensitivity analysis of the association between hypothermia at hospital admission and mortality in a cohort of client-owned rabbits. In the graph, vertical marks represent ORs and horizontal lines represent 95% CIs.
Citation: Journal of the American Veterinary Medical Association 248, 3; 10.2460/javma.248.3.288
Discussion
For many years, the clinical value of measuring rectal temperature in client-owned rabbits has been criticized. More than a decade ago, a popular rabbit compendium reported, “The rectal temperature can be taken at the clinician's own judgment, as the findings are often not helpful in making a diagnosis.”30 The most recent compendium of rabbit medicine published in 2014 states that “[m]any practitioners do not routinely take the rectal temperature as part of their clinical examination because of the risk of trauma and the limitations in interpreting its significance.”27 On the basis of results of the present study, rectal temperature at hospital admission was found to be a relevant prognostic factor in client-owned rabbits. Even though the study did not have sufficient statisical power to evaluate the effect of hypothermia in healthy rabbits, we found that hypothermia was associated with increased mortality in client-owned rabbits that did not show clinical signs of systemic diseases (eg, patients undergoing elective surgery). This suggested that rectal temperature may also be a useful screening tool for healthy rabbits and therefore that rectal temperature should be measured during every physical examination.
Hypothermia affects multiple organ systems and physiologic processes.43 Understanding the mechanisms by which temperature is associated with an increased mortality risk in client-owned rabbits goes beyond the scope of the present study. A possible explanation for our findings is that hypothermia represents a response to stress44 or an early decompensatory stage of shock.45 With this hypothesis, hypothermia represents a proxy for disease severity. A further explanation is that hypothermia actively decreases the chances of survival. Hypothermia may cause depressed cardiac function, thus contributing to poor perfusion and myocardial irritability with secondary arrhythmias.46 The function of enzymes, such as those involved in cellular respiration, may decrease in hypothermic patients having implications throughout all organ systems.13,45
In the adjusted analysis performed in the present study, we found that 2 factors other than hypothermia were also associated with a significantly increased risk of mortality. The finding that rabbits suspected to have a systemic disease were more likely to die during hospitalization was intuitive. More interestingly, we observed that rabbits had not defecated in the 12 hours before admission had odds of dying twice the odds for other rabbits. Gastrointestinal stasis is a common problem in client-owned rabbits,39 and our finding supported the recommendation that supportive care for rabbits with signs of gastrointestinal stasis should be initiated promptly.38
Currently, several prognostic factors have been described for client-owned rabbits. In a large cohort of client-owned rabbits, it was reported that hyperglycemia was associated with poorer clinical outcomes.47 In particular, 2.6% of rabbits in a previous study47 of 907 rabbits (including 238 that were overtly healthy) had blood glucose concentrations > 20 mmol/L (360 mg/dL), and all of these had a critical illness. In a different cohort of client-owned rabbits, hyponatremia was associated with increased mortality.48 In that study48 of 356 ill rabbits, those with plasma sodium concentration < 129 mEq/L had a 2.3-fold increase in mortality risk. We suggest that future prospective studies should incorporate all of these prognostic variables.
Although it has been reported that risk of death during anesthesia in client-owned rabbits is higher than the risk for other species,26 we observed a protective effect of anesthesia on mortality in the present study. In this study, mortality rate for rabbits undergoing anesthesia was 16.8%, compared with a rate of 38.2% for rabbits that did not undergo anesthesia. We suggest that the apparent protective effect of anesthesia was probably secondary to the fact that anesthesia was commonly performed for patients undergoing elective procedures, whereas rabbits that were hospitalized without undergoing anesthesia had more severe diseases. In the previous report,25 which included a large multicenter cohort of rabbits (n = 8,209), risk of anesthetic death was 1.3%. In the present study, risk of death in rabbits undergoing anesthesia was more than 10-fold greater (16.8%). This difference may be secondary to several factors, including type of procedures performed under anesthesia, anesthetic protocol used, type of monitoring during anesthesia, and severity of disease state for the patients admitted. We suggest that the most likely reason for the difference in mortality rate between the 2 patient cohorts is the difference in follow-up. The present study was designed to detect mortality after hospitalization; therefore, follow-up information on the rabbits was collected at least 7 days after discharge, whereas the previous study26 focused on perioperative deaths with follow-up occurring 48 hours after surgery. Further multicenter studies with longer follow-up periods after elective procedures are suggested to better understand this phenomenon.
