In female dogs undergoing elective neutering is ovariectomy or ovariohysterectomy superior? A Knowledge Summary by

An 11-month-old, female entire crossbreed dog presents for elective neutering. You discuss surgical techniques with the client. You are confident at performing both ovariectomy and ovariohysterectomy via midline coeliotomy. Other techniques, including minimally invasive surgery and surgery through a flank incision, are not locally available or financially feasible and so are not considered. considering anaesthetic duration, complication rate and postoperative to OVE is to OVH or vice versa?


The evidence
From the literature reviewed here there is some evidence to suggest that when performed by an experienced veterinary surgeon OVE is associated with a shorter incision length and reduced surgical time compared to OVH. However the studies reviewed differ markedly in case selection, technique and study objectives. The sample sizes are typically small and the experience of the operating veterinary surgeons may not accurately reflect an 'average' general practitioner. Additionally the differences in between procedures reported may not be clinically significant.
There is convincing evidence that, when properly performed, OVE is not associated with a risk of pyometra. There is currently insufficient evidence to determine if there is a difference in postoperative pain following OVH compared to OVE. No papers reported an improvement in any outcome with OVH compared to OVE.
In conclusion, whilst the evidence does suggest OVE may be associated with some modest improvement in surgical time and incision length, further studies are required before definitive conclusions can be made.

Summary of the evidence 1. Harris et al. (2013)
Population: Female dogs of various breeds presented to a British veterinary teaching hospital for elective neutering. Age range 6-120 months. Weight range 2.9-51.5 kg. Dogs were excluded from recruitment if there were signs of pregnancy, oestrus, pseudopregnancy or ill health on clinical examination. Dogs were excluded at the time of surgery if pregnancy or abnormalities of the reproductive system were visualised or palpated.

Sample size: 108 dogs
Intervention details:  Dogs were randomly allocated to undergo OVE (n=54) or OVH (n=54) via midline coeliotomy.  A final year student was allocated to each case and given written instructions of the procedure to be performed based on a standard open protocol.
 The procedures were performed by the final year student assigned to the case with a qualified veterinary surgeon assisting, this assistant would complete the procedure if the total surgical duration exceeded 2 hours or major complications occurred.  Two dogs in the OVH group experienced major complications (one bladder laceration, one pedicle rupture prior to ligation) which required the assistant veterinary surgeon to complete the procedure, these were not included in the final analysis.  11/108 cases exceeded the 2 hour time limit, 8/51 of the OVH group and 3/53 of the OVE group -this was not statistically significant.  Mean total surgical time was 88.7 +/-20.6 minutes in the OVE group and 92.0 +/-27.6 minutes in the OVH group. This was not significantly different.  Mean incision length was 8.7 +/-2.6 cm in the OVE group and 9.6 +/-3.4 cm in the OVH group. This was not significantly different.  No significant difference was observed in time from first incision to start of closure or time of closure between groups.  Minor intra-operative complication rates occurred in 12/52 (23%) of the OVH group and 21/54 (39%) of the OVE groupthis did not differ significantly between groups.

Limitations:
 This is a single centre study which may limit application to other centres.  Surgeries were performed by final year students and so results may not be applicable to experienced veterinarians.  Students may have been more familiar with OVH from time in general practice where OVH is more commonly performed.  An assistant veterinary surgeon more experienced in OVH was scrubbed in to procedures.  Analgesia used is not recorded.  The text instructions supplied recommended a set incision length with extension at the discretion of supervising veterinary surgeon.  Incision length was recorded as an absolute length rather than as a proportion of the dog's length.  Different supervising veterinary surgeons of different skill levels were assisting and this was not controlled between groups.  There was no assessment of difference in skill levels of students, this was likely to be variable and was not controlled between groups.

Limitations:
 This is a single centre study which may limit application to other centres.  All procedures were performed by a single board certified veterinary surgeon which may limit application to general practice.  Wound appearance was assessed by subjective methods.  Collection of blood via a jugular catheter was performed at 1 and 6 hours post-surgery, and by venipuncture at 24 hours post-surgery, this may have affected pain scores.  Incision length was determined by the operating veterinary surgeon and may not reflect the required minimum incision length for each procedure.  Dogs received buprenorphine analgesia at 6 hourly intervals for 24 hours post-surgery. This is unlikely to reflect general practice and may have hindered ability to detect differences in pain scores between groups.

