In bitches, is ovariectomy/ovariohysterectomy by laparoscopy less painful postoperatively than by midline open laparotomy?

a Knowledge Summary by

Chris Webb MA VetMB PgC(SAS) PgDip(VCP) MRCVS ^1*

Julia Deutsch DipECVAA MRCVS ²

¹Fitzpatrick Referrals Ltd., Halfway Lane, Eashing, Godalming, GU7 2QQ
²Langford Vets, Small Animal Referral Hospital, University of Bristol Vet School, Langford House, North Somerset, BS40 5DU
^*Corresponding Author (cwebb@fitzpatrickreferrals.co.uk)

Clinical scenario

During a rabies vaccination appointment, you discuss with your client about the benefits of spaying her young French bulldog bitch in the near future. She enquires about a keyhole approach as your practice started performing laparoscopic spays last year. Her previous practice was not in support of the technique as the veterinary surgeons had many years of great success with small incisions. The client feels her bitch is very sensitive however, so would like to know if there is any evidence that she will be in any less pain following a laparoscopic approach.

The evidence

Seven studies were identified that compared postoperative pain in bitches spayed by laparoscopy with bitches spayed by a traditional midline laparotomy. They comprised of three blinded randomised controlled trials, two non-blinded randomised controlled trials and two non-blinded non-randomised controlled trials. One paper provided weak-moderate evidence and the remaining six papers provided weak evidence to support laparoscopy as being superior to an open midline laparotomy in causing less postoperative pain.

Summary of the evidence

Appraisal, application and reflection

A total of seven papers were appraised, comprising three blinded randomised controlled trials, two non-blinded randomised controlled trials and two non-blinded non-randomised controlled trials.

Whilst they supported the use of laparoscopy (to varying degrees) over traditional laparotomy by causing less postoperative pain, small patient numbers and the large variability between studies (and in some cases, groups) were significant constraints to the strength of evidence. Although each may be considered minimally invasive relative to their comparative (control) open procedure, the difference in laparoscopic technique across studies (i.e. port number, trocar size, method of access for pneumoperitoneum, insufflation pressures and variation in method of haemostasis) makes consideration as a single group very difficult. It is also worth noting that none of the referenced pain scoring systems used in these studies are validated for ovariectomy/ovariohysterectomy pain, but are commonly used in canine studies as a measure of postoperative pain.

In humans, laparoscopic techniques have been shown to cause less postoperative pain, shorten hospital admissions and ensure faster return to normal daily activity when compared with open surgery. As such, it is currently considered the gold standard approach for many gynaecological surgeries (Rajvinderet al. 2018). A common anatomy and physiology in pain processing between vertebrates (Pelligand & Mora, 2016) would lead us to believe that a laparoscopic approach is also less painful in dogs. It is likely that smaller incisional wounds, as well as reduced tissue handling of sensitive structures, would be responsible for any difference seen in these patients.

Underpowered studies (all seven studies here may be considered such) are typically associated with increased risk of false negative results (type II error) but will also increase the likelihood that statistically significant findings are false positives (type I error). Though either scenario would render any results misleading in the context of answering the clinical question, prior knowledge of human laparoscopic surgery would suggest that the probability of the null hypothesis (i.e. no significant difference in postoperative pain scores exists between laparoscopic and open ovariectomy/ovariohysterectomy) being true, is low. This would imply that there may still be some utility in significantly improved laparoscopic group pain scores.

The lowest risk of bias was found in the well-designed blinded randomised controlled trial by Hancock et al. (2005). Despite no age data, the purpose-bred population was considered homogeneous ensuring no difference between groups. Dogs were housed for 96 hours prior to surgery to limit the effects of stress on postoperative behaviour changes and pain scores. Although three different surgeons were responsible for intervention, block randomisation ensured that no surgeon performed an unequal number of surgeries in each group. A multidimensional composite pain scale (University of Melbourne pain scale (UMPS)) was combined with an objective nociceptive threshold measurement technique and assessed by a single (blinded) observer, removing interobserver bias. Statistics were performed appropriately, and confidence intervals recorded. Results showed significantly lower pain scores and higher nociceptive thresholds in dogs operated by laparoscopy. The strength of evidence is considered weak to moderate and could principally have been strengthened by larger group sizes.

