Factors associated with major complications in the short-term postoperative period in dogs undergoing surgery for brachycephalic airway syndrome

Abstract

Surgical correction of brachycephalic airway syndrome (BAS) in dogs has been reported to result in low complication rates and good long-term outcomes. Previous reports have not identified risk factors for the development of complications following BAS surgery. This retrospective study evaluated a wide variety of patient- and procedure-related, pre-operative, intra-operative, and post-operative factors for an association with the development of major postoperative complications in the short-term period following BAS surgery. The overall major complication rate, including death or euthanasia, was 4/55 (7%) dogs. Temporary tracheostomy was the only major surgical complication identified (n = 3). Multiple logistic regression identified postoperative radiographic evidence of pneumonia as associated with the development of any major complication overall, requirement of a temporary tracheostomy postoperatively, and death or euthanasia, within the short-term postoperative period. Future prospective studies should evaluate specific risk factors for an association with major complications following BAS surgery in dogs to improve patient outcomes.

Introduction

Brachycephalic airway syndrome (BAS) is an increasingly prevalent disease in dogs due to the increase in popularity of brachycephalic breeds (1,2). The components of BAS, stenotic nares, elongated soft palate, and hypoplastic trachea, result in increased negative pressure within the upper airway leading to clinical signs varying from mild respiratory abnormalities to severe respiratory distress and collapse (1–4). Recent literature also purports the presence of nasal turbinates protruding beyond the choana as playing a role in upper airway obstruction in dogs with BAS (5). Persistently increased negative pressure within the upper airway can also lead to various grades of laryngeal collapse. Grade I laryngeal collapse is defined as everted laryngeal saccules and is reported in 56% to 59% of cases (1,4,6). Gastrointestinal signs have also been documented to be associated with BAS and medical treatment of these signs, along with surgical correction of BAS, have improved outcome and prognosis (7,8).

Riecks et al (4) reported a good to excellent outcome in 94% of dogs that had long-term follow-up. Even with the persistence or recurrence of clinical signs following surgery, including laryngeal collapse, Torrez et al (6) identified that 57% of dogs with long-term follow-up were reported by owners to be much improved clinically. Poncet et al (7) determined that 92% of dogs undergoing BAS surgery had a significant improvement in respiratory signs within 6 wk after surgery.

Surgical correction of BAS abnormalities, including rhinoplasty, laryngeal sacculectomy, and caudal staphylectomy, has been reported to result in low overall complication rates, ranging from 6% to 26% without any correlation to the type of procedure and complication (4,7). Intraoperative complications have included hypotension, tachycardia, and atrioventricular block in 5% of dogs (1). Postoperative complications have included vomiting and mucoid regurgitation (18%); aspiration pneumonia (11%); nasal discharge (5%); respiratory noise, dehiscence, or regurgitation (3% to 10%); temporary tracheostomy (3% to 5%); and severe dyspnea or death (3% to 5%) (1,4,7,9).

To the authors’ knowledge, no previous studies have identified risk factors for the development of complications following BAS surgery in dogs. The specific objective of this retrospective study was to investigate possible associations between various patient- and procedure-related factors and the development of major complications in the short-term postoperative period in dogs undergoing BAS surgery. The hypothesis was that 1 or more factors would be identified to be associated with an increased risk of major complications in the short-term postoperative period.

Materials and methods

Medical records of client-owned dogs presented to the Lois Bates Acheson Small Animal Veterinary Teaching Hospital at Oregon State University for BAS surgery from 2006 to 2014 were reviewed. Inclusion criteria were dogs undergoing any of the following BAS surgical procedures: rhinoplasty, laryngeal sacculectomy, and/or caudal staphylectomy. No exclusion criteria were applied. Dogs which had BAS surgeries on more than 1 date had their most recent BAS surgical date used for purposes of this study in order to evaluate either a negative or positive effect of previous BAS surgery on short-term outcome. Patient- and procedure-related factors were collected and included signalment, history including prior BAS surgery, whether the surgery was performed at the study institution or with the primary care veterinarian, rectal temperature at presentation, premedication and anesthetic agents used, surgical procedure(s) performed, anesthetic and surgical times, and postoperative observations and treatments. Grading of laryngeal collapse was performed via sedated laryngeal examination: Grade I: laryngeal saccule eversion; Grade II: laryngeal saccule eversion and cuneiform process of arytenoid cartilage collapse; Grade III: corniculate process of arytenoid cartilage collapse. Pre- and postoperative radiographic reports were evaluated when available for evidence of pneumonia and defined as interstitial or alveolar pulmonary patterns diffusely or focally. For the purposes of this study, details of major complication were recorded and major complications were defined as those requiring surgical intervention and/or resulting in euthanasia or death within the short-term period between surgery and 2-week recheck examination.

