Complications after canine tonsillectomy without suturing were investigated from an intra- and postoperative perspective, and the most common complication was bleeding from the surgical site. Although no fatal complications were found in the study population, the identification of 10 minor and 2 major complications out of 39 cases raises concerns about the use of the technique.
Comparisons between earlier studies and the present study have to be done cautiously due to different patient groups, diagnosis, differences in duration of ongoing pathology and clinical signs. Both Cook et al.  and Belch et al.  performed tonsillectomy in brachycephalic dogs with obstructive airway syndrome amongst other surgical interventions such as staphylectomy, laryngeal sacculectomy and rhinoplasty. The most common complications, described in both studies were coughing, regurgitation, inappetence and laryngeal oedema. There is an increased risk of complications in brachycephalic dogs [9, 10]. In the present study, BOAS surgery cases were excluded.
Veterinary medical publications report a low risk for bleeding post tonsillectomy when using vessel sealing devices. This study found a higher incidence of bleeding, 11 out of 39 dogs, compared to previous studies using energy-based vessel sealing devices [5, 6]. In the study by Belch et al.  the LigaSure device had to be reapplied once in 20 tonsils. Cook et al.  also used the LigaSure device and reported no bleeding in 22 dogs. The present technique, with over 20 minutes of clamping of the tonsillar base, indicated a significantly longer duration of surgery compared to Belch’s use of LigaSure with less than one minute for seal and cut . The present study involved electrosurgery but use of the technique was not standardised, which is different compared to use of energy based vessel sealing devices such as LigaSure. Shorter surgical time and less bleeding are beneficial for the patient. Considering the results of the present study and the available data from other studies, a preference for the use of standardised energy-based techniques, vessel sealing devices, may be suggested for tonsillectomy. However, a randomized prospective study to evaluate advantages and drawbacks of each technique should be done.
Tonsillectomy is a common surgical procedure in humans. Considering species differences, comparison of human studies with veterinary medical studies should be done cautiously. Furthermore, surgical facilities and perioperative care may also differ substantially. Large retrospective studies of human data reported a wide range of complications such as bleeding from the surgical site, infections, other surgical site complications, dehydration, upper respiratory obstruction, postoperative pain and inadequate oral nutrition as commonly recognized complications [11–13]. A study by Seshamani et al.  reported postoperative complication rate for human tonsillectomies to be 20%, mainly haemorrhage, dehydration and pain, whereas another study by Chen et al.  reported a lower rate, 1.2%, mainly infections.
In several reported human case series, the rate of post tonsillectomy bleeding ranged from 2.0–7.0% [12, 14–21]. However, definitions varied slightly. Sarny et al. reported higher rates of post tonsillectomy bleeding, from 14–16% in four different studies [22–25]. There is conflicting evidence about surgical technique and equipment in studies evaluating the risk for post tonsillectomy bleeding. A meta-analysis by Francis et al.  reported that frequency of post tonsillectomy bleeding across different techniques did not differ. Krishna et al.  reported benefits for monopolar surgery but two other studies described higher risk of bleeding for “hot” tonsillectomy techniques, when diathermy was used for dissection, compared to traditional “cold steel” [21, 27].
The present study had several limitations. Surgeries were performed by different surgical teams and reports of complications were therefore inconsistent. The decision to treat intra- and postoperative bleeding was at the discretion of the surgeon and may have differed between surgeons, which influenced the classification of a complication. Furthermore, an incident may not be regarded as a complication, but rather an expected sequela of surgery. Other limitations were the small number of dogs, weight variation, age and type of dogs. An additional limitation was the lack of long-term follow-up; owners of the dogs were not contacted and potential late complications may have been treated at another clinic.
This study was not designed to evaluate or compare safety of different techniques. Due to the involvement of several different surgeons and the retrospective nature of the study, the results of the present study should be used cautiously in recommendation of any particular surgical technique. A prospective randomized study comparing several techniques is needed.