Dalmolin et al. (2020) used blinded randomised controlled trial design with three pain scoring systems (including two multi parameter systems) but the benefit of multiple assessments was negated by using three pain assessors with only light to moderate agreement between them. Anaesthetic protocol was well controlled and a single surgeon was used for all procedures. The only significant differences in pain scores between groups showed the laparoscopic group to be more comfortable. There were, however, individual time points on each of the pain scoring systems where pain scores were lower for the laparotomy group. Significantly faster return to voluntary feeding was noted for the laparoscopy group which has previously been suggested as a sign of reduced postoperative pain (Sarrau et al., 2007). Unlike the other studies, ongoing pain medication was provided throughout the recovery period which would reduce differences in pain scores between groups.

The benefit of laparoscopy over laparotomy appears limited using this analgesic combination but, again, this is an underpowered study. As such we believe that this study only provides weak evidence in favour of laparoscopy.

Devitt et al. (2005) used a non-blinded, randomised controlled trial design but with few other limitations. The main source of bias was in using two non-blinded observers to perform pain scoring. Reporting of breeds may have been beneficial but the weight and age of animals were well controlled. Standardisation of procedures was particularly rigorous (although there was limited description of the laparotomy technique), with the same surgeon and anaesthetic protocol for all animals. Statistical analysis was appropriate, with recording of confidence intervals. Pain scores and requirement for rescue analgesia (as well as neuroendocrine markers for some time points) were significantly higher in the dogs operated by laparotomy. The relative risk for rescue analgesia in the laparotomy was 10 fold, but the 95% confidence interval was enormous (1.6 to 64.2) indicating an underpowered study. The pain scoring system was adapted from a system validated for use in human paediatric patients, not dogs. We therefore consider the strength of evidence in favour of laparoscopy to be weak.

A non-blinded, non-randomised controlled trial design was used by Freeman et al. (2010). Animals in both laparotomy and laparoscopy groups were from a local animal shelter and comparable in weight (though breeds and ages were not stated). Anaesthetic protocol was reportedly standardised (details of protocol not shown) but surgical team was unclear. Like Hancock et al. (2005), a combination of UMPS and nociceptive threshold testing was used, but two observers limit this strength. Reporting of statistics between laparoscopy and laparotomy was limited. The power calculation (the only study to report one) significantly overestimated the effect of the intervention group on reduction of IL-6, c-reactive protein (CRP), cortisol and increase in nociceptive threshold readings, this demonstrates the study is considerably underpowered.

Significantly lower pain scores and higher nociceptive thresholds were found in the laparoscopy group at certain time points. This may easily be missed as it is not reported in the results or discussion but is derived from the figures. We would consider this to be weak evidence in favour of laparoscopy.

Coutinho et al. (2018) used a blinded randomised controlled trial design but several limitations were encountered; principally two observers were responsible for pain scoring. Both UMPS and a visual analogue scale (VAS) were used, with the latter known to be more subjective (Pelligand & Mora, 2016). The groups were assessed thoroughly before inclusion to ensure they were healthy, but breeds were not reported. Significantly faster return to eating (and in greater proportion) was found in the laparoscopy group, but this behaviour is not captured by any pain scoring system. Postoperative glucose concentrations (which Walsh et al. (1999) report to be a possible marker of stress caused by discomfort) were also significantly lower in dogs in this group, suggesting they may be more comfortable postoperatively than dogs operated by laparotomy. Although it was reported that no physiological parameters changed (suggesting no change in nociception) during the two complications in the laparoscopic group, the elimination of these from follow-up was an additional weakness. Given the lack of significant difference in pain scores combined with the above weaknesses, we cannot see this study being used as evidence to support use of laparoscopy over laparotomy to reduce postoperative pain.

