Hysterectomy: abdominal, vaginal or laparoscopic?

Hysterectomy is the commonest major gynaecological procedure1,2. The optimal route of a hysterectomy will depend on several factors. The clinical indication; uterine size; uterine descent; presence of endometriosis, pelvic adhesions or adnexal masses; previous pelvic surgery; surgeon’s and client’s preference are among them. Although there are many methods of performing a hysterectomy, primarily there are three approaches; laparotomy, vaginal and laparoscopic routes.


Introduction
Hysterectomy is the commonest major gynaecological procedure 1,2 . The optimal route of a hysterectomy will depend on several factors. The clinical indication; uterine size; uterine descent; presence of endometriosis, pelvic adhesions or adnexal masses; previous pelvic surgery; surgeon's and client's preference are among them. Although there are many methods of performing a hysterectomy, primarily there are three approaches; laparotomy, vaginal and laparoscopic routes.
In some instances, the optimal route is obvious; e.g., if the uterus is larger than 20-weeks, a total abdominal hysterectomy (TAH) may be the method of choice. While in malignancy a more radical approach with other additional surgical procedures would be needed. In uterine prolapse, with coexistent cystocele or rectocele a vaginal hysterectomy and repair would be the best approach to handle all the problems. Therefore, the route is obvious in some instances and there is no need for a debate on the optimal route in these situations.
However, there is a ubiquitous group of gynaecological conditions with menorrhagia or dysmenorrhoea where the uterus is less than 14-weeks with no descent, for Sri Lanka Journal of Obstetrics and Gynaecology 2021; 43: [5][6][7][8][9][10][11][12][13][14][15] DOI: http://doi.org/10.4038/sljog.v43i1.7992 whom any one of the three main routes of hysterectomy can be applied. In such situations alternatives such as levonorgestrel-releasing intrauterine system (LNG-IUS) and endometrial ablation should also be considered. Hysterectomy was similar in terms of quality of life and psychological well-being when compared to LNG-IUS 3 . A systematic review found that hysterectomy is preferable to LNG-IUS and endometrial ablation in terms of clinical effectiveness and cost-effectiveness 4 . Despite being clinically effective and cost-effective hysterectomy is generally not considered an initial option owing to its invasiveness and the possibility of complications 4 .
However, in Sri Lanka, management in the form of LNG-IUS and endometrial ablation are virtually nonexistent due to financial constraints in the public sector. As a result, a significant proportion of these women who could have been managed with these methods also undergo hysterectomy. Although statistics for Sri Lanka are not collated it can be presumed that the hysterectomy rate is relatively high. Therefore, as hysterectomy is the commonest gynaecological major surgery, a Lower-Middle Income Country (LMIC) such as Sri Lanka must find out the Prof. D A Ranasinghe Memorial Oration 2020 optimal route in terms of cost-effectiveness for this ubiquitous group which can be operated in any one of the three main methods. In order to address this a multi-centre randomized controlled trial was done at District General Hospital, Mannar and North Colombo Teaching Hospital, Ragama 5 . Figure 1 and Table 1 show the CONSORT diagram and the basic characteristics of the study groups. Obtained from 'Randomized controlled trial on non-descent vaginal hysterectomy and total laparoscopic hysterectomy versus total abdominal hysterectomy: a cost-effectiveness analysis' submitted to the Journal of South Asian Federation of Obstetrics and Gynaecology.
The optimal route of hysterectomy was studied in terms of operative time, hospital stay, complications, time to recover, quality of life, pelvic organ function and cost etc. which will eventually determine cost-effectiveness.

Operative time
Vaginal hysterectomy has a shorter operative time compared to laparoscopic hysterectomy and abdominal hysterectomy 6 . Data from studies involving LAVH or TLH versus TAH all suggest a shorter operative time for the latter although this may become insignificant when the learning curve has been overcome [6][7][8] Table 2).

Post-operative hospital stay
It is generally accepted that VH has a shorter postoperative hospital stay compared to TAH 6,9,10 . Similarly, a shorter hospital stay is seen for the laparoscopic route when LAVH or TLH is compared with TAH 6-11 . There is conflicting evidence in terms of post-operative hospital stay between VH and LH with Morelli et al, suggesting a significant difference in hospital stay between LH and VH in favour of LH, and data from eVALuate trial and Ottosen not showing a significant difference between the two approaches 9,12,13 . Our study  (Table 2).

Complications
Trials by Benassi and Ottosen did not show a difference in the incidence of urinary tract, bowel and vascular injuries between VH and TAH 6,9 . However all these studies were under-powered to detect a significant difference. Even a systematic review on the approach to hysterectomy by Nieboer et al was under-powered to detect a difference in visceral and vascular injuries between the two approaches, although fewer febrile episodes or unspecified infections [OR 0.42 (95% CI (0.21-0.83)] was observed in VH 10 . A lower postoperative pain score for TAH was observed by Silva-Filho AL, when VH was compared to TAH for women undergoing hysterectomy for uterine myoma 14 .
The abdominal arm of the eVALuate trial, designed to compare effects of laparoscopic hysterectomy with TAH showed a higher rate of major complications for laparoscopic hysterectomy [11.1% vs 6.2%, P=0.02] 13 . Most studies were under-powered to detect a difference in urinary tract injuries between LH and TAH when analysed separately. However when results were pooled together the risk of either a bladder or ureteric injury was significantly more in LH compared to   10 . TLH appears to be less significantly painful in both the immediate post-operative period and even after convalescence compared to VH 12 .
Our study was also under powered to detect a difference in complications between the three routes ( Table 2). In terms of major complications; there were three cases in the NDVH group and two in the TLH group which had to be converted to laparotomy. There were two bladder injuries in the NDVH group with one ureteric injury in the TLH group. One patient had a rectal injury in the NDVH group. There was no significant difference in the blood loss or change in haematocrit (pre-op -post-op) between the three routes.
In terms of post-operative pain, TLH group had a lower pain score on the first two days after surgery ( Table 2).