Statistical analysis was performed using computer software (SAS V 9.3; Stata, Cary, North Carolina, USA). Simple logistic regression was performed to test for relationships between the patient- and procedure-related factors and the probability of major complications (both overall and individual types of major complications). If quasi-separation of variable occurred, Firth’s penalized maximum likelihood estimation method was used to reduce bias in logistic model parameter estimation. Multiple logistic regressions were performed between all significant risk factors in the simple logistic regressions and the probability of major complications (both overall and individual types of major complications). A step-wise selection procedure was used. Odds ratios are reported with 95% confidence intervals (95% CI). Kappa statistics were calculated to test for correlations between occurrences of any major complications. Significance was set at P ≤ 0.05. Continuous data are reported as mean ± standard deviation with interquartile ranges (IQR).

Results

Fifty-five dogs met the inclusion criteria for the study. There were 25 (46%) neutered males, 15 (27%) spayed females, 10 (18%) intact males, and 5 (9%) intact females. Breeds represented included 18 (33%) pugs, 16 (29%) English bulldogs, 7 (13%) French bulldogs, 4 (7%) Boston terriers, 3 (5%) shih tzu, and 1 (2%) each of the following: mixed breed dog, Staffordshire bull terrier, Dogue de Bordeaux, chow chow, dachshund, Chihuahua, and Cavalier King Charles spaniel. The mean age was 53 ± 42 mo (IQR 74 mo).

Historical data indicated that 48 (87%) dogs had stertor or stridor, 12 (22%) dogs had gastrointestinal signs (vomiting or regurgitation), and 8 (15%) dogs had previous BAS surgery. Of the 8 dogs which had previous BAS surgery, 5 (63%) had their previous surgery performed at the study hospital, while 3 (37%) had surgery performed by their referring veterinarian. Of the 5 dogs with previous surgery at the study hospital, 4 had rhinoplasty, 4 had caudal staphylectomy, and 5 had laryngeal sacculectomy, performed at a mean of 5.2 mo prior to presentation for the surgical episode used for this study.

Mean rectal temperature at presentation was 38.3 ± 0.46°C (IQR 0.56°C). Ten (18%) dogs had no evidence of laryngeal collapse (i.e., no everted laryngeal saccules). Forty-three (78%) dogs had evidence of grade 1 laryngeal collapse and 1 (2%) dog had grade 3 laryngeal collapse. No dog had grade 2 laryngeal collapse recorded and 1 (2%) dog did not have a laryngeal examination documented.

Premedication drugs used in the study population included various combinations of the following: acepromazine, dexmeditomidine, full and partial opioids, and anticholinergic agents. Induction agents also varied and included propofol, ketamine, midazolam, etomidate, and gas induction by mask. All dogs were maintained under anesthesia using isoflurane, desflurane, or sevoflurane administered by endotracheal tube in 100% oxygen.

Three dogs (5%) required emergency surgery secondary to upper airway obstruction while the remaining 52 (95%) dogs had a routine preoperative fast and surgery on a scheduled basis. Forty-nine (89%) dogs had a caudal staphylectomy, 42 (76%) had a laryngeal sacculectomy, and 35 (64%) had rhinoplasty performed. Twenty-one (38%) dogs had other non-BAS surgical procedures performed concurrently. These included 7 (13%) castrations or ovariohysterectomies, 4 (7%) orthopedic procedures, 2 (4%) urethropexies, 2 (4%) excisional biopsies, and 1 each (2%) of the following: unilateral left arytenoid lateralization, bilateral enucleation, cystotomy, cholecystectomy, upper gastrointestinal endoscopy, and a cutaneous mast cell tumor excision. Of these 21 dogs with additional surgical procedures, 9 had their BAS surgery performed after additional procedures, 7 had their BAS surgery performed before the additional procedures, and the order of surgeries could not be determined in 5 dogs. No difference was identified in outcomes between dogs that had their BAS surgery performed before or after additional procedures.

