Total versus subtotal hysterectomy for benign gynaecological conditions

Abstract

Background

Hysterectomy using an abdominal approach removes either the uterus alone (subtotal hysterectomy) or both the uterus and the cervix (total hysterectomy). The latter is more common but the outcomes have not been systematically compared.

Objectives

To compare short term and long term outcomes of subtotal hysterectomy (STH) with total hysterectomy (TH) for benign gynaecological conditions.

Search methods

We searched the Cochrane Menstrual Disorders and Subfertility Group Specialised Register of controlled trials (July 2011), CENTRAL (July 2011), MEDLINE (1966 to July 2011), EMBASE (1980 to July 2011), CINAHL (January 2005 to July 2011), Biological Abstracts (1980 to December 2005), the National Research Register and relevant citation lists.

Selection criteria

Only randomised controlled trials of women undergoing either total or subtotal hysterectomy for benign gynaecological conditions were included.

Data collection and analysis

Nine trials including 1553 participants were included. Independent selection of trials, assessment for risk of bias and data extraction were undertaken by two review authors and the results compared.

Main results

There was no evidence of a difference in the rates of multiple outcomes that assessed urinary, bowel or sexual function between TH and STH, either in the short term (up to two years post‐surgery) or long term (nine years post‐surgery). Length of operation (difference of 11 min) and amount of blood lost during surgery (difference of 57 ml) were significantly reduced during subtotal hysterectomy when compared with total hysterectomy. These differences are unlikely to constitute a clinical benefit and there was no evidence of a difference in the odds of blood transfusion. Post‐operative fever and urinary retention were less likely (fever: OR 0.48, 95% CI 0.3 to 0.8; retention: OR 0.23, 95% CI 0.1 to 0.8) and ongoing cyclical vaginal bleeding up to two years after surgery was more likely (OR 16.0, 95% CI 6.1 to 41.6) after STH compared with TH. There was no evidence of a difference in the rates of other complications, recovery from surgery, alleviation of pre‐surgery symptoms or readmission rates between the two types of hysterectomy carried out through the abdominal or laparoscopic route, although trials comparing the laparoscopic route were underpowered to detect some differences.

Authors' conclusions

This review has not confirmed the perception that subtotal hysterectomy offers improved outcomes for sexual, urinary or bowel function when compared with total abdominal hysterectomy. Women are more likely to experience ongoing cyclical bleeding up to a year after surgery with subtotal hysterectomy compared to total hysterectomy.

Plain language summary

Subtotal versus total hysterectomy

When hysterectomy is required for non‐cancerous conditions, either the uterus alone (subtotal hysterectomy) or the uterus and the cervix (total hysterectomy) are removed. It has been suggested that not removing the cervix (subtotal hysterectomy) would reduce the chances of sexual difficulties or problems with passing urine or stools. This review has found no evidence of a difference between these two different operations for these outcomes. Surgery is faster with subtotal hysterectomy and there is less blood loss during or just after surgery, although these benefits are not large. With subtotal hysterectomy, women are less likely to experience fever during or just after surgery but are more likely to have long term ongoing menstrual bleeding when compared with total hysterectomy.

Summary of findings

for the main comparison.

Subtotal hysterectomy compared with total hysterectomy for benign gynaecological conditions
Patient or population: Women requiring hysterectomy for benign conditions Settings: Hospital Intervention: Subtotal hysterectomy Comparison: Total hysterectomy
Outcomes	Illustrative comparative risks* (95% CI)		Relative effect (95% CI)	No of Participants (studies)	Quality of the evidence (GRADE)	Comments
	Assumed risk	Corresponding risk
	[control]	[experimental]
Prevalence of stress urinary incontinence within 2 years post surgery [follow‐up]	Low risk population		OR 1.45 (0.85 to 2.47)	955	⊕⊕⊕⊝ moderate	Stress incontinence was measured in most trials by questionnaire, but in Thakar it was more objectively determined by the use of twin‐channel subtracted cystometrography
	52 per 1000	73 per 1000
	Medium risk population
	[value] per 1000	[value] per 1000 ([value] to [value])
	High risk population
	[value] per 1000	[value] per 1000 ([value] to [value])
Prevalence of constipation within 2 years post surgery [follow‐up]	Low risk population		OR 0.80 (0.49 to 1.31)	555	⊕⊕⊝⊝ low	Evaluated by questionnaire. Inconsistency between trials.
	[value] per 1000	[value] per 1000 ([value] to [value])
	Medium risk population
	150 per 1000	126 per 1000 ([value] to [value])
	High risk population
	[value] per 1000	[value] per 1000 ([value] to [value])
Satisfaction with sex within 2 years post surgery [follow‐up]	Low risk population		OR 1.04 (0.68 to 1.59)	454	⊕⊕⊝⊝ low	Evaluated by questionnaire. Inconsistency between trials may have resulted from different questions on sexual satisfaction being asked.
	[value] per 1000	[value] per 1000 ([value] to [value])
	Medium risk population
	[value] per 1000	[value] per 1000 ([value] to [value])
	High risk population
	725 per 1000	736 per 1000 ([value] to [value])
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk Ratio; [other abbreviations, eg. OR, etc]
GRADE Working Group grades of evidence High quality: Further research is very unlikely to change our confidence in the estimate of effect. Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very low quality: We are very uncertain about the estimate.

Background

Description of the intervention

Anatomically, the uterus consists of two parts, the uterine body and the cervix. Hysterectomy is the surgical removal of the uterus. It is the most frequently performed major surgical intervention in gynaecology. A total hysterectomy involves the removal of both the uterine body and the cervix; a subtotal hysterectomy involves the removal of only the uterine body, leaving the cervix intact. Subtotal hysterectomy is also referred to as supracervical hysterectomy.

