Introduction

Heart disease is estimated to affect approximately 11% of the dog population, with 75–80% of canine heart disease classified as chronic valvular heart disease, also called endocardiosis.^1–6 The prevalence of chronic valvular heart disease varies by size, breed, gender, and age; small- to medium-sized breeds (generally less than 20 kg body weight) are dramatically predisposed, with males more frequently (approximately 1.5x) affected, and the disease becomes more prevalent with advancing age.^1–6

Periodontal disease is also common in dogs—potentially among the most common diseases of older dogs.⁷ A study performed at the University of Minnesota demonstrated that oral disease accounted for 5.8% of the presenting diagnoses in dogs aged 0–7 yr of age, 13.7% in dogs aged 7–10 yr, and 13.6% in dogs aged 10–25 yr.⁸ Periodontal disease commonly causes bad breath and can cause abnormal eating behavior, pain, facial bone loss, local abscesses, and, in extreme cases, systemic sepsis. Dental cleaning (prophylaxis) and minor dental surgery are common reasons that dogs undergo general anesthesia. In the authors' experience, the presence of heart disease in dogs requiring general anesthesia for dental treatment represents an important source of anxiety among both practicing veterinarians and pet owners. In some dogs afflicted with both heart and periodontal disease, needed dental procedures are denied or delayed because of this fear. Many other such dogs are referred to specialist centers to undergo general anesthesia for their dental procedure. We designed this study to begin to evaluate whether our perception of the widespread fear of anesthetizing dogs with cardiac disease for routine dental procedures was justified in a referral hospital setting.

Previous studies have found the risk of anesthetic death in dogs to be between 0.11–0.43% of cases, with anesthetic mortality increasing with increasing age and the American Society of Anesthesiologists (ASA) status.^9–15 Two previous studies that evaluated the overall incidence of anesthetic complications in dogs found that complications occurred in 2.1–12% of anesthetized dogs; however another study performed in a veterinary teaching hospital found minor complications in 0.9–38.4% of dogs undergoing general anesthesia.^12–14 Although these studies were thorough in evaluating anesthetic complications and mortality in general populations of dogs undergoing sedation or general anesthesia, to the authors' knowledge, no previous study has looked specifically at the influence of the presence of heart disease on the occurrence of anesthetic complications and mortality.

This study compares the occurrence of several predefined anesthetic complications and death associated with general anesthesia for dental prophylaxis or minor dental surgery in dogs with heart disease to the occurrence of those events in a matched cohort of dogs without evidence of heart disease. Our primary aim was to test the hypothesis that the risk of anesthetic complications or death during routine dental procedures performed in dogs with heart disease did not differ significantly from that risk in dogs without heart disease when performed in a referral setting.

Materials and Methods

Data Collection

Medical records were reviewed for all included cases, and the following information was extracted: patient identification, breed, sex, age, body weight, dental diagnosis and procedure, cardiac diagnosis, cardiac medications, ACVIM/American College of Cardiology heart failure classification, ASA classification, total anesthesia time, premedications received, induction drugs received, and anesthetic maintenance protocol used. In addition, specific complications that may occur under anesthesia were evaluated as present or absent by evaluating the anesthesia record every 5 min during the procedure: bradycardia (heart rate [HR] <60 beats min⁻¹), tachycardia (HR > 160 breaths min⁻¹), hypotension (mean arterial pressure <60 mmHg or systolic arterial pressure <80 mmHg, both measured from noninvasive devices), hypoventilation (partial pressure of end-tidal carbon dioxide >55 mmHg), hypoxemia (hemoglobin oxygen saturation <95%), cardiac arrhythmia, hypothermia (body temperature <35°C), and cardiopulmonary arrest and are summarized in Table 1.

TABLE 1 Perianaesthetic Morbidity Definitions Used in This Study

View Fullscreen

When bradycardia, tachycardia, hypotension, or a cardiac arrhythmia were identified, the following information was also recorded: percentage of total anesthetic time spent bradycardic, percentage of total anesthetic time spent tachycardic, percentage of total anesthetic time spent hypotensive, and the complete cardiac rhythm diagnosis. Finally, all records were evaluated to determine whether the dog survived to hospital discharge.