Though Vasiljevic et al. (2015) also used a non-blinded, randomised controlled trial design, there was moderately high risk of bias. Aside from study design, a significant factor limiting the strength of evidence was the lack of recording. Species and approximate breed size of animals were stated, but we have no other information on group characteristics so are unable to confirm if they were comparable. The minimum patient age was 7 months and the smallest patient weight was 5 kg, which does not fit with the description of medium and large breed dogs. Surgical team was not specified nor surgical techniques described, but there was a similar induction and maintenance protocol between groups. Additional analgesia was used for all patients in the laparotomy group but none in the laparoscopic group, which would bias the laparotomy group to lower postoperative pain scores (which was not the case). A single postoperative pain scoring method was used by one observer, based on four parameters to give a score from 0 to 9 (incorrectly presented in figures as 0 to 10). The parameters had similarities to the short form composite measure pain scale (CMPS-SF), but the description was brief and not referenced to other studies so was difficult to compare. Statistics were lacking but there is clear difference in pain scores between groups – much higher for dogs operated by laparotomy. This may, however, be a result of the reported slower recovery from general anaesthesia in the laparotomy group. Though unrelated to strength of evidence, ethics are of concern. The study was approved by the local ethics committee but no intervention level was set for rescue analgesia during the 6 hour assessment period. Two dogs were considered to be in severe pain at 1, 3 and 6 hours but analgesia was withheld until after the final pain score assessment. Overall concerns with reporting bias are significant and results must be viewed with skepticism and therefore this study cannot be used to answer the question in the clinical scenario.

A truly non-blinded non-randomised study was used by Davidson et al. (2004). The groups may not be comparable as there is no breed data, health status was not clear, and they were derived from different populations. In addition, there was no standardisation of anaesthetic protocol or surgical team between groups. The same surgeon was responsible for laparoscopic surgery, but multiple veterinary students (who are likely to have poorer tissue handling) performed open surgery, increasing bias in favour of laparoscopic techniques. We see that surgical times for OHE were as long as 140 minutes which is not consistent with what is typically expected* for this procedure. Equally, laparoscopic surgical times were significantly longer than expected, leading us to believe that this is a novel technique for the surgeon and the outcomes therefore unreliable. Ketamine (6 mg/kg IV) was used preoperatively for four of the dogs undergoing laparoscopic surgery and only one of the dogs receiving open surgery. Preoperative use has been shown to confer significant postoperative analgesic effects (Slingsby & Waterman‐Pearson, 2000), so the results here are further biased. Unique to this study, a 4-port technique with a surgical wire ligation device was used. These techniques are not applicable for ovariectomy or ovariohysterectomy in general practice. *Typical surgical times for laparotomy or laparoscopy would be 30–60 minutes depending on patient size.

Although they report a number of parameters within the subjective pain scale and UMPS to be significantly higher for open surgery, this belies the lack of significant difference in overall scores. Overall, this study cannot be used to answer the question in the clinical scenario.

Even in human surgery, we must remember that a laparoscopic approach is not without problems.­ Insufflation pressures above 15 mmHg, unnecessary pressure peaks and prolonged insufflation with CO₂ have been shown to contribute to pain after laparoscopy in humans (Moulton et al., 1999).

In dogs, there is a suggestion that the rate of complications are reduced for laparoscopic ovariectomy compared with open ovariectomy (Charlesworth & Sanchez, 2019), but experience in using the technique will affect the outcomes (Pope & Knowles, 2014) along with surgical time. None of these factors, however, have been critically reviewed here but would require careful consideration when making recommendations for this technique.