The first reported elective hysterectomy was performed through a vaginal approach by Conrad Langenbeck in 1813. The first elective abdominal hysterectomy, a subtotal operation, was performed by Charles Clay of Manchester in 1863 (Sutton 1997). Subtotal abdominal hysterectomy remained the operation of choice until 1929, when EH Richardson performed the first total abdominal hysterectomy. Subsequent concerns over the potential for the development of cancer in a conserved cervix, combined with improvements in operative and anaesthetic techniques, meant that total hysterectomy almost completely replaced subtotal hysterectomy.

One of the few indications for subtotal hysterectomy was recto‐vaginal endometriosis, which would have made removal of the cervix difficult or hazardous. However, consistent with developments in endometriosis surgery, it is now believed that retaining the cervix is likely to lead to residual disease and future symptoms. Therefore, subtotal hysterectomy should be seen as a relative contraindication for this type of endometriosis (Nezhat 1996). The other arguments presented in favour of subtotal hysterectomy include retaining the supporting structures of the uterus and vagina (cardinal and uterosacral ligaments) thereby in the long term reducing the risk of prolapse after a hysterectomy. In addition, by causing less damage to the nerves supplying the vagina, bladder and bowel, it is possible that subtotal hysterectomy might cause fewer urinary, bowel and sexual symptoms. A recent increase in the use of the laparoscopic approach to performing hysterectomies has led to an increase in numbers of subtotal procedures as well, as they would appear to be easier to perform than total hysterectomies. These proposed benefits of subtotal hysterectomy need to be reviewed and compared to outcomes with the standard procedure of total hysterectomy.

On the other hand, concern is often expressed regarding the risk of developing stump carcinoma of the cervix after performing subtotal hysterectomy. This has not been substantiated. The risk of cervical stump carcinoma in women with a previously normal Pap smear is no more than 0.3% (Storm 1992), approximately the same risk as for vaginal carcinoma after hysterectomy for a benign condition (Lyons 1993). However, caution should be taken so that women with subtotal hysterectomy fully understand the need for complying with the existing cervical screening program and they are not inappropriately excluded from screening.

How the intervention might work

Subtotal hysterectomy requires less dissection of surrounding tissue than total hysterectomy. Thus, there has been a suggestion it might be associated with:

• a reduced risk of damage to the bladder and ureter (Kilkku 1981; Parys 1990);

• a reduced risk of a post‐operative pelvic haematoma (collection of blood in the pelvis) (Nathorst‐Boos 1992);

• a reduced risk of prolapse after surgery;

• better sexual function (Helstrom 1994); and

• less damage to neuro‐anatomical structures compared to total hysterectomy (Thakar 2002), thereby preserving the nerve supply to vagina, bladder and bowel sphincters.

Why it is important to do this review

Comparative rates of subtotal and total hysterectomy vary in different parts of the world. There has been some evidence of a resurgence in the use of subtotal hysterectomy, particularly in Scandinavia. In Sweden, the ratio of subtotal to total hysterectomy is 0.56 (Culhed 1993) and in Denmark, the number of total abdominal hysterectomies decreased by 38% and the number of subtotal hysterectomies increased by 458% during the years 1988 to 1998, by which time 22% of all hysterectomies were subtotal (Gimbel 2001). There has been a smaller rise in the proportion of subtotal hysterectomies performed in the United States from 0.7% in 1990 to 1.1% in 1993 (Farquhar 2002) and subsequently to 1.6% in 1997 and 7.5% in 2004 (Merrill 2008). Changes were more pronounced in California where the rate was 6.9% in 1994 and rose to 20.8% in 2003 (Jacobson 2006). In contrast, the rate in the United Kingdom remains very low with a ratio of only 0.04 (Gimbel 2005), although a UK survey suggests that the ratio of subtotal to total hysterectomy will increase in the future (Esdaile 2006). Given the wide global variation in rates, there is uncertainty about the potential advantages and disadvantages of subtotal hysterectomy when compared to a total procedure and a review is required to clarify the uncertainty.

Objectives

To compare short term and long term outcomes of subtotal hysterectomy with total hysterectomy for benign gynaecological conditions.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCTs) where subtotal hysterectomy is compared with total hysterectomy, by any approach (laparoscopic, abdominal or vaginal).

Types of participants

Inclusion criteria

Women undergoing hysterectomy for benign gynaecological conditions. Subgroup analysis will be performed according to the indication for hysterectomy, if there are sufficient trials.

Exclusion criteria

Women with gynaecological cancer.

Types of interventions

Subtotal versus total hysterectomy.

Comparison of the surgical approaches to removal of the uterus, for example abdominal hysterectomy, vaginal hysterectomy and laparoscopic hysterectomy, are the focus of another Cochrane review and such comparative trials will not be included in this review. However, subgroup analysis will be performed in this review according to the approach used if there are sufficient trials.

Types of outcome measures

After the publication of the protocol, the review authors decided to list outcomes according to their status as primary or secondary outcomes. In the 2011 update of the review, the outcomes were re‐ordered and some new outcomes added. In the original review, outcomes were analysed at the time points six months, 12 months and 24 months, as described in the included studies. In the 'Implications for research' section of this review, the authors advised that longer term follow up was needed for full assessment of the comparative safety of subtotal versus total hysterectomy. In the 2011 update studies with longer term follow up (> five years post‐surgery) were also considered eligible for the review. In order to avoid analysing too many outcomes in this update of the review, we have simplified the multiple follow‐up times and defined 'short term' as outcomes occurring up to and including two years after surgery and 'long term' as outcomes occurring at longer follow‐up times.

For some of the short term measures, for example, blood loss, hospital stay and operating times, it is clinically relevant to assume that the outcome measures would be different when performed by the laparoscopic route as compared to the open approach. Also, compared to the original review, the more recent clinical trials are based on surgery by the laparoscopic route. Therefore, we have stratified the analysis under each outcome measure (both long term and short term) by the route of surgery, open or laparoscopic, where possible and wherever clinically relevant.

Studies were only included if they assessed one or more of the primary outcomes.