Mean anesthesia time was 107 ± 70 min (IQR 80 min). Mean total surgical time was 73 ± 64 min (IQR 48 min). Mean BAS surgical time was 37 ± 14 min (IQR 20 min). The lowest intraoperative rectal temperature was a mean of 36.7 ± 1.3°C (1.7°C).

Radiographs were performed before surgery in 26 (47%) dogs and 3 (12%) had radiographic evidence of pneumonia. Radiographs were performed after surgery in 5 (9%) dogs and 3 (5%) had radiographic evidence of pneumonia. Two of these dogs with postoperative radiographic evidence of pneumonia required placement of a temporary tracheostomy.

Mean postoperative rectal temperature during hospitalization was 38.3 ± 0.6°C (IQR 0.56°C). Thirteen dogs (24%) had vomiting and/or regurgitation after surgery. Perioperative medications varied by patient but included the following: anti-inflammatory doses of corticosteroids (n = 41, 75%), metoclopramide constant rate infusion (CRI) (n = 5, 9%), acepromazine (n = 35, 64%), maropitant (n = 7, 13%), trazodone (n = 1, 2%), full opioid agonists (n = 15, 27%), and dexmedetomidine (n = 3, 5%). Of the 41 dogs receiving perioperative corticosteroids, 25 dogs (61%) had them administered before or during surgery.

The overall major complication rate was 4/55 (7%) dogs. The only surgical intervention recorded was placement of a temporary tracheostomy tube in 3 of these dogs, all of whom died due to respiratory complications between 1 and 3 d after surgery. Two of the 4 dogs experiencing major complications were presented on an emergency basis. One of these dogs had grade 3 laryngeal collapse and had a unilateral left arytenoid lateralization concurrently with BAS surgery (caudal staphylectomy and sacculectomy). This dog was not presented on an emergency basis. The remaining dog in this subgroup of 4 dogs was euthanized by the referring veterinarian 3 d after surgery. A definitive diagnosis leading to this decision was not available but clinical signs of supraventricular tachyarrhythmia and systemic hypertension were noted immediately prior to expiration.

All of the above recorded variables underwent statistical analysis for an association with development of overall and individual major complications, specifically placement of a temporary tracheostomy or death or euthanasia. Simple logistic regression identified 2 factors negatively associated with the development of a major complication overall as well as death and euthanasia within the short-term postoperative period: perioperative use of a metoclopramide CRI (OR: 16; 95% CI: 1.64 to 156.55; P = 0.0172) and postoperative radiographic evidence of pneumonia (OR: 49.99; 95% CI: 3.08 to 810.38; P = 0.0059). Three factors were associated with requirement of a temporary tracheostomy after surgery: perioperative use of a metoclopramide CRI (OR: 32.0; 95% CI: 2.2 to 461.5; P = 0.0109), postoperative radiographic evidence of pneumonia (OR: 100.0; 95% CI: 4.5 to 2245.1; P = 0.0037), and longer overall surgical time (OR: 1.02; 95% CI: 1.001 to 1.03; P = 0.0307). The differences between the subgroups (e.g., major complication versus no major complication) for these particular variables are provided in Table 1. There was a concordance between dogs having a tracheostomy performed and dogs experiencing death or euthanasia (Kappa 0.8475; P < 0.0001).

Table 1.

Differences in variables identified as significant on univariate analysis between dogs with and without major complications in the short-term postoperative period after brachycephalic airway syndrome surgery

Patient variable	Overall major complication (n = 4)	No overall major complication (n = 51)	Tracheostomy performed (n = 3)	No tracheostomy performed (n = 52)	Death or euthanasia (n = 4)	No death or euthanasia (n = 51)
Use of metoclopramide CRI in the perioperative period (number and %)	3 (75%)	2 (4%)	2 (66%)	3 (6%)	3 (75%)	2 (4%)
Postoperative radiographic evidence of pneumonia (number and %)	2 (50%)	1 (2%)	2 (66%)	1 (2%)	2 (50%)	1 (2%)
Median overal surgical time	NS	NS	Sa	Sb	NS	NS

NS — Not significant; S — Significant;

^a197 min;

^b49 min.