Statistical Analysis

Tests of association between specific anesthetic complications, patient characteristics, drugs used, and case or control status were performed using traditional tests of hypotheses. For categorical variables, χ² tests of association were used when the expected count for each contingency table cell was greater than five, and Fisher's exact tests were employed when any expected values were less than five. For continuous variables, t tests with one degree of freedom were used. Two-sided hypothesis tests were used, and the reported P values were not corrected for multiple comparisons. Using highly anticonservative, uncorrected P values was done as a strategy to maximize the likelihood of identifying an effect of the presence of heart disease if one existed, because the primary hypothesis of the study was null (i.e., that cardiac disease status did not influence the occurrence of anesthetic complications or mortality). In light of the primary null hypothesis, it was also important to know the power of the study to detect differences in outcome or event rates between groups, if those differences actually existed. Power curves were constructed for outcomes or events with different prevalence rates (each different prevalence rate is represented by a different power curve in Figure 1) in the range of those reported in the veterinary anesthesia literature for anesthetic complications and mortality in dogs. In addition, the size of the differential effect caused by the presence of heart disease on the occurrence of anesthetic complications that this study would have an 80% chance of detecting was calculated. All statistical analyses were performed in statistical software^a.

View Figure

• Download as Powerpoint
• Download Figure

Results

The 200 dogs (100 with and 100 without heart disease) included 47 different breeds plus 11 mixed-breed dogs. There were 17 toy or miniature poodles, 13 Italian greyhounds, and 10 boxers, with all other breeds represented by fewer than 10 dogs. Eighty-six of the 100 dogs in the CD group had chronic valvular heart disease with primarily mitral valve insufficiency. Other cardiac diseases represented included four dogs with arrhythmogenic right ventricular cardiomyopathy, three with severe pulmonic stenosis (two more than 3 yr following balloon valvuloplasty), two with ventricular septal defects, two with a history of treated heartworm disease with severe residual pulmonary hypertension and tricuspid regurgitation, and one each with dilated cardiomyopathy, severe subaortic stenosis, and tricuspid valve dysplasia. The distribution of cardiac disease included in the study is summarized in Table 2. Case matching ensured that there was no difference between groups with respect to gender (P > .24; the CD group had 46 and the NCD group 60 female or female spayed dogs; the CD group had 54 and the NCD group had 40 male or castrated male dogs). There was also no difference between groups with respect to body weight (P > .56; mean body weights were CD = 12.53 kg and NCD = 13.42 kg). Dogs in the CD group were, despite attempted matching, significantly older than those in the NCD group (P < .02; mean ages CD = 11.24 yr, NCD = 10.26 yr). Dogs in the CD group had a significantly higher ASA classification than those in the NCD group (P < .001) with 57% of dogs in the CD group having a score of three or greater, compared to only 9% of the NCD group. Table 3 summarizes the distribution of ASA scores in each of the study groups.

TABLE 2 Types and Frequencies of Canine Cardiac Disease for Dogs Included in the Cardiac Disease Group

View Fullscreen

TABLE 3 American Society of Anesthesiologists Score Distribution for Dogs Included in Each Study Group

View Fullscreen

All dogs in both groups survived to hospital discharge (i.e., no anesthetic deaths occurred in this study). There was no significant difference in time to hospital discharge between the groups (P = .73), with 85% of dogs from group CD and 88% of dogs from group NCD discharged on the same day as their anesthetic event. There were also no significant differences identified between groups in the overall prevalence of anesthetic complications (CD = 82%, NCD = 83%, P > .84). The most common anesthetic complication was hypotension, with 49% of dogs in both groups affected. Bradycardia was the second-most common complication, with 35% of dogs in the CD group and 36% of dogs in the NCD group affected. During the anesthetic period, 21% of CD dogs and 24% of NCD dogs experienced hypoventilation. The remainder of the anesthetic complications catalogued were less common. There were no significant associations between any of the anesthetic complications evaluated and cardiac disease status (Table 4). Cardiac arrhythmias were infrequent in both groups. In the CD group, six dogs experienced episodes of second-degree atrioventricular block, three dogs experienced ventricular escape beats, and isolated atrial or ventricular premature contractions each occurred in one dog. In the NCD group, four dogs experienced episodes of second-degree atrioventricular block, and one dog experienced isolated ventricular premature contractions.

TABLE 4 Association Analysis of Each Perianaesthetic Morbidity Between Dogs with Cardiac Disease and Those Without

View Fullscreen

There was no significant difference between groups with respect to the average total duration of anesthesia (CD = 152 min, NCD = 139 min; P = .1137). Comparison of the percentage of total anesthesia time spent bradycardic, tachycardic, or hypotensive also revealed no significant differences between the two groups (P > .09, .12, and .68, respectively).

There was no significant difference between the two groups with regard to the premedication drugs administered, as nearly every case received an opioid-based premedication; the most common choice in both groups was hydromorphone. There was a significant difference in the induction drugs administered to the two groups (P < .001), with many more dogs in the CD group receiving the combination of midazolam and etomidate. Lastly, there was no significant difference between the maintenance inhalant anesthetic agents used in the two groups (P < .99). Table 5 lists the anesthetic drugs used and the frequency of their use. The doses of the drugs that were used most commonly in both groups were compared, and there was no significant difference between the doses used in the two groups for hydromorphone (P = .432), acepromazine (P = .962), midazolam used during anesthetic induction (P = .727), propofol (P = .642), or isoflurane (P = .092).