Primary outcomes

1. Urinary function

• stress incontinence

• urinary urgency

• voiding dysfunction (incomplete emptying)

2. Bowel function

• constipation

• incontinence (stool)

3. Sexual function

• Pain symptoms or dyspareunia

• Satisfaction, relationship and functioning combined

Secondary outcomes

1. Quality of life (measured by validated scales)

2. Operating time

3. Recovery from surgery

• length of hospital stay

• return to normal activities

4. Short term complications (pre‐discharge)

• surgical injury

• blood loss (amount in ml)

• requirement for blood transfusion

• pelvic haematoma

• vaginal bleeding

• urinary tract infection

• any other infection

• pyrexia

• urinary retention

• bowel obstruction

5. Intermediate term complications (post‐discharge, up to two years post‐surgery)

• ongoing cyclical bleeding

• persistent pain

• removal of cervical stump

• pelvic prolapse

• gynaecological cancer

6. Long term complications (> two years post‐surgery)

• fistula (abnormal opening)

• prolapse

• gynaecological cancer

7. Alleviation of pre‐surgery symptoms

8. Readmission to hospital (related to surgery)

Search methods for identification of studies

Electronic searches

The Trials Search Coordinator of the Cochrane Menstrual Disorders and Subfertility Group (MDSG) searched the following electronic databases for trials meeting the inclusion criteria:

• Cochrane Central Register of Controlled Trials (04/07/2011);

• MDSG specialised register (04/07/2011);

• MEDLINE (1966 to 04/07/2011);

• EMBASE (1980 to 04/07/2011);

• CINAHL (01/01/2005 to 04/07/2011);

• PsycINFO (01/01/2005 to 04/07/2011).

The search strategies for these searches are itemised in Appendix 1.

Searching other resources

1. The principal author searched the National Research Register (NRR), a register of ongoing and recently completed research projects funded by, or of interest to, the United Kingdom's National Health Service (NHS), as well as entries from the Medical Research Council's Clinical Trials Register, and details on reviews in progress collected by the NHS Centre for Reviews and Dissemination. The ClinicalTrials.gov register, a registry of federally and privately funded US clinical trials, was also searched.

2. The citation lists of relevant publications, review articles, abstracts of scientific meetings and included studies were also searched.

A search will be conducted for trials every two years and the review updated if new trials are found.

Data collection and analysis

Selection of studies

The independent selection of trials for inclusion in the review was performed by two review authors (VI and AL) in 2002 and for the update in 2011 (AL and AM) after employing the search strategy described previously. Differences of opinion were to be resolved by consensus after consultation with a third review author (NJ or RN) but this proved unnecessary.

Trials were excluded from the review if they did not meet the inclusion criteria and the references to these trials and the reasons for exclusion are listed in the table 'Characteristics of excluded studies'.

Data extraction and management

Data were extracted independently by two review authors, in 2002 (VI and AL) and for the 2011 update (AL and AM), using a form designed according to Cochrane guidelines. This information is presented in the table 'Characteristics of included studies' and provides a context for assessing the reliability of results. Additional information on trial methodology or actual original trial data were sought from the corresponding author of one trial which was initially published in a conference abstract but the information was subsequently published. The following information was extracted.

(1) Trial methods

1. Method of randomisation (either low risk, unclear risk or high risk)

2. Allocation concealment (either low risk, unclear risk or high risk)

3. Number of centres

4. Study design (parallel or crossover)

5. Blinding (of participants, investigators, assessors)

6. Number of participants randomised

7. Number of participants analysed

8. Methods used to describe missing data

9. Whether a power calculation was performed and adhered to

10. Whether 'intention‐to‐treat' analysis was performed by authors, possible from data but not performed by authors, not possible or uncertain

11. Source of funding stated or not

(2) Characteristics of the study participants

1. Inclusion criteria

2. Exclusion criteria

3. Age of participants

4. Source of participants

(3) Interventions

1. Approach to hysterectomy ‐ abdominal, vaginal, laparo‐vaginal, laparoscopic

2. Timing of follow‐up assessments after surgery

(4) Outcomes

1. Methods for measuring urinary, bowel and sexual function

2. Methods for measuring quality of life

Assessment of risk of bias in included studies

Risk of bias was assessed independently by two review authors (AL and AM) during the 2011 review, using the risk of bias tool developed by Julian Higgins (Higgins 2011). The following domains of the risk of bias tool were assessed:

• sequence generation (whether the allocation sequence was adequately generated, for example, random number table, computer random number generator, coin tossing, throwing dice);

• allocation concealment (whether the allocation was adequately concealed, for example, sequentially numbered containers of identical appearance, central allocation, sequentially numbered opaque sealed envelopes);

• blinding of participants, personnel and outcome assessors (whether knowledge of the allocated intervention was adequately prevented during the study, for example, by ensuring blinding of participants and key personnel or, where there is no blinding, knowledge of the intervention is not likely to influence the outcomes);

• incomplete outcome data (whether incomplete outcome data were adequately addressed, for example, missing data balanced in numbers across intervention groups, proportion of missing outcomes not sufficient to affect estimates, reasons for missing data unlikely to be related to the outcomes);

• selective outcome reporting (whether the reports of the study were free of suggestion of selective outcome reporting, for example, previous publication of a study protocol, other evidence that the study contains all of the prespecified outcomes);

• other sources of bias (whether the study was apparently free of other problems that could put it at a high risk of bias, for example, baseline imbalance, bias related to study design, early termination of the study).

Each domain was scored as either:

• low risk (criterion met);

• unclear risk (unclear whether criterion met);

• high risk (criterion not met).

The individual scores for each included study are found in the table 'Characteristics of included studies'. A summary is also included in Figure 1, and in graphic form in Figure 2.

1.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Measures of treatment effect

Dichotomous data were expressed as an odds ratio with 95% confidence interval. Where ordinal data were used to measure outcomes, for example, satisfaction rates, the categories were collapsed and the data dichotomised. The distributions of continuous data from the included studies were inspected for evidence of skew according to guidance from the Cochrane Handbook for Systematic Reviews of Interventions. If means and standard deviations (SDs) were available, or could be calculated, continuous data were expressed as the mean difference between groups with 95% confidence interval. Where there was strong evidence of skew in continuous data, results from the trial were not meta‐analysed but expressed in narrative format in the text of the review.