A main reason that rectal temperature measurement has been discouraged during clinical examination of client-owned rabbits is stress related to the procedure and the risk for development of rectal lesions.27 It has been suggested that the use of an auricular thermometer may mitigate such problems.36 In cats, auricular thermometry has been reported to be a reliable alternative to rectal thermometry for assessment of core body temperature.49 Unfortunately, in a previous study50 of rabbits, results of auricular thermometry agreed poorly with rectal temperature, thus making the measurements obtained with the 2 techniques not interchangeable. Given the lack of agreement between the 2 methods, it is unclear whether measurement of auricular temperature has clinical value in client-owned rabbits. Future studies are indicated to evaluate whether use of a correction factor for auricular thermometry in client-owned rabbits may facilitate application of this modality and to further evaluate the clinical use of auricular thermometry. Currently, clinicians should continue to measure rectal temperature in rabbits whenever possible.
Expert opinions and anecdotic reports suggest that hypothermia in rabbits should be treated.37 It remains to be determined whether aggressive or passive rewarming procedures are preferred.51 In neonates or depressed patients that may be hypoxemic or have reduced energy stores, the use of radiant heat sources may help maintain body temperature while allowing access to the patient.9 Other recommendations to prevent hypothermia in newborns include use of a warm delivery room, immediate drying, skin-to-skin contact with the mother, and use of an incubator.52 Administration of IV fluids may also play a role in rewarming, and a recent review suggested that all IV fluids should be warmed to body temperature prior to administration.37 An in vitro study53 from 2014 demonstrated that there was no benefit to prewarming of fluids with use of a distance-dependent veterinary-specific IV fluid warmer and that the fluid warmer should be placed close to the patient. Considering that rabbits usually require low IV fluid rates, the output temperature for the devices tested in that study may be suboptimal, and the authors suggested use of both active and passive heating devices for patients requiring low IV fluid flow rates.53 Nonetheless, we suggest that the provision of a warm ambient temperature (eg, incubators) and use of radiant heat devices, warmed IV fluids, and warmed food may be not sufficient to restore normothermia in sick rabbits, in view of the high mortality rate observed in our data set. As such, we suggest further study.
Enrollment of animals in the present study was performed in a 24-hour, exotics-only animal clinic that provided emergency, primary, and specialist care to client-owned rabbits. Performing the same study in a different setting (eg, mixed practice or primary careonly practice) may result in substantially different prevalences of hypothermic rabbits and overall rabbit mortality rate. However, we suggest that if similar strategies to treat hypothermia are used, the association of hypothermia and mortality risk would likely remain constant. Body temperature in captive rabbits may follow a circannual cycle.35 The present study enrolled patients during a 1-year period to avoid confounding associated with variations in environmental temperature. Regardless, no significant association was found between month of admission and presence of hypothermia. Although this finding suggested that environmental temperature does not play a role in the development of hypothermia, it should be considered that the present study was performed in a clinic situated in a temperate Mediterranean climate. Results may differ with latitude, with colder environments potentially associated with more severe hypothermia in the patient population examined.
The present study should be considered hypothesis testing as it was a priori designed and a moderate to large effect size was observed.54 Nevertheless, there are natural limitations related to the design of the study, and cohort studies have important strengths and weaknesses.55 First, we emphasize that association does not prove causation.54 Even though the present study demonstrated that rectal temperature was significantly linked with increased risk of death, the results of this study do not indicate that hypothermia causes death in rabbits. Hypothermia may be an indication of the severity of the disease rather than an independent predictor of death.56 In a large study14 of human trauma patients, hypothermia was considered to be an independent contributor to mortality, and some evidence suggests that hypothermia may be a result of injury severity and therefore should not be considered an independent predictor of mortality.17
A further limitation of cohort studies is that they do not account for confounders that are unknown (ie, unmeasured confounders). Briefly, the observed association between hypothermia and mortality in this study may have been a function of unmeasured patient characteristics that were related to the exposure (hypothermia) or to the outcome of interest (mortality). For instance, if practicing veterinarians believed that hypothermia was an indicator of poor prognosis, they may have had a less involved plan of treatment or may have suggested euthanasia to owners of hypothermic rabbits. However, we suggest that this scenario was unlikely because the practicing veterinarians were not aware of the prognostic value of temperature and the clinic had policies such that euthanasia was reserved for a relatively small number of patients. Less than 3% of the rabbits that died in the clinic each year were euthanized, and funds were available to assist owners with economic constraints. Thus, we believe that we can exclude a spurious association between hypothermia and death (ie, suggestion of euthanasia in hypothermic rabbits).