Stepwise multiple logistic regression identified only 1 factor, postoperative radiographic evidence of pneumonia, associated with the development of a major complication overall (OR: 49.99; 95% CI: 3.08 to 810.38; P = 0.0059), requirement of a temporary tracheostomy after surgery (OR: 90.00; 95% CI: 4.45 to 2023.38; P = 0.0046), and death or euthanasia within the short-term postoperative period (OR: 49.99; 95% CI: 3.09 to 810.38; P = 0.0059).

Discussion

This study identified specific patient- and procedure-related factors that were associated with an increased risk of development of major postoperative complications in dogs undergoing BAS surgery. Univariate analysis identified postoperative radiographic evidence of pneumonia, the perioperative use of a metoclopramide CRI, and longer overall surgical time as associated with development of a major complication after surgery. However, multivariate analysis identified only 1 risk factor, postoperative radiographic evidence of pneumonia, which increased the risk of development of any major overall or individual postoperative complication. The results of the multivariate analysis suggest that the use of a metoclopramide CRI in the perioperative period and a longer overall surgical time may represent confounding factors. These results support acceptance of our hypothesis that 1 or more factors associated with major complications would be identified.

There were differences in breed distribution in our study population compared to previous reports on outcome following BAS surgery in dogs (1,4,7). The most common breed reported in our study was pugs (32%), which was higher than in previous reports (9% to 21%), and may represent a difference in regional breed distribution (1,4,7). Mean age has been previously reported as 30 to 38 mo with a range of 2.5 to 189 mo (1,4,6,7,9). The present study reported a mean age of 53 mo, which is higher than in previous reports, but was not found to be a significant risk factor for complications. A previous report hypothesized that older dogs may have more significant complications due to secondary changes, such as laryngeal collapse (9); however, this was not found in our results or in another study (10). A higher incidence of dogs with everted laryngeal saccules (i.e., grade 1 laryngeal collapse) was noted in the present study (78%), compared to historic reports with ranges of 40% to 66% (1,11). Although not confirmed in the results of this study and difficult to evaluate due to its retrospective nature, the higher incidence of everted laryngeal saccules may have been related to the increased mean age in these dogs, which may align with the progressive nature of BAS.

Previous postoperative complication rates have been reported between 6% and 26% (1,4,7,9). Two of these reports separated major from minor complication and their data indicated that major complication rates ranged from 5% to 8% (1,7). The other 2 reports only identified major complications, including temporary tracheostomy and death, ranging from 6% to 14% (4,9). These reported major complication rates are comparable to our overall major complication rate of 7%.

Postoperative radiographic evidence of pneumonia was documented in 5% of the dogs in our study population. In a 1997 study, Lorinson et al (9) reported aspiration pneumonia affecting 11% of the dogs in the study as the most common reason for death or euthanasia after surgery. However, aspiration pneumonia has not been identified as a postoperative complication in more recent studies. Improvement in management of gastrointestinal disease in dogs undergoing BAS surgery has been proposed as the cause for reduction in incidence of aspiration pneumonia (1,4,7). Also, the lack of aspiration pneumonia in this study may show that other respiratory disease may be associated with BAS and warrant further investigation.

Dogs with BAS have signs of gastrointestinal disease that may predispose them to being treated with prokinetic and anti-emetic medications such as metoclopramide in the perioperative period (7,8). Gastrointestinal disease can also lead to a higher risk of aspiration pneumonia following surgery in BAS dogs (7,9). The present study found that perioperative use of a metoclopramide CRI was significantly associated with an increased risk of development of a major complication. Three of the 5 dogs which had metoclopramide CRI in the perioperative period developed a major complication. This study also evaluated a variety of other perioperative medications, including preoperative sedatives and analgesics, intraoperative gas inhalants, and postoperative gastrointestinal, analgesic, and sedative medications. These drugs had not been specifically evaluated previously but were not found to be significantly associated with the development of major complications. Most full opioid agonists can cause gastrointestinal signs, such as vomiting and nausea (12). Vomiting and/or regurgitation occurred in 24% of dogs, which is higher than in previous studies (7% to 18%) but was not found to be associated with the development of major complications (1,7). The infrequent use of a metoclopramide CRI, despite the high frequency of gastrointestinal signs found in this study, is likely related to variations in individual managing clinician preferences. The use of gastrointestinal medications has been supported by previous publications documenting gastrointestinal disease in dogs with BAS (7,8). However, the limitation in this study was the lack of characterization of the gastrointestinal signs to correlate the severity of these signs to the use of metoclopramide CRI’s.