Unit of analysis issues

Due to the nature of the pre‐specified interventions and outcomes, it was not anticipated that there would be unit of analysis issues.

Dealing with missing data

Reasons for missing data in the included studies were documented and are included in the table 'Characteristics of included studies'. An assessment was made in the 'Risk of bias' table for each study whether the missing data in the trial were likely to affect the calculation of summary effect estimates.

Assessment of heterogeneity

The included studies were carefully inspected for evidence of clinical heterogeneity, in either the characteristics of the participants, the interventions, the outcomes or the trial duration. Where pooling the studies was appropriate, statistical heterogeneity between the results of different studies was examined by inspecting the scatter in the data points on the graphs and the overlap in their confidence intervals and, more formally, by checking the results of the Chi² test, using a P value of less than 0.10 as evidence of significant heterogeneity. The I² statistic was also checked to determine the percentage of total variation across studies that was due to heterogeneity rather than chance (Higgins 2003). These values can be categorised as follows: an I² of 25% represents mild heterogeneity, 50% represents moderate heterogeneity and 75% or more is evidence of extreme heterogeneity. In cases with extreme statistical heterogeneity which could not be explained by differences between studies, the estimates were not pooled in the meta‐analyses.

Assessment of reporting biases

It was planned to check for evidence of publication bias by assessment of the amount of asymmetry in a funnel plot. There were insufficient trials identified to undertake this assessment.

Data synthesis

The outcome data were pooled in a meta‐analysis where no significant clinical heterogeneity was apparent and there was no evidence of a major skew in the data.

Dichotomous data were combined for meta‐analysis with RevMan software using the Peto‐modified Mantel‐Haenszel method to estimate pooled odds ratios. For negative outcomes (for example, urinary incontinence) an increase in the odds of a particular outcome for the experimental group is displayed graphically in the meta‐analyses to the right of the centre‐line and a decrease in the odds of an outcome is displayed graphically to the left of the centre‐line. For positive outcomes (for example, satisfaction with treatment) an increase in odds is shown on the reverse axis. Graphs have been labelled for ease of interpretation.

Continuous data were combined for meta‐analysis with RevMan software using an inverse variance method to estimate the pooled mean difference (MD) with 95% confidence interval. Fixed‐effect models were used in the meta‐analysis.

Subgroup analysis and investigation of heterogeneity

Subgroup analysis was planned according to: 
 (a) indication for hysterectomy; 
 (b) time of follow up; 
 (c) hysterectomy approach (abdominal, laparoscopic or vaginal).

There were insufficient trials to undertake subgroup analysis according to indication for hysterectomy, but subgroups were used for separate analyses of those trials using abdominal procedures and those using laparoscopic procedures. Separate comparisons were made of outcomes assessed prior to two years post‐surgery and outcomes assessed after two years post‐surgery.

Statistical heterogeneity was assessed by the Chi² test (with P < 0.1 as evidence of significant heterogeneity) and the I² statistic. Where I² was found to be greater than 50%, sensitivity analysis was planned to compare results using a fixed‐effect model with the results using a random‐effects model.

Sensitivity analysis

It was planned to perform sensitivity analyses to examine the stability of the results in relation to: 
 (a) allocation concealment (adequate versus all trials); 
 (b) source of data (published only versus all trials); 
 (c) prior experience of the surgeon (experienced versus all trials).

There were insufficient trials to undertake these analyses.

Results

Description of studies

Results of the search

We identified 10 trials that were potentially relevant to the original publication of the review. Of these 10 studies, three were excluded, one because it was not randomised (Lyons 1993) and two because they did not measure any of the major outcomes in the review (Lalos 1986; Showstack 2004). Of the seven remaining studies, four were subsequent publications of a primary study (mostly assessing different outcomes from the primary publication). Thus, three RCTs met our inclusion criteria and were included in the original publication of the review.

For the 2011 update, an additional 15 trials were identified that were potentially relevant to the review. Of these 15 studies, two were excluded because they were meta‐analyses, one was excluded because it was a guideline, one was excluded because it was not randomised, and two were unobtainable and have been included in the awaiting classification section of the review. Of the remaining nine trials, one was a duplicate, one was a longer term follow up of a trial already included, one was a publication giving more details of a trial already included and six were new trials.

Included studies

For the 2011 update, nine trials met the inclusion criteria and were included in the review. The nine trials randomised a total of 1553 women, but not all participants were included in the analysis of every outcome.

Study design

All trials were randomised parallel group studies. Five trials were undertaken in a single centre, one trial involved 11 different centres in Denmark, one trial involved eight different centres in Sweden, one trial had four different centres in the US and one trial had two different centres in the UK. Two trials were undertaken in North America, two in Italy, two in Sweden, one in Denmark, one in Iran, and one in the UK. Power calculations for sample size were reported for seven of the included studies and were appropriate (although one study was concluded prematurely because of difficulties in recruitment); for the other two trials it was not clear whether power calculations were undertaken.

Six studies claimed that analysis of outcomes was by intention to treat. One study had true intention to treat for the primary outcomes but minimal loss to follow up for secondary outcomes (Learman 2003). Another study performed four analyses: 'regular' intention to treat (based on outcome data only for those participants whose results were known, that is, excluding exclusions from the analysis and those lost to follow‐up); 'best case scenario' intention to treat (the analysis considered all randomised participants and estimated dropouts as not having the primary undesirable outcome of interest); 'worst case scenario' intention to treat (the analysis considered all randomised participants and estimated dropouts as having the primary undesirable outcome of interest); and 'carry forward' intention to treat (analysis considered the last registered information on the outcome of interest among those dropping out as being the result at the end of the study period). Conclusions were based on the 'regular' intention to treat analysis, which excluded 13.2% of participants after randomisation (Gimbel 2003). Two studies (Thakar 2002; Morelli 2007) assessed peri‐operative outcomes, but not other outcomes, in full intention‐to‐treat (ITT) analyses. Dropouts and withdrawals, where reported, were similar between randomised groups but analysis of the primary outcomes was only undertaken where data were available. Two other studies claiming ITT analyses (Ellstrom 2010; Persson 2010) had exclusions from all analyses that were similar between randomised groups. Two studies did not have any ITT analyses and for one other study ITT analysis was not reported. There was no evidence that funding of the trials was from groups that could have benefited from the results of the studies; seven trials reported the source of their funding and two did not report how funding was provided. Follow up after surgery ranged from six to seven months (Flory 2006; Asnafi 2010), one year (Gimbel 2003; Gorlero 2008; Ellstrom 2010; Persson 2010), two years (Learman 2003; Morelli 2007) to seven to 11 years (mean nine years) (Thakar 2002).