Another limitation of the present study was that there were potential imprecisions in how some variables were measured (eg, temperature and presence of gastrointestinal stasis) because of the the pragmatic design of the study.57 Finally, because this was a clinical study, the treatment was different for hypothermic versus normothermic rabbits. Therefore, although it is highly unlikely on the basis of current evidence from human medicine,21,22,58 we cannot exclude that the higher mortality risk in hypothermic rabbits was caused by the treatment for hypothermia, rather than the hypothermia itself. Consideration should be given to the hypothesis that high ambient temperatures may be detrimental in rabbits. In a randomized controlled study59 performed on New Zealand White rabbits, sudden exposure to high ambient temperatures (30°C [86°F]) provoked a decrease in feed consumption, a change in cecal fermentation (decrease in volatile fatty acid concentrations and acetic, propionic, and butyric acid concentrations), and a change in the intestinal flora, with an increase in anaerobes and clostridia. This evidence that a change in environmental temperature has an effect on metabolism and bacterial flora and that rabbits treated for hypothermia had a high mortality rate should be the basis for further research. In particular, we suggest that there is a strong need for a trial in which hypothermic rabbits are randomized to different rewarming treatments to determine whether high environmental temperature (eg, the use of incubators) negatively impacts survival.
In the present study, we demonstrated a significant association between the presence of hypothermia at admission and mortality in client-owned rabbits. Because of the size of the effect, this association should be also considered clinically important. Thus, rectal temperature should be considered a major prognostic predictor in client-owned rabbits and should always be measured during a physical examination. Treatment of hypothermia in rabbits needs further research.
Acknowledgments
Presented in abstract form at the 13th Annual Conference of the Association of Exotic Mammal Veterinarians, Orlando, Fla, October 2014.
The authors thank Drs. Tommaso Collarile, Ivano A. Ciraci, Valeria Del Duca, Alessandra Carnimeo, Alessandro Montani, and Sabrina Sinisi for assistance with clinical management of rabbits.
ABBREVIATION
| CI | Confidence interval |
Footnotes
Vedodigit II, PiC Solutions, Artsana SpA, Grandate, Italy.
Intensive care unit, Joe Freed's Petiatric Supply, Wichita, Kan.
Critical Care, Oxbow Animal Health, Murdock, Neb.
JC-236, Joycare SpA, Rome, Italy.
IBM SPSS 22.0, SPSS Inc, Chicago, Ill.
Review Manager (RevMan) 5.2, The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark.
References
1. Kayser C. Physiological aspects of hypothermia. Annu Rev Physiol 1957; 19: 83–120.
2. Jalan R, Rose C. Hypothermia in acute liver failure. Metab Brain Dis 2004; 19: 215–221.
3. Romanovsky AA, Shido O, Sakurada S, et al. Endotoxin shockassociated hypothermia. How and why does it occur? Ann NY Acad Sci 1997; 813: 733–737.
4. Moffatt SE. Hypothermia in trauma. Emerg Med J 2013; 30: 989–996.
5. Haskins SC. Hypothermia and its prevention during general anesthesia in cats. Am J Vet Res 1981; 42: 856–861.
6. Oliver TK Jr. Temperature regulation and heat production in the newborn. Pediatr Clin North Am 1965; 12: 765–779.
7. Hey EN. The relation between environmental temperature and oxygen consumption in the new-born baby. J Physiol 1969; 200: 589–603.
8. Adamson SK Jr, Gandy GM, James LS. The influence of thermal factors upon oxygen consumption of the newborn human infant. J Pediatr 1965; 66: 495–508.
