Objective
Recent randomized trials comparing total laparoscopic hysterectomy (TLH) and vaginal hysterectomy (VH) have produced conflicting results. The role of TLH in women suitable for VH remains uncertain.
Study Design
This study was a metaanalysis of randomized studies comparing TLH and VH for benign disease. Pooled outcome measures (odds ratio [OR] and weighted mean difference [WMD]) were calculated using random-effects models.
Results
No differences in perioperative complications, either total (pooled odds ratio, 0.87; P = .74) or by grade of severity, were demonstrated. TLH was associated with reduced postoperative pain scores (WMD −2.1; P = .03) and reduced hospital stay (WMD −0.62 days; P < .0001) but took longer to perform (WMD 29.3 minutes; P = .003). No differences in blood loss, rate of conversion to laparotomy, or urinary tract injury were identified.
Conclusion
TLH may offer benefits compared with VH for benign disease, although this analysis is likely underpowered for rare complications. Further studies of long-term outcomes, including prolapse, urinary incontinence, and sexual function, are required.
Hysterectomy is a commonly performed gynecological procedure. Although rates of benign hysterectomy in the United States have fallen in the last 2 decades, substantial numbers of women continue to undergo this operation annually. Seventy percent of hysterectomies are performed for benign indications, which include menorrhagia, fibroids, and uterine prolapse. Whereas abdominal and vaginal routes have traditionally been the most common surgical approaches, increasing numbers of hysterectomies are now performed laparoscopically. Several modifications to a laparoscopic hysterectomy have been described, depending on the extent of surgery carried out via the laparoscope. Total laparoscopic hysterectomy (TLH) is technically the most difficult laparoscopic technique and is performed entirely by the laparoscope without any vaginal component.
Currently, standard gynaecological practice dictates that, when feasible, vaginal hysterectomy (VH) is the surgical route of choice for benign hysterectomy. This is based on numerous studies, including a Cochrane review, which have shown VH to be associated with reduced infective morbidity and earlier return to normal activities compared with abdominal hysterectomy (AH). In cases in which VH is not technically possible, TLH appears to offer benefits as compared with AH. The recent Cochrane review on benign hysterectomy concluded that as a group, laparoscopic hysterectomies were slower and associated with more bleeding than VH. A subanalysis of TLH vs VH found no significant differences, although it included only 2 trials.
As experience with TLH increases, gynecologists have begun to debate the role of TLH in women otherwise suitable for VH. TLH facilitates better anatomical views, allows performance of concomitant surgery, and is suitable for larger uteri and those with little or no descent, which may prove difficult to remove vaginally. Several recent randomized trials comparing TLH with VH have been published, with conflicting conclusions. We present a metaanalysis of randomized controlled trials (RCTs) comparing VH and TLH for benign disease.
Materials and Methods
Search strategy
This metaanalysis was prepared in accordance with the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement. PubMed, Medline, SCOPUS, and Cochrane Databases were searched in June 2010 for combinations of the terms “laparoscopic,” “hysterectomy,” “vaginal,” “outcome,” “randomiz(s)ed,” and “benign.” The primary search was performed individually by 2 authors (R.G. and C.A.W.) for relevant trials published between January 1989 (the year that LH was first reported ) and June 2010, without language restrictions. Online trial registries ( controlled-trials.com and clinicaltrials.gov ) were also searched. Reference lists from articles retrieved in the primary searched were screened for additional citations.
Eligibility
Studies were eligible for inclusion if they met each of the following inclusion criteria: original prospective RCTs; hysterectomy performed for benign gynecological disease, and VH outcomes compared with TLH, as defined by the American Association of Gynecologic Laparoscopists. Studies were excluded from the analysis if any 1 of the inclusion criteria was not met.