Time to recover and quality of life
When considering time to recover in days, VH and LH performs better than TAH, with a shorter recovery 9 .
There is no significant difference between VH and LH when considering time to recovery 9   NDVH -non-descent vaginal hysterectomy, TAH -total abdominal hysterectomy, TLH -total laparoscopic hysterectomy Obtained from 'Randomized controlled trial on non-descent vaginal hysterectomy and total laparoscopic hysterectomy versus total abdominal hysterectomy: a cost-effectiveness analysis' submitted to the Journal of South Asian Federation of Obstetrics and Gynaecology.

Cost-effectiveness
Typically the cost-effectiveness ratio that compares two alternatives is calculated as the difference in costs (net costs) between the alternatives divided by the difference in outcomes (net effectiveness) 19 . In the eVALuate trial, the incremental cost per QALY gained was $ 471,789 between LH and VH 11 . When considering the abdominal arm of eVALuate trial, an incremental cost per QALY gained of $ 46893 between LH and TAH 11 . Similarly the incremental costs for reducing one patient with major complications between LH and TAH was $35750 17 . Cost- effectiveness trials between VH and TAH were not found in literature.
Our trial used time to recover as the denominator to calculate cost-effectiveness. The incremental costeffectiveness ratio (ICER, 95% CI) for the study setting showed that it was $ 11 (TAH dominance to 351) for TLH. It was preferable to do a TAH instead of a NDVH [TAH dominance (TAH dominance to $ 477)]. The probability of cost-effectiveness at a threshold of 3 USD/day were 1.15% and 0% for NDVH and TLH.
The corresponding values at a threshold of 10 USD/ day were 14.1% and 4.2% for NDVH and TLH respectively (Figure 3).
It was preferable to do a TAH instead of either a NDVH or a TLH in the worst case scenario which considered probable complications and prolonged hospital stay with readmissions ( Table 3). The probability of costeffectiveness were unchanged at 0.2% and 0% for NDVH and TLH at a threshold of 3 USD/day and 10 USD/day (Figure 3).

Prof. D A Ranasinghe Memorial Oration 2020
The best case scenario which considered optimal conditions with no complications and readmissions showed that NDVH and TLH were both superior to TAH ( Table 3). The probability of cost-effectiveness at a threshold of 3 USD/day were 76.1% and 79% for NDVH and TLH. The probability of cost-effectiveness at a threshold of 10 USD/day were 76.3% and 79% for NDVH and TLH respectively (Figure 3).
This result illustrates the fact that alternate routes; NDVH and TLH would be superior to the conventional TAH in specialist centres whereas the generalist would be better off confining himself to the usual TAH.

Urinary function
A review on the effects of hysterectomy on vesicourethral function by Long confirms the hypothesis that hysterectomy is likely to improve lower urinary tract symptoms (LUTS) 20 . Roovers et al. found a higher prevalence of urge incontinence after VH compared with TAH and subtotal hysterectomy (STAH) 21 . Figure 3. Cost-effectiveness acceptability curves for study data, worst-case scenario and best-case scenario.
However this study was also an observational study which did not control for confounders such as asymptomatic prolapse. A prospective multi-centre observational study also found that a significant number of patients had been treated for micturition symptoms after VH compared to TAH [

Sexual function
An observational study over six months by Roovers et al. found a reduction in sexual problems after vaginal, total or subtotal hysterectomy 26 . Galyer et al. also evaluated differences in libido and genital sexual sensitivity and found that the type of hysterectomy made no difference at 12-months 27 . Radosa et al. did an observational cohort study on VH versus TLH versus supra-cervical laparoscopic hysterectomy (SLH) and found that there was a significant improvement post-hysterectomy regardless of the approach 28 .
Kluivers also confirmed these findings on a RCT between TLH and TAH 23 .
Vaginal and sexual symptoms were analysed using validated Sinhala and Tamil ICIQ-VS questionnaire 29

Bowel function
Prospective studies have failed to demonstrate an adverse bowel function after hysterectomy. Prior

Conclusion
There is a dilemma on how to assess clinical outcomes following a surgical procedure. Many indicators that have been used to compare outcomes for different routes of hysterectomy. Available evidence and study data suggests that TLH and NDVH is favourable to TAH in terms of post-operative hospital stay, time to recover, quality of life and cost-effectiveness. The operative time and cost appears to be more for TLH. There is inadequate data to suggest a difference in complications between the routes although early evidence suggested more urological injuries with TLH. There is also no difference in pelvic organ symptoms between the three routes. The study data further suggests that alternate routes would be preferable for specialist centres whilst the conventional TAH would be more suited for the generalist.
In an era where quality of care is an evolving area of interest, studies such as this which deals with postoperative convalescence are of importance, especially as it is from a low-resource setting. Patient reported outcome measures are vital as they consider the patient's viewpoint and functionality, something that cannot be assessed by clinical evaluation or investigations. One of the major findings was that there was an improvement in pelvic organ symptoms at 1-year with no significant difference between the three routes.
Adequate emphasis should be laid on cost; quality of life; and post-procedural convalescence in addition to conventional clinical indicators. A change in attitude among gynaecologists, appropriate allocation of resources, and refining postgraduate training are also required when selecting surgical options for women with benign uterine conditions in addition to patient factors and logistical factors in the health care system.
Prof. D A Ranasinghe Memorial Oration 2020