Participants

Two of the studies specified that participants needed to be between 30 and 50 years of age with evidence that they were pre‐menopausal (follicle stimulating hormone (FSH) ≤ 30 mIU/mL) (Learman 2003; Morelli 2007), one study required participants to be less than 60 years (Thakar 2002), one study enrolled only pre‐menopausal women (Asnafi 2010; Ellstrom 2010), one study required participants to be between 18 and 55 years (Flory 2006), one study accepted participants aged less than 75 years (Gorlero 2008) and two studies did not mention age criteria (Gimbel 2003; Persson 2010). The mean age of the women in the trials varied from 42 to 49 years, although one study undertook longer follow up (an average of nine years after surgery). All women were eligible for hysterectomy for benign conditions, mostly fibroids or heavy menstrual bleeding. Women were excluded if they had known or suspected malignant conditions or other pathology. Two trials (Thakar 2002; Gorlero 2008) excluded women with known endometriosis.

Interventions

Six studies compared total abdominal hysterectomy with subtotal abdominal hysterectomy (Thakar 2002; Gimbel 2003; Learman 2003; Gorlero 2008; Asnafi 2010; Persson 2010); for two studies the procedures were performed laparoscopically (Flory 2006; Morelli 2007) and for one study the decision whether to use an abdominal, vaginal or laparoscopic approach was left to the surgeon (Ellstrom 2010). Two of the studies using the abdominal route specified that the total hysterectomy be done by the clamp‐cut‐ligate method (Kaser 1985) with polyglycolic sutures and antibiotic prophylaxis, and that the endocervical canal be electrocoagulated (surgical coagulation of tissue by an electrical heat process) after removing the uterus in subtotal hysterectomy. No other detailed instructions were provided and the remaining studies allowed surgeons to perform the operations using their customary techniques. Seven trials did not include any information on the experience or number of surgeons performing the procedures, one trial stated that only experienced surgeons were used (Thakar 2002) and one other trial stated that all the laparoscopic operations were performed by one experienced surgeon who was a consultant (Gorlero 2008).

Outcomes

In one trial, the primary outcomes were various measures of peri‐operative morbidity and sexual function at one and two years (Learman 2003), another trial assessed the effects of surgery on a wide range of urinary tract symptoms at one year follow up (Gimbel 2003), another assessed psychological wellbeing and sexual function together with clinical outcomes at one year (Persson 2010), and another evaluated measures of bladder, bowel and sexual function in detail both at one year and at an average of nine years follow up (Thakar 2002). For two other trials, the primary outcomes were satisfaction, sexual activity, body image and health status at one year follow up (Gorlero 2008; Ellstrom 2010), another assessed psychosocial functioning (defined as sexual, pain and psychological outcomes) at six months follow up (Flory 2006), another assessed sexual function as well as clinical outcomes at six months follow up (Asnafi 2010), and the remaining trial measured a wide range of outcomes, pelvic and urinary symptoms, surgical complications and clinical outcome, at two years follow up (Morelli 2007).

Risk of bias in included studies

Each included study was assessed for risk of bias (see 'Risk of bias' tables after each study in the table 'Characteristics of included studies').

Allocation

Most included studies allocated participants randomly into groups using computer generated numbers; in two studies the method of randomisation was not reported (Asnafi 2010;Ellstrom 2010). Four studies used block randomisation (Gimbel 2003; Learman 2003; Flory 2006; Persson 2010). Seven of the nine studies had adequate concealment of allocation (Thakar 2002; Gimbel 2003; Learman 2003; Flory 2006; Gorlero 2008; Ellstrom 2010; Persson 2010).

Blinding

One study (Thakar 2002) blinded participants and investigators for the first year of the study. Although self examination by participants could break the blinding, this was strongly discouraged and the investigators considered that the women were highly motivated and willing to participate in the interests of the study. Two other studies did not report whether blinding was undertaken, but it was considered unlikely (Asnafi 2010; Morelli 2007). Six studies reported that there was no blinding (Learman 2003; Flory 2006; Gimbel 2003; Gorlero 2008; Ellstrom 2010; Persson 2010).

Incomplete outcome data

Four studies adequately addressed their incomplete data by clearly specifying reasons for dropouts, which were balanced between groups and thus unlikely to affect estimates (Thakar 2002; Learman 2003; Morelli 2007; Persson 2010). For four other studies, there was either insufficient reporting of attrition and exclusions to permit judgments of whether incomplete data were adequately addressed or incomplete data were substantial (> 20%) (Gimbel 2003; Flory 2006; Gorlero 2008; Ellstrom 2010). In one study it appeared that there were no exclusions after randomisation (Asnafi 2010).

Selective reporting

No protocols were identified to check whether all specified outcomes were reported. However, all studies reported the results of the pre‐specified outcomes in the methods sections of their publications.

Other potential sources of bias

Seven of the included studies had no evidence of other potential sources of bias. One study (Flory 2006) reported a greater percentage of women with fibroids in the group that had subtotal hysterectomy compared to the group that had total hysterectomy. Another study (Asnafi 2010) analysed outcomes only in subgroups of women who were sexually active or who had previous dyspareunia.