Assessment of study quality
Eligible studies were assessed for methodological quality using the Jadad scoring system. This system assigns scores ranging from 0 to 5, with higher scores reflecting higher-quality evidence based on 3 parameters: randomization, blinding, and reporting of dropouts and withdrawals ( Table 1 ).
| Variable | Ribeiro et al 2003 | Morelli et al 2007 | Candiani et al 2009 | Drahonovsky et al 2010 | Ghezzi et al 2010 |
|---|---|---|---|---|---|
| Inclusion/exclusion criteria specified | Yes | Yes | Yes | Yes | Yes |
| Randomization process described | No | No | Yes | Yes | Yes |
| Blinding | No | No | No | No | No |
| Withdrawals explained | No | No | No | No | Yes |
| Power calculation | No | No | Yes | No | Yes |
| Primary outcome | N/A | Major complication | LOS | N/A | Pain |
| Baseline differences | Unclear | No | No | No | No |
| Jadad score | 1 | 1 | 2 | 2 | 3 |
Study outcomes
The primary outcome of the present analysis was incidence of perioperative complications (defined as intraoperative complications and postoperative complications presenting within 6 weeks of hysterectomy). Operative complications were classified using the validated Dindo 5 scale system, which ranks complications primarily on the therapy required to treat them ( Table 2 ). The rationale behind this system is to eliminate subjective interpretation of serious adverse advents and a tendency to downgrade serious complications. Secondary outcomes were operating time, blood loss, urinary tract injury, rate of conversion to laparotomy, postoperative pain, and length of postoperative stay.
| Dindo classification | Definition |
|---|---|
| Grade I | Any deviation from normal postoperative course, without the need for pharmacological intervention (except antipyretics, analgesics, antiemetics, electrolytes, diuretics) |
| Grade II | Requiring pharmacological treatment with drugs not included in grade I; also blood transfusions and TPN |
| Grade III | Requiring surgical, endoscopic, or radiological intervention (not under GA = IIIa; under GA = IIIb) |
| Grade IV | Life-threatening complication requiring intensive care |
| Grade V | Death |
Statistical analysis
Data were abstracted into an Excel spreadsheet for analysis. Pooled odds ratios (ORs) were calculated for categorical variables, and the weighted mean difference (WMD) was determined for continuous variables. These pooled outcome measures were calculated using random-effects models, as per DerSimonian and Laird. Where possible, data were analyzed on an intention-to-treat basis. Heterogeneity was assessed using Cochran’s Q test, a negative hypothesis test in which a P value less than 5% indicates the presence of significant statistical heterogeneity. Heterogeneity refers to the variation in outcomes between studies. Although some variation in individual treatment effects is expected between studies because of randomization, heterogeneity measures whether this variation is greater than would be expected by chance alone. Bias is a systematic error that leads to an incorrect estimate of effect. Bias was assessed visually by inspection of a funnel plot and statistically by use of the Horbold-Egger test. Bias is likely to cause asymmetry in such plots. The 5% level was considered significant. The statistical analysis was performed using Statsdirect 2.5.7 (Statsdirect Ltd, Altrincham, UK).
Results
The primary search retrieved 269 abstracts, which were screened for eligible studies. Thirty-one potentially eligible citations were identified and examined in detail ( Figure 1 ). Of these, 19 were not RCTs and were excluded. Of the remaining 12 studies, 7 studies described “laparoscopically assisted hysterectomy,” which did not meet a strict definition of TLH. The other 5 papers reported results from RCTs comparing VH with TLH and were included in the present analysis. Four studies were English-language articles and the remaining paper was reported in Italian, which we translated for the purposes of our analysis. Thus, we performed a metaanalysis of 5 RCTs comparing TLH (n = 332 women) and VH (n = 331 women; Table 3 ).
| Variable | Ribeiro et al 2003 | Morelli et al 2007 | Candiani et al 2009 | Drahonovsky et al 2010 | Ghezzi et al 2010 |
|---|---|---|---|---|---|
| Number | 40 | 400 | 60 | 81 | 82 |
| Inclusion criteria | Hysterectomy indicated | Benign or premalignant indication |
| Benign indication | Benign indication |
| Exclusion criteria |
|
|
|
|
|
| Antibiotic prophylaxis | Yes | N/A | Yes | Yes | Yes |
| Number of surgeons | 1 (experience?) | N/A | 4 (≥50 each) | 17 (experience?) |
|
| Estimation of EBL | Serial estimates of Hb/erythrocyte | N/A | Not stated | Intraoperative estimation and serial Hb | Not stated |
| Definition of postoperative fever | Not defined | ≥38°C | > 38°C | >38°C at longer than 24 h | Not defined |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