TABLE 5 Premedication, Induction, and Maintenance Drugs Used in Each Group

View Fullscreen

Figure 1 shows the power of the study to identify differences in anesthetic complication rates between the CD and NCD groups, had differences actually existed, over the range of potential complication rates from the literature. The study had 80% power (i.e., an 80% chance) to detect differences in the complications listed in Table 3. These calculations show that the study was reasonably (though certainly not perfectly) powered to detect clinically meaningful differences between the groups for frequently encountered anesthetic complications. For rare outcomes or events (such as death), however, the power of the study to identify or detect small differences between the CD and NCD groups was low.

Discussion

This investigation compared the anesthetic complications and mortality of dogs with and without underlying heart disease that underwent dental prophylaxis or minor dental surgery at a single, large referral center. While it is the largest study of its kind that the authors are aware of, the power of this study to detect differences between groups remains limited, with low power to detect differences between groups with respect to rare events. Fortunately, there were no anesthetic deaths in either group in this study. This lack of mortality was not surprising, since previous studies have estimated anesthetic mortality in dogs at 0.11–0.43%, and only 200 dogs were evaluated for this study.^9–15 Those same studies found an increased risk of anesthetic mortality with increasing age and ASA status.^9–15 In this study, the CD group was significantly older and had a higher ASA classification compared to the NCD group. Despite these apparent risk factors, there was no significant difference in the overall prevalence of anesthetic complications between the two groups. It was, however, interesting to observe that the overall complication frequency observed in this study, while not different between groups, was much higher (82.5%) than the previously reported range of 0.9–38.4%.¹⁴ The procedures described here all took place at a veterinary teaching hospital, and while students may have been involved in case management, they were always under the direct supervision of anesthesiologists and trained technicians; it seems unlikely that student inexperience played a role in the high complication rate observed. One potential influence on the high frequency of complications observed was the relatively high average age of the dogs reported here. Their mean age was 10.75 yr, while in the previous study, the median age was 6 yr, and older dogs in that study had nearly twice the odds of serious anesthetic complications.¹⁴ Another possible explanation for the large number of anesthetic complications observed are the relatively strict criteria utilized to diagnose anesthetic complications in this study combined with the monitoring frequency (every 5 min). Those criteria (summarized in Table 1) represented reasonable clinical definitions of complicating events, and dogs that met the definition for an event for even one anesthetic recording period in the record were considered to have experienced an anesthetic complication. As defined here, anesthetic complications are probably more common in dogs than the current literature suggests.

Multiple studies have documented the association between preexisting disease or organ system dysfunction and increased anesthetic risk. One human study evaluating risk of death under anesthesia demonstrated five variables that predicted an increase in risk: increased age, presence of chronic cardiac disease, presence of renal disease, emergency surgery, and the type of operation.¹⁶ Another human study evaluating adverse perioperative outcomes demonstrated a recent history of cardiac failure or myocardial infarction or an ASA status of three or four, amongst others, as predictors of increased anesthetic risk.¹⁷ Several veterinary studies have supported this assertion, although most evaluate ASA classification rather than any particular preexisting disease condition. Three studies evaluating the influence of ASA classification on anesthetic mortality demonstrated an increased frequency of adverse anesthetic events with ASA classifications of three or greater.^9,10,18 One study demonstrated a higher incidence of complications and cardiac arrest among dogs with an ASA status of three or greater when compared to those with an ASA status of one or two.¹² Another study, performed at a veterinary teaching hospital, also demonstrated a higher frequency of anesthetic complications in animals with an ASA status of three or greater.¹⁴ Those findings are not surprising, since the ASA classification scheme was originally intended to serve as an indicator of anesthetic risk.¹⁹

In contrast to those studies, the results of this study show no significant difference in the occurrence of any of the anesthetic complications measured, despite the presence of a significant difference in ASA status and age between two groups matched for other relevant factors except the presence or absence of heart disease. These findings may reflect the different nature of heart disease in dogs versus humans (primarily valvular versus primarily vascular, respectively). In the face of human coronary artery disease, the common physiologic disturbances that accompany general anesthesia (arterial vasodilation, systemic hypotension, autonomic dysregulation, brady- or tachyarrhythmias) are far more threatening to myocardial perfusion than those same disturbances in dogs with mitral valve insufficiency (in which systemic vasodilation may reduce mitral valvular regurgitant fraction and actually improve cardiac output, and the risk of myocardial infarction is profoundly lower). The age difference between the two groups, while statistically significant, was less than 1 yr, and probably of little clinical significance.