Effects of interventions

See: Table 1

Where relevant, analyses of outcomes were subgrouped according to type of surgery, abdominal or laparoscopic. Separate comparisons were made of outcomes measured up to two years after surgery (with the later time interval used where outcomes were measured at multiple time intervals) and outcomes measured greater than two years after surgery (all measured at a mean of nine years after surgery).

Primary outcomes

Urinary function

There was no evidence of a significant difference in the prevalence of stress incontinence (within 2 years: OR 1.45, 95% CI 0.85 to 2.47; 4 studies; >2 years: OR 1.15, 95% CI 0.63 to 2.08; 1 study), incomplete emptying (within 2 years: OR 0.94, 95% CI 0.59 to 1.47, three studies; > 2 years: OR 0.69, 95% CI 0.37 to 1.29, 1 study) or urinary urgency (within 2 years: OR 1.05, 95% CI 0.47 to 2.37, 1 study; > 2 years: OR 1.26, 95% CI 0.68 to 2.32; 1 study) between randomised groups having either abdominal or laparoscopic surgery. There was also no evidence of significant differences between groups in any of the subgroup analyses according to type of surgery, abdominal or laparoscopic. There was moderate heterogeneity in the comparison of the abdominal subtotal with total hysterectomy with respect to incomplete emptying.

Bowel function

There was no evidence of a difference in the rates of constipation (within 2 years: OR 0.8, 95% CI 0.49 to 1.31, 2 studies; > 2 years: OR 1.52, 95% CI 0.67 to 3.45, 1 study) or incontinence of stools (> 2 years: OR 0.52, 95% CI 0.05 to 5.83, 1 study). Bowel function outcomes were not measured by the trials where laparoscopic surgery was undertaken. Substantial heterogeneity (I² = 73%) was found in the analysis of constipation rates within two years between the abdominal subtotal and total hysterectomy. When the data were carefully checked, the values for each outcome were dissimilar at baseline for groups in the Thakar trial. For neither outcome was there evidence of a significant difference between groups with or without the trial with imbalances at baseline.

Sexual function

Six trials measured multiple outcomes related to sexual function but the outcomes were measured in different ways, making it inappropriate to pool the results from some studies.

Some measure of sexual satisfaction was measured by five studies, using dichotomous or continuous data. There was no evidence of a difference in sexual satisfaction between randomised groups in meta‐analyses (dichotomous data: OR 1.04, 95% CI 0.68 to 1.59, 2 studies; continuous data SMD ‐0.15, 95% CI ‐0.43 to 0.13, 2 studies). One other trial that couldn't be pooled (Ellstrom 2010) also reported no evidence of significant differences. Substantial heterogeneity (I² = 76%) was found in the analysis of sexual satisfaction within two years (dichotomous data) between the abdominal subtotal and total hysterectomy. In these two pooled trials, satisfaction was assessed differently; one trial assessed whether women had a good sexual relationship with their partner and the other trial asked women whether they were satisfied with their sexual life, with or without a partner. Neither trial reported a significant difference between groups.

Dyspareunia (pain during intercourse) was measured by four trials. There was no evidence of a difference in dyspareunia (defined as either deep dyspareunia or dyspareunia not otherwise specified) between randomised groups (< 2 years: OR 0.87, 95% CI 0.46 to 1.67, 2 studies; > 2 years: OR 0.56, 95% CI 0.25 to 1.23, 1 study). Substantial heterogeneity (I² = 71%) was found in the analysis of dyspareunia within two years between the abdominal subtotal and total hysterectomy. The differences were likely to have arisen from different ways of measuring dyspareunia in the two trials. Two studies that couldn't be included in the meta‐analyses (Flory 2006; Asnafi 2010) also confirmed that there were no significant differences between groups. One of these studies used a laparoscopic approach to surgery.

Secondary outcomes

Quality of life

Quality of life was measured by five trials where women underwent abdominal surgery and one trial where women had laparoscopy. There was no evidence of a statistically significant difference in any of the quality of life scales measured within two years of surgery, although only a few studies contributed data to each outcome (General health (high better): MD 0.30, 95% CI ‐0.27 to 0.97, 3 studies; Physical domain (high better): MD ‐0.52, 95% CI ‐21.8 to 1.14, 3 studies; Mental domain (high better): MD ‐0.61, 95% CI ‐2.05 to 0.82, 4 studies; General health (low better): MD ‐1.0, 95% CI ‐4.92 to 2.92; 1 study; Anxiety (low better): MD 0.20, 95% CI ‐2.68 to 3.08, 1 study; Depression (low better): MD ‐0.27, 95% CI ‐1.55 to 1.00, 2 studies; Psychological domain (low better): MD ‐2.00, 95% CI ‐15.66 to 11.66, 1 study).

The data from the study that assessed quality of life outcomes at a mean of nine years after surgery were not suitable for pooling; this study also reported no significant differences between groups on any quality of life scales measured (Short Form‐36: Physical functioning, Mental Health and General Health and General Health Questionnaire: somatic and anxiety symptoms) (Thakar 2002).

In the majority of studies, quality of life improved from baseline (before surgery), regardless of the type of surgery.

Operation time and recovery

The length of time required for surgery was significantly shorter for subtotal abdominal hysterectomy when compared with total abdominal hysterectomy in four trials (MD ‐11.26 mins, 95% CI ‐15.1 to ‐7.5) (Thakar 2002; Learman 2003; Gorlero 2008; Persson 2010), but there was no evidence of a significant difference between the laparoscopic subtotal or total hysterectomy (MD ‐5.0, 95% CI ‐14.8 to 4.8, 1 study). There was no evidence of significant differences between types of hysterectomy for hospital stay or time to resume normal activities (hospital stay ‐ abdominal hysterectomy: MD ‐0.17, 95% CI ‐0.39 to 0.04, 5 studies; hospital stay ‐ laparoscopic hysterectomy: MD ‐0.2, 95% CI ‐0.6 to ‐0.2, 1 study; normal activities ‐ abdominal hysterectomy: MD ‐0.14, 95% CI ‐0.53 to 0.25).