9. Dahm LS, James LS. Newborn temperature and calculated heat loss in the delivery room. Pediatrics 1972; 49: 504–513.
10. Kirkpatrick AW, Chun R, Brown R, et al. Hypothermia and the trauma patient. Can J Surg 1999; 42: 333–343.
11. Kongsayreepong S, Chaibundit C, Chadpaibool J, et al. Predictor of core hypothermia and the surgical intensive care unit. Anesth Analg 2003; 96: 826–833.
12. Redondo JI, Suesta P, Serra I, et al. Retrospective study of the prevalence of postanaesthetic hypothermia in dogs. Vet Rec 2012; 171: 374.
13. Waibel BH. Hypothermia in trauma patients: predicting the big chill. Crit Care 2012; 16: 155.
14. Martin RS, Kilgo PD, Miller PR, et al. Injury-associated hypothermia: an analysis of the 2004 National Trauma Data Bank. Shock 2005; 24: 114–118.
15. Silverman WA, Fertig JW, Berger AP. The influence of the thermal environment upon the survival of newly born premature infants. Pediatrics 1958; 22: 876–886.
16. Day RL, Caliguiri L, Kamenski C, et al. Body temperature and survival of premature infants. Pediatrics 1964; 34: 171–181.
17. Trentzsch H, Huber-Wagner S, Hildebrand F, et al. Hypothermia for prediction of death in severely injured blunt trauma patients. Shock 2012; 37: 131–139.
18. Ahmed A, Aboshady I, Munir SM, et al. Decreasing body temperature predicts early rehospitalization in congestive heart failure. J Card Fail 2008; 14: 489–496.
19. Singer AJ, Taira BR, Thode, HC Jr, et al. The association between hypothermia, prehospital cooling, and mortality in burn victims. Acad Emerg Med 2010; 17: 456–459.
20. Quiroga E, Tran NT, Hatsukami T, et al. Hypothermia is associated with increased mortality in patients undergoing repair of ruptured abdominal aortic aneurysm. J Endovasc Ther 2010; 17: 434–438.
21. Gentilello LM, Jurkovich GJ, Stark MS, et al. Is hypothermia in the victim of major trauma protective or harmful? A randomized, prospective study. Ann Surg 1997; 226: 439–447.
22. Wang Y, Feng J, You G, et al. Heating pad for the bleeding: external warming during hemorrhage improves survival. J Trauma 2011; 71: 1915–1919.
23. AVMA. US pet ownership & demographics sourcebook. Schaumburg, Ill: AVMA, 2012.
24. Schepers F, Koene P, Beerda B. Welfare assessment in client-owned rabbits. Anim Welf 2009; 18: 477–485.
25. Fisher PG. Standards of care in the 21st century: the rabbit. J Exot Pet Med 2010; 19: 22–35.
26. Brodbelt DC, Blissitt KJ, Hammond RA, et al. The risk of death: the confidential enquiry into perioperative small animal fatalities. Vet Anaesth Analg 2008; 35: 365–373.
27. Varga M. Textbook of rabbit medicine. Oxford, England: Butterworth-Heinemann Elsevier, 2014.
28. Harcourt-Brown F. Textbook of rabbit medicine. Oxford, England: Butterworth-Heinemann, 2002.
29. Hull D. Oxygen consumption and body temperature of new-born rabbits and kittens exposed to cold. J Physiol 1965; 177: 192–202.
30. Richardson V. Rabbits: health, husbandry and diseases. Oxford, England: Blackwell Science, 2000;25:.
31. Chalmers I, Bracken MB, Djulbegovic B, et al. How to increase value and reduce waste when research priorities are set. Lancet 2014; 383: 156–165.
32. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol 2008; 61: 344–349.
33. Kirkewood BR, Sterne JA. Essential medical statistics. Oxford, England: Blackwell Science Ltd, 2003.
34. Graham J, Mader DR. Basic approach to veterinary care. In: Quesenberry KE, Carpenter JW, eds. Ferrets, rabbits and rodents: clinical medicine and surgery. 3rd ed. St Louis: Saunders Elsevier, 2012; 174–182.