Previous veterinary studies have indicated an increased anesthetic risk for certain breeds of dogs.^10,12,20 This study did not attempt to evaluate the influence of breed on anesthetic complications but rather controlled for the potential influence of breed differences by the case-matching scheme.

In humans, the total duration of anesthesia has been correlated with increased risk of mortality, although this same correlation has not been clearly demonstrated in veterinary studies, as one study demonstrated an increased odds ratio for anesthetic mortality while two others failed to show a correlation.^9,12,14,16 This study did not specifically evaluate the overall effect of the duration of anesthesia on the occurrence of complications, but there was no significant difference between groups with respect to average anesthetic duration, effectively eliminating a difference in average anesthetic duration as a potential influence on the outcome of the study.

It was beyond the scope of this retrospective study to evaluate the effect of any drug or drug combination on the frequency or duration of any of the predefined anesthetic complications. It is important to note, however, that the drug regimens differed between groups. Significantly more dogs in the CD group were induced for general anesthesia with midazolam and etomidate compared to dogs in the NCD (control) group. This difference reflects a reasonable clinical choice on the part of the anesthetists, since etomidate is an injectable general anesthetic that produces minimal changes in myocardial function, HR, or blood pressure in dogs.²¹ No differences were identified between groups with regard to the doses of hydromorphone, acepromazine, midazolam (at induction), propofol, or isoflurane, suggesting that use of any of these drugs in dogs with cardiac disease did not influence the occurrence of anesthetic complications.

Previous studies have highlighted the use of xylazine, other alpha-2 adrenergic agonists, and isoflurane as factors that increase the risk of anesthesia-related morbidity and mortality.^10–12 The cardiovascular depressant effects of alpha-2 adrenergic agents have been well described and these agents are, therefore, usually avoided in dogs with known cardiac disease.²² Significantly fewer dogs (only one) in the CD group received an alpha-2 adrenergic agonist, dexmedetomidine, compared to 11 dogs in the NCD (control) group in this study. This reduced rate of alpha-2 agonist usage in the CD again reflects reasonable clinical preference, and while this study was not designed to test drug efficacy, it is important to interpret the results of this study in light of these significant differences.

While we have no way of knowing what role these between-group drug differences played in the outcome of this study, the results (failure to observe increased anesthetic complications in animals with heart disease) should not be extrapolated to clinical situations in which these drug differences are not present.

One major weakness of any retrospective study is the inability to judge intent and clinical discretion that may have existed at the time the cases were managed. Anesthetic complications, as defined in this study, are routinely treated by the anesthetist in charge of the case, and these treatments (such as intravenous fluid boluses, dopamine infusions, and fentanyl infusions) may have influenced the outcome of the anesthetic events. Interpreting the anesthetic record for decisions to implement a treatment may have caused flawed assumptions in this study and, as such, the potential impact of any treatment on patient outcome was not explored.

Finally, it is important to recognize that all of these procedures were performed at a veterinary teaching hospital under the supervision of highly trained veterinary technicians and anesthesiologists. In addition, all of the dogs included in this study were carefully monitored by trained observers using multiparameter veterinary anesthesia monitors.^bc This may represent a clinically meaningful difference from veterinary environments in which less-intensive monitoring by less-trained individuals routinely occurs. Poor anesthetic monitoring has been implicated as a risk factor for anesthetic complications and mortality in previous studies, with the Confidential Enquiry into Perioperative Small Animal Fatalities indicating that monitoring of pulse and utilization of pulse oximetry reduce the risk of mortality in cats.^{9,10,12,14,23} Although the frequency of anesthetic complications was high in this study, the absence of serious morbidity (such as cardiac arrest), like the high complication rate itself, is likely to be at least in part a reflection of good monitoring practices and the ability of trained individuals to recognize and respond quickly and appropriately to changes in patient parameters. The results of this study cannot be extrapolated to less-intensively monitored environments. The results of this study do, however, provide a reference point for veterinarians when counseling owners of dogs with cardiac disease regarding the safety of general anesthesia in a referral setting.

Conclusion

In a veterinary teaching hospital setting, dogs with cardiac disease undergoing general anesthesia for dental prophylaxis or minor dental surgery experienced no more anesthetic complications or mortality than dogs without cardiac disease. Because the study was too small to provide adequate power to identify small but potentially significant differences between groups for rare events such as anesthetic death, the influence of heart disease on the absolute risk of anesthetic death remains uncertain.