Complications

Separate subgroups, according to whether hysterectomy was performed through the abdominal or laparoscopic route, were retained for the evaluation of transfusion requirements and blood loss since route of surgery is likely to cause variation. For all other complications of surgery, abdominal and laparoscopic trials were combined in the meta‐analyses.

Blood transfusion

There was no evidence of significant differences between randomised groups in the proportion of women who required blood transfusions with either laparoscopic or abdominal surgery.

Blood loss

Abdominal subtotal hysterectomy was associated with significantly less intraoperative blood loss than total hysterectomy (MD ‐56.6, 95% CI ‐99.6 to ‐13.7, three studies); one other trial where the data were not suitable for pooling also found a reduction in blood loss with subtotal hysterectomy (Gimbel 2003). There was no evidence of a significant difference in blood loss between a subtotal and total laparoscopic hysterectomy in one trial.

Other short term complications

Pyrexia and urinary retention were significantly reduced with the subtotal when compared to total hysterectomy (pyrexia: OR 0.48, 95% CI 0.31 to 0.75, 5 studies; urinary retention: OR 0.23, 95% CI 0.06 to 0.81, 5 studies). There was no evidence of significant differences in the rates of other short term complications such as surgical injury, pelvic haematoma, vaginal bleeding, wound infection, or bowel obstruction between groups having a subtotal or total hysterectomy.

Intermediate outcomes

Ongoing cyclical bleeding was significantly increased with the subtotal when compared to total hysterectomy (OR 16.0, 95% CI 6.1 to 41.6, 5 studies). There was no evidence of significant differences in the rates of other intermediate outcomes: persistent pain after discharge, removal of the cervical stump, pelvic prolapse or gynaecological cancer.

Long term outcomes:

At a mean of nine years after surgery, one trial found no significant difference in the rate of pelvic prolapse between randomised groups.

The included trials did not have long enough follow up to compare the odds of gynaecological cancer in the two groups.

Alleviation of symptoms

There was no evidence of significant differences in the alleviation of pre‐surgery symptoms such as back pain, pelvic pressure, menstrual abnormalities or pelvic pain according to whether a subtotal or total hysterectomy was performed. These outcomes were assessed by one to two trials.

Readmission rate

There was no evidence of a difference in the readmission rate with respect to abdominal surgery (four trials) or laparoscopic surgery (one trial) between groups randomised to subtotal or total hysterectomy.

Sensitivity analysis

There were too few trials in the analysis to conduct sensitivity analyses.

Discussion

Summary of main results

Primary outcomes

The rationale for undertaking this review was to examine the perception by women and some gynaecologists and health professionals that the retention of the cervix was necessary to maintain sexual pleasure and that total hysterectomy may lead to damage of the pelvic nerves or pelvic support structures that potentially could increase the risk of urinary incontinence, bowel dysfunction and reduce sexual pleasure (Thakar 2005). These were therefore considered the primary outcome measures.

This review has not demonstrated confirmatory evidence of subtotal hysterectomy causing less damage to neuroanatomical structures than total hysterectomy. The outcomes most indicative of such damage, including urinary, bowel and sexual dysfunction, have shown no consistent evidence of a benefit in women undergoing subtotal hysterectomy. Use of a laparoscopic or abdominal approach to hysterectomy did not alter the findings, although some of these analyses were underpowered.

Secondary outcomes

Quality of life measures after surgery did not appear to vary according to type of hysterectomy, whether an abdominal or laparoscopic approach was used, and up to nine years after surgery was completed. In all studies, quality of life improved significantly from baseline regardless of type of hysterectomy. This benefit of hysterectomy has been reported in another Cochrane systematic review evaluating the effect of hysterectomy for benign disease on women's well being (Lethaby 2009).

A significant benefit of subtotal, as compared to total, abdominal hysterectomy was reduced operating time and reduced blood loss, although no differences were reported in the requirement for blood transfusion. These benefits were not found for the laparoscopic approach possibly because only one trial contributed data. The average difference in operation time of 12 minutes was statistically different between groups but these differences are unlikely to be clinically significant benefits. There was no evidence of any difference in recovery from surgery, either hospital stay or return to normal activities after surgery.

Overall completeness and applicability of evidence

The findings in this review are based on only nine randomised trials including 1553 women. A wide range of outcomes relating to urinary, bowel and sexual function have been studied comprehensively in some of the trials included in this review. However, in order to provide a focus to the review and to improve readability, we restricted the potential major outcomes. Nevertheless, inclusion of studies comparing both open and laparoscopic routes, as well as assessing both short term and long term outcomes, have improved the scope for external validity and generalisation of study results.

Quality of the evidence

The studies included in this review are mostly small, and with some methodological flaws. Many of the studies were underpowered to find differences, which is reflected in the wide confidence intervals in the findings. In particular, lack of blinding means we cannot exclude the likelihood that some outcomes may have been influenced by knowledge of the treatment. However, it is almost impossible to guarantee blinding in the comparison of a subtotal with total hysterectomy, because of the need for women with an intact cervix to continue screening for cervical cancer.

Potential biases in the review process

Many of the included studies assessed a multiplicity of outcomes, often correlated, relating to urinary, bowel and sexual function, and measured at a number of different time points without adjustment or correction. The risks of multiplicity of data and clinical heterogeneity among the studies in terms of the types of symptoms, time of reporting of symptoms, scales used to analyse the symptoms and route of surgery have been addressed by stratified analysis and avoiding inclusion of too many outcome measures. In general, a reporting bias could occur when combining outcome measures reported at various time points in the individual studies under a somewhat broad and arbitrary categorisation of less than two years and greater than two years. This is especially relevant for some of the time‐related short term outcomes such as urinary symptoms, which potentially could have shown improvement or deterioration over time when measured at intervals like six months, 12 months or 24 months. While combining them all under the less than two years category, we have chosen the time point with the greatest interval since surgery to represent effects persisting after recovery from surgery; and a separate long term analysis more than two years after surgery for some outcomes where longer term information was provided. As mentioned above, this allowed us to reduce the bias due to multiplicity of variables. It will always be recommended to refer to the individual studies for detailed descriptions.