35. Pericin C, Grieve AP. Seasonal variation of temperatures in rabbits. Lab Anim 1984; 18: 230–236.
36. Paul-Murphy J. Critical care of the rabbit. Vet Clin North Am Exot Anim Pract 2007; 10: 437–461.
37. Lichtenberger M, Lennox AM. Critical care of the exotic companion mammal (with a focus on herbivorous species): the first twenty-four hours. J Exot Pet Med 2012; 21: 284–292.
38. Lichtenberger M, Lennox A. Updates and advanced therapies for gastrointestinal stasis in rabbits. Vet Clin North Am Exot Anim Pract 2010; 13: 525–541.
39. Huynh M, Vilmouth S, Gonzalez MS, et al. Retrospective cohort study of gastrointestinal stasis in pet rabbits. Vet Rec 2014; 175: 225.
40. Peduzzi P, Concato J, Feinstein AR, et al. Importance of events per independent variable in proportional hazards regression analysis. II. Accuracy and precision of regression estimates. J Clin Epidemiol 1995; 48: 1503–1510.
41. Hosmer DW, Lemeshow S, Sturdivant RX. Applied logistic regression. 3rd ed. NJ: John Wiley and Sons Inc, 2013.
42. Little RJA, Rubin DB. Statistical analysis with missing data. New York: Wiley, 1987.
43. Black PR, van Devanter S, Cohn LH. Effects of hypothermia on systemic and organ system metabolism and function. J Surg Res 1976; 20: 49–63.
44. Grant R. Emotional hypothermia in rabbits. Am J Physiol 1950; 160: 285–290.
45. Lichtenberger M, Ko J. Critical care monitoring. Vet Clin North Am Exot Anim Pract 2007; 10: 317–344.
46. Peng RY, Bongard FS. Hypothermia in trauma patients. J Am Coll Surg 1999; 188: 685–696.
47. Harcourt-Brown FM, Harcourt-Brown SF. Clinical value of blood glucose measurement in pet rabbits. Vet Rec 2012; 170: 674.
48. Bonvehi C, Ardiaca M, Barrera S, et al. Prevalence and types of hyponatraemia, its relationship with hyperglycaemia and mortality in ill pet rabbits. Vet Rec 2014; 174: 554.
49. Sousa MG, Carareto R, Pereira-Junior VA, et al. Agreement between auricular and rectal measurements of body temperature in healthy cats. J Feline Med Surg 2013; 15: 275–279.
50. Chen PH, White CE. Comparison of rectal, microchip transponder, and infrared thermometry techniques for obtaining body temperature in the laboratory rabbit (Oryctolagus cuniculus). J Am Assoc Lab Anim Sci 2006; 45: 57–63.
51. Vassal T, Benoit-Gonin B, Carrat F, et al. Severe accidental hypothermia treated in an ICU: prognosis and outcome. Chest 2001; 120: 1998–2003.
52. World Health Organization, Department of Reproductive Health and Research. Thermal protection of the newborn: a practical guide. Geneva: World Health Organization, 1997.
53. Lee RA, Towle Millard HA, Weil AB, et al. In vitro evaluation of three intravenous fluid line warmers. J Am Vet Med Assoc 2014; 244: 1423–1428.
54. Hennekens CH, DeMets D. Statistical association and causation: contributions of different types of evidence. JAMA 2011; 305: 1134–1135.
55. Grimes DA, Schulz KF. An overview of clinical research: the lay of the land. Lancet 2002; 359: 57–61.
56. Steinemann S, Shackford SR, Davis JW. Implications of admission hypothermia in trauma patients. J Trauma 1990; 30: 200–202.
57. Gaglio B, Phillips SM, Heurtin-Roberts S, et al. How pragmatic is it? Lessons learned using PRECIS and RE-AIM for determining pragmatic characteristics of research. Implement Sci 2014; 9: 96.
58. Moola S, Lockwood C. Effectiveness of strategies for the management and/or prevention of hypothermia within the adult perioperative environment. Int J Evid Based Healthc 2011; 9: 337–345.
59. Amici A, Canganella F, Bevilacqua L. Effects of high ambient temperature in rabbits: metabolic changes, caecal fermentation and bacterial flora. World Rabbit Sci 1998; 6: 319–324.