This study identified an increase in overall surgical time as a risk factor for placement of a temporary tracheostomy. Logically, dogs having additional surgical procedures performed at the time of BAS surgery have increased overall surgical time. Previous studies have reported that longer surgical times are associated with increased risks of postoperative complications, although not specifically for BAS surgeries. For example, the risk of surgical wound infection has been documented to be increased with longer durations of surgery (13). In another study, risk of infection doubled for every 70 min of surgery (14). Nicholson et al (15) found that dogs that developed a postoperative wound infection in clean-contaminated procedures had significantly longer surgical times. The findings in the present study are consistent with these reports, and suggest that limiting additional procedures in dogs with BAS surgery should be considered to reduce postoperative complication rates.

A history of prior BAS surgery was not associated with the development of major complications post-surgery. There was also no difference based on whether the previous surgery was performed at the study institution or by the primary care veterinarian. These are not factors that had been evaluated previously (1,4,7). Due to the low incidence of major complications experienced in this study, these findings may represent a type 2 statistical error and could be valid points of investigation for future, larger studies.

There was a statistical concordance identified in the study population between dogs requiring a temporary tracheostomy after surgery and those that died or were euthanized in the postoperative period. The dogs that required a temporary tracheostomy after surgery were the same dogs that had significant major complications, generating a selection bias for this subpopulation and may suggest that other factors, such as pneumonia or other chronic respiratory disease, were responsible for their poor outcome. Due to the retrospective nature of this study, the details of the timing and factors driving the decision to place the temporary tracheostomies in relation to death or euthanasia are unclear. Preoperative placement of a temporary tracheostomy prior to surgery in dogs at risk for respiratory compromise, such as grade III laryngeal collapse, may prevent major complications in dogs undergoing surgery for upper airway obstruction (16).

Only 1 dog in the present study population had grade III laryngeal collapse. This dog had a left-sided cricoarytenoid lateralization along with caudal staphylectomy and sacculectomy and also required a temporary tracheostomy after surgery. Dogs evaluated for grades 2 and 3 laryngeal collapse have been reported to have a 83% to 100% survival rate in the postoperative period if a temporary tracheostomy was performed concurrently with a left-sided cricoarytenoid lateralization (6,16). Unlike these previous reports, this dog was only responsive to mechanical ventilation. He was euthanized despite these medical and surgical interventions and was suspected to have suffered systemic and pulmonary inflammation.

The primary limitations of this study are the low rate of major complications and limitations typical for retrospective studies (1,4,7). This emphasizes the need for larger studies to be conducted in the future to obtain a more representative sample of this subpopulation of dogs undergoing BAS. The low sample size for the subpopulation of dogs that experienced a major postoperative complication is reflected in the wide 95% CI ranges reported herein. Another limitation is the lack of characterization of GI signs in these patients, which could contribute to pneumonia but were not able to be correlated in this study. Limitations inherent to retrospective studies include: reliance upon medical records with incomplete information, non-standardized treatments or descriptions, and incorrect determination of case outcomes. Despite these limitations, the present study identified certain factors associated with major complications which represent valuable topics for future investigation via prospective clinical studies designed to specifically track these variables in a larger population of dogs undergoing BAS surgery.

In conclusion, this study identified specific patient- and procedure-related factors associated with major overall or individual complications following BAS surgery in dogs. Although a number of other variables were found to be significant based on univariate analysis, after multivariate regression analysis was performed, only postoperative radiographic evidence of pneumonia was identified as a significant risk factor for major complications. Future prospective studies could evaluate management strategies, such as prokinetic therapy, that can improve outcomes for dogs undergoing BAS surgery.