Despite the widely held belief that laparoscopic surgery is associated with less postoperative pain, the available veterinary literature only provides weak evidence to support this in bitches undergoing ovariectomy/ovariohysterectomy. There are many uncontrolled variables to consider across the underpowered studies here including surgeon number (and experience), the choice of perioperative analgesia, method of pain scoring and the laparoscopic technique. It is therefore clear that laparoscopic procedures cannot be viewed equally and the strength of the answer to the clinical question may change based upon these variables.

Methodology Section

Search Strategy
Databases searched and dates covered:	CAB Abstracts on OVID Platform 1973–2020 Week 15 Ovid Medline(R) 1946–June 2020
Search strategy:	Search performed was the same for each database: dogs/ (dog* OR canine* OR pupp* OR bitch*).mp 1 OR 2 (laparoscop* OR celioscop* OR coelioscop* OR peritoneoscop* OR NOTES OR HALO OR keyhole OR “key-hole”).mp (laparotom* OR celiotom* OR coeliotom* OR open OR traditional OR conventional).mp 4 AND 5 (spay* OR spey* OR ovarectom* OR ovariectom* OR ovariohysterectom* OR oophorectom* OR OVE OR OHE OR OVH OR neuter* OR desex* OR sterilis* OR steriliz* OR gonadectom*).mp 3 AND 6 AND 7
Dates searches performed:	30 June 2020

Exclusion / Inclusion Criteria
Exclusion:	Not available in English. Reviews, letters, book chapters, case reports or conference proceedings. Species other than canines. No report of outcomes relevant to postoperative pain. Studies evaluating different laparoscopic techniques only. Papers comparing laparoscopy to open flank approaches. The NOTES technique was not considered as a comparison to open procedures as it is not a technique that is currently widely used for neutering in general practice – papers were not excluded if they still compared a standard laparoscopic technique with open surgery.
Inclusion:	Comparison of midline open ovariectomy/ovariohysterectomy with laparoscopic ovariectomy/ovariohysterectomy. Laparoscopic techniques and methods considered reasonable in general practice. This included any number of direct abdominal ports for access, as well as the use of harmonic scalpel, mono- or bi-polar electro surgery and suture, wire or clip ligation. Studies that used a postoperative pain scoring system.

Search Outcome
Database	Number of results	Excluded – [Not English] Language	Excluded – [Review, case re-port or conference proceedings]	Excluded – [Did not answer PICO]	Total relevant papers
CAB Abstracts	92	29	14	43	6
Medline	44	1	5	34	4
Total relevant papers when duplicates removed					7

The authors declare no conflicts of interest.