Tallant et al. (2016)
Population: Adult, female entire dogs obtained from local humane society shelters (country of origin not specified). Weight ranged from 3.3-30.1 kg. Dogs were excluded if there were signs of illness or cardiovascular abnormalities, evidence of oestrus or pregnancy on physical examination.

Sample size: 20 dogs
Intervention details:  Dogs were individually kenneled a minimum of 24 hours prior to surgery.  Dogs were randomly assigned to receive either OVE or OVH.  Anaesthetic protocol was standardised with carprofen (4 mg/kg) given once subcutaneously prior to surgery.  Surgery was performed as per a standardised technique via a median coeliotomy. Haemostasis was achieved via a vessel sealing device, this was also used to seal and divide the uterine body where this was under 9 mm, where the uterine body exceeded 9 mm a single circumferential ligature was applied prior to transection.  Aftercare was standardised. Rescue analgesia (buprenorphine) was administered to dogs with pain scores of 5/24 or greater.

Main findings: (relevant to PICO question):
 The change in mean arterial pressure between phase 1 and 2 was greater in the OVH group (increase of 25 +/-14 mmHg) than the OVE group (increase of 9 +/-15 mmHg).  The change in diastolic pressure between phase 1 and 2 was greater in the OVH group (increase of 27 +/-13 mmHg) than the OVE group (increase of 6 +/-14 mmHg).  There were no significant differences between groups in blood pressure between other phases, or heart rate and systolic pressure changes between any phases.  The mean heart rate of the OVE group was significantly greater than that of the OVH group during phase 0 and during phase 3.  The end-tidal isoflurane was significantly lower for the OVH group compared to the OVE group during phase 1 and phase 2. There was no difference during phase 3 or between the phases. between groups in the duration of phases 0-2, however, the duration of phase 3 was longer in the OVH group than the OVE group.  Mean skin incision length was significantly greater in the OVH group (6.4 +/-0.7 cm) compared to the OVE group (5.3 +/-1.1 cm).  One dog from each group required rescue analgesia -this was not significantly different.  Neither visual analog scores nor algometer readings were significantly different between groups.  There were no complications in any of the dogs intraoperatively or up to 24 hours postoperatively.

Limitations:
 This was a single centre study, which may limit the application of results to other centres.  All procedures were performed by the same veterinary surgeon which may limit application to other veterinary surgeons.  The operating veterinary surgeon was board certified which may not reflect general practice.  Incision length was reported as an absolute value rather than percentage of body length.  Pre-operative overnight kenneling is not reflective of a typical general practice setting.  The sample size was small, limiting power to detect differences between groups and increasing the effect of any individual variation.  Incision length was determined by the operating veterinary surgeon and may not reflect the required minimum incision length for each procedure.  A vessel sealing device was used during the procedure and so results may not be applicable to procedures using ligatures.

Lee et al. (2013)
Population: Female entire, purpose-bred, crossbreed dogs (country of origin not specified). Dogs were excluded if abnormalities were found on clinical examination or serum biochemical profile/complete blood count 24 hours prior to surgery.  Butorphanol (0.4 mg/kg) was administered intravenously prior to extubation, no further analgesia was given.  Blood sampling was performed by jugular venipuncture prior to surgery then at 1, 2, 4, 6, 12 and 24 hours post-surgery.  Aftercare was standardised.

Limitations:
 This was a single centre study which may limit application of results to other centres.  All procedures were performed by the same veterinary surgeon which may limit application to other veterinary surgeons.  Venipuncture was performed regularly throughout the study which may have affected pain scores.  Incision length was reported as an absolute value rather than percentage of body length.  The study population were purpose bred crossbreeds which may limit application to other breeds.  Pre-operative overnight kenneling is not reflective of a typical general practice setting.  The sample size was small, limiting power to detect differences between groups and increasing the effect of any individual variation. p a g e | 10 of 16  Incision length was determined by the operating veterinary surgeon and may not reflect the required minimum incision length for each procedure.  Dogs received one injection of butorphanol analgesia postoperatively only. This was in contrast to other studies where a non-steroidal anti-inflammatory drug (NSAID) was administered which may be more typical of general practice. This may account for the differences in pain score noted.

Okkens et al. (1997)
Population: Female dogs of various breeds which had undergone either an OVE or OVH procedure at a Dutch, teaching hospital 8-11 years prior to the study. Weight range at time of surgery 1.6-37.5 kg, age range at time of surgery 9.6 months to 9 years.

Sample size: 135 dogs
Intervention details:  Dogs had been randomly selected to receive either OVH (n=66) or OVE (n=69). Anaesthesia and surgical technique were standardised.  A questionnaire was sent to owners 8-11 years post-surgery.
Questions included whether the dog had experienced abdominal pain, vaginal discharge, endometritis/pyometra attractiveness to male dogs and urinary incontinence postsurgery. If any of the questions were answered positively a follow-up phone call was carried out.

Study design: Retrospective single centre, cohort study
Outcome studied: Incidence of urinary incontinence, ovarian remnant syndrome, attractiveness to male dogs, abdominal pain, vaginal discharge, endometritis/pyometra post-surgery.

Main findings: (relevant to PICO question):
 No attractiveness to male dogs or abdominal pain as a consequence of elective neutering was reported in any dog.  Asymptomatic vaginal discharge was observed in two dogs from either group.  Six dogs in the OVE group and nine in the OVH group developed urinary incontinence.  There was no incidence of endometritis/pyometra in either group.  There was no significant difference between groups in the incidence of any of the urogenital problems studied during the follow-up period.

Limitations:
 The method of randomisation was not reported and thus cannot be critiqued.  The questions asked were not reported and it is not clear if these may have biased client answers.  The study is retrospective and follow-up was performed.  Group matching (of age, weight etc.) was not possible and therefore confounding factors may have influenced results.  The study may be underpowered to detect differences in the rate of urinary incontinence between groups due to the relatively uncommon nature of the condition.  Due to the long follow-up time, accuracy of client recollection would be expected to have reduced.  The level of experience of the veterinary surgeons performing the procedures is not reported.  Bias may have been introduced by cases with negative outcomes being less likely to have complete case records and therefore being more likely to be excluded. . Likewise the group reported there was no significant difference in urinary incontinence between groups. The findings of this study must be considered with caution however, as there may be some inherent bias with dissatisfied owners less likely to provide follow-up and therefore to be included. Administration of additional analgesia beyond what is commonly used in general practice may hinder ability to detect differences in pain scores between groups, however withholding analgesia that is commonly used, may artificially increase differences in groups beyond what would be anticipated.  Lee et al. (2013) and Tallant et al. (2016) also recorded intra-operative complications, however none were noted in either group.

Appraisal, application and reflection
Gonadectomy in female dogs is generally performed to prevent misalliance and pyometra, for behavioural reasons and to reduce the risk of mammary and uterine neoplasia. Removal of the ovaries alone would be expected to prevent unwanted pregnancy, pyometra and to have an equal impact on the risk of mammary neoplasia as OVH, due to the cessation of ovarian hormone production. The effect of leaving the uterus in situ on the development of uterine neoplasia has not been evaluated, however as this disease is rare -the incidence of canine malignant uterine neoplasia is approximately 0.003% (Van Goetham et al., 2006) -and anticipated to be at least in part hormonally mediated, the impact of a change in practice from OVE to OVH on the uterine neoplasia related morbidity/mortality is likely to be negligible.
Of the studies reviewed here only two major complications were reported, both by Harris et al. (2013). Of these complications one, bladder laceration, would be expected to occur more commonly in OVH procedures due to the more caudal placement of the distal ligature. However, this study did not find a significant difference in complications between groups overall. It has also been proposed that OVE may be associated with a reduced risk of ureteral ligation; as the distal ureter is located caudal to the placement of the distal ovarian ligatures in OVE, but in the vicinity of the uterine ligature when an OVH is performed ( No difference in complication rates between groups has been shown and differences in postoperative pain scores between procedures have not been convincingly demonstrated. No papers identified an advantage of performing OVH over OVE. These results apply only to animals with grossly normal uteri at coeliotomy, and hysterectomy is still recommended when uterine pathology is present.
Further indicated research includes large scale studies allowing identification of differences in rare intraoperative complications and randomised, clinical trials in larger numbers of animals in a setting more reflective of general practice, to determine if a significant difference between procedures is found when power is increased.  Total relevant papers when duplicates removed