1. Charlesworth, T. M. & Sanchez, F. T. (2019). A comparison of the rates of postoperative complications between dogs undergoing laproscopic and open ovariectomy. Journal of Small Animal Practice. 60(4), 218–222. DOI: https://doi.org/10.1111/jsap.12993
2. Coutinho, A. J., Gasser, B., Rodriguez, M. G. K., Uscategui, A. A. R., Santos, V. J. C., Tiosso, C. de F., Barros, F. F. P. da C. & Toniollo, G. H. (2018). Comparison between single port videolaparoscopy and miniceliotomy with snook hot ovariohysterectomy techniques in bitches. Ciênca Rural. 48(10), e20180345. DOI: https://doi.org/10.1590/0103-8478cr20180345 
3. Dalmolin, F., Oliveira, M. T., Filho, S. T. L. P., Vaz, M. A. B., de Cecco, B. S., Feranti, J. P. S., Carvalho, E. R., Silva, M. A. M. & Brun, M. V. (2020). Dipyrone, Scopolamine, and meloxicam for conventional or two-port laparoscopic assisted ovariohysterectomy in female dogs. Semina: Ciências Agrárias (Londrina). 41(3), 887–896. DOI: https://dx.doi.org/10.5433/1679-0359.2020v41n3p887
4. Davidson, E. B., Moll, H. D. & Payton, M. E. (2004). Comparison of Laparoscopic Ovariohysterectomy and Ovariohysterectomy in Dogs. Veterinary Surgery. 33(1), 62–69. DOI: https://doi.org/10.1111/j.1532-950X.2004.04003.x
5. Devitt, C. M., Cox, R. E. & Hailey, J. J. (2005). Duration, complications, stress, and pain of open ovariohysterectomy versus a simple method of laparoscopic-assisted ovariohysterectomy in dogs. Journal of the American Veterinary Medical Association. 227(6), 921–927. DOI: https://doi.org/10.2460/javma.2005.227.921
6. Freeman, L. J., Rahmani, E. Y., Al-Haddad, M., Sherman, S., Chiorean, M. V., Selzer, D. J., Snyder, P. W. & Constable, P. D. (2010). Comparison of pain and postoperative stress in dogs undergoing natural orifice transluminal endoscopic surgery, laparoscopic, and open oophorectomy. Gastrointestinal Endoscopy. 72(2), 373–380. DOI: https://doi.org/10.1016/j.gie.2010.01.066
7. Hancock, R. B., Lanz, O. I., Waldron, D. R., Duncan, R. B., Broadstone, R. V. & Hendrix, P. K. (2005). Comparison of postoperative pain after ovariohysterectomy by harmonic scalpel-assisted laparoscopy compared with median celiotomy and ligation in dogs. Veterinary Surgery. 34(3), 273–282. DOI: https://doi.org/10.1111/j.1532-950x.2005.00041.x
8. Khasriya, R., Vashisht, A. & Cutner, A. (2018). Laparoscopy in urogynaecology. The Obstetrician and Gynaecologist. 20(2), 101–108. DOI: https://doi.org/10.1111/tog.12449
9. Mouton, W. G., Bessell, J. R., Otten, K. T. & Maddern G. J. (1999). Pain after laparoscopy. Surgical Endoscopy. 13(5), 445–448. DOI: http://dx.doi.org/10.1007/s004649901011
10. Pelligand, L. & Mora, S. S. (2016). Pain assessment methods. In: Duke-Novakovski T., de Vries M. & Seymour C. (Eds.) BSAVA Manual of Canine an Feline Anaesthesia and Analgesia, Third Edition. Gloucester, UK. British Small Animal Veterinary Association, pp113–123.
11. Pope, J. F. A. & Knowles, T. G. (2014). Retrospecitve Analysis of the Learning Curve Associated With Laparoscopic Ovariectomy in Dogs and Associated Perioperative Complication Rates. Veterinary Surgery. 43, 668–677. DOI: https://doi.org/10.1111/j.1532-950X.2014.12216.x
12. Sarrau, S., Jourdan, J., Dupuis-Soyris, F. & Verwaerde, P. (2007). Effects of postoperative ketamine infusion on pain control and feeding behaviour in bitches undergoing mastectomy. Journal of Small Animal Practice. 48(12), 670–676. DOI: https://doi.org/10.1111/j.1748-5827.2007.00362.x
13. Slingsby, L. S. & Waterman‐Pearson, A. E. (2000). The postoperative analgesic effects of ketamine after canine ovariohysterectomy—a comparison between pre and postoperative administration. Research in Veterinary Science. 69(2), 147–152. DOI: https://doi.org/10.1053/rvsc.2000.0406
14. Vasiljević, M., Ristanović, D., Jovanović M., Davitkov, D., Bošnjak, I., Krstić, V. & Stanimirović, Z. (2015) Comparative analysis of parameters of intraoperative and postoperative pain in bitches undergoing laparoscopic or conventional ovariectomy. Acta Veterinaria-Beograd. 65(4), 488–495. DOI: https://doi.org/10.1515/acve-2015-0041
15. Walsh, P. J., Remedios, A. M., Ferguson, J. F., Walker, D. D., Cantwell, S. & Duke, T. (1999). Thoracoscopic versus open partial pericardectomy in dogs: Comparison of postoperative pain and morbidity. Veterinary Surgery. 28(6), 472–479. DOI: https://doi.org/10.1111/j.1532-950x.1999.00472.x