Agreements and disagreements with other studies or reviews

Four retrospective observational studies (Kilkku 1981; Kilkku 1985; Roovers 2001; Neumann 2004) and three systematic reviews (Brown 2000; Gimbel 2007; Robert 2008) have also assessed the effects of type of hysterectomy on various measures of urinary function after surgery. Results were inconsistent in the observational studies and two of the three systematic reviews confirmed most of the results of the RCTs in this review. There was no evidence of a difference in stress incontinence according to type of hysterectomy. Although two of the three systematic reviews in broad terms confirmed the results of this review, their authors reached slightly different conclusions. The Brown systematic review did suggest that, for women aged more than 60 years, urinary incontinence after hysterectomy is about 60% higher than for women in the same age group who had not undergone hysterectomy. This has not been confirmed by other studies, in particular the RCTs in this review which measured incontinence at baseline and post‐surgery and found improved urinary function. It is possible that the conditions that may lead to hysterectomy adversely affect lower urinary tract function and surgery provides a benefit. The outcomes measured here were measures of women's perception of urinary symptoms rather than urodynamic investigation since the association between clinical symptoms and urodynamic findings is poor (Abrams 1983). The Robert systematic review suggested that although there was no evidence of statistical differences, their meta‐analysis showed a non‐significant trend toward increased risk of stress incontinence (RR 1.3, 95% CI 0.94, 1.78) and incomplete emptying (RR 0.9, 95% CI 0.59 to 1.38) in women who underwent a subtotal compared with total hysterectomy. They suggested that the included trials may have been underpowered to detect effects and that longer follow up was needed to allow symptoms to emerge. However, this present review has included additional data with much longer follow up (average of nine years), published since the Robert meta analysis, which indicates similar rates of stress incontinence and incomplete emptying regardless of whether subtotal or total hysterectomy was performed. The systematic review by Gimbel concluded that overall incontinence was less likely for women undergoing total hysterectomy. However, the Gimbel review included a trial which was excluded by this review and groups were not comparable at baseline; in this trial, 64% of those undergoing subtotal hysterectomy had stress incontinence prior to surgery compared to only 18% of women in the total hysterectomy group. Moreover, Gimbel assessed prevalence of total incontinence by pooling different types of incontinence, stress, urge and mixed. This approach is not appropriate as stress and urge incontinence are considered to develop from different causes.

Few studies have assessed bowel function according to type of hysterectomy surgery. One retrospective study not included in the review has reported an increased prevalence of disturbed bowel function within one month of hysterectomy that waned over time but no differences according to type of surgery (van Dam 1997). In a prospective multicentre study (Roovers 2007), defecation complaints such as constipation and incomplete evacuation were more prevalent in women undergoing subtotal hysterectomy when compared to women undergoing total hysterectomy.

Sexual response after hysterectomy has been extensively studied in a number of observational studies but results were inconsistent and some of the studies had methodological flaws (Kilkku 1983; Kilkku 1983a; Saini 2002; Roovers 2003; Roussis 2004; Lonnee‐Hoffmann 2006). The poorer quality retrospective studies (Kilkku 1983; Kilkku 1983a; Saini 2002) reported that women undergoing subtotal hysterectomy reported better sexual function and satisfaction than those undergoing total hysterectomy, but two more recent better quality prospective studies and one retrospective study (Roovers 2003; Roussis 2004; Lonnee‐Hoffmann 2006) reported no differences between groups. These latter three studies also reported that perceived sexual function appeared to improve after hysterectomy regardless of technique.

Authors' conclusions

Implications for practice.

There appears to be a limited resurgence in rates of subtotal hysterectomy in the Western world. This review, however, has not confirmed the perception that subtotal hysterectomy offers improved outcomes for urinary, sexual and bowel function when compared with total hysterectomy. Although surgery is significantly faster and blood loss reduced, these may not translate to clinical benefits. Post‐operative febrile morbidity is reduced with subtotal hysterectomy but ongoing cyclical vaginal bleeding is likely to be increased up to a year after surgery. A consensus opinion published by the American College of Obstetricians and Gynecologists concluded that subtotal hysterectomy should not be recommended by the surgeon as superior to total hysterectomy when indicated for benign disease (ACOG 2007 (reaffirmed 2010)). Women requiring hysterectomy need to be given information based on the evidence presented in this review so that they can make well informed choices, and this is not often routine. An American survey reported that fewer than 20% of gynaecologists offered women a choice between subtotal and total hysterectomy (Zekam 2003). Women can be informed about the route of hysterectomy by reference to another Cochrane systematic review (Nieboer 2006).

One of the rationales for total, as opposed to subtotal, hysterectomy is the potential risk of cervical cancer when the cervix is left in place, although this review has not been able to assess this risk. The incidence of cervical cancer in women who have had subtotal hysterectomy is rare; a study of 1104 women having this surgery in Denmark between 1978 and 1988 found an incidence of 0.3% (Storm 1992). The same study, however, reported a 3.3 to 5 fold increased rate of cervical cancer if subtotal hysterectomy was carried out in women aged 50 years or older. Another study of cervical cancer screening in the Midwestern United States showed no differences in screening rates between women who did not have hysterectomy (Eaker 1998). This potential risk is not such an issue for women in countries that have routine cervical screening programs. However, it may be prudent to advise against subtotal hysterectomy in women with a history of high grade cervical lesions, a fear of developing cervical cancer, or cervical cancer screening that is not up to date or unlikely to occur regularly in the future. Even in countries where routine cervical screening exists, it is important that women are adequately counselled prior to and after subtotal hysterectomy.

Implications for research.

Although this review has not confirmed the presumed superiority of subtotal hysterectomy for preserving urinary, sexual and bowel function, the conclusions are based on only nine RCTs, only one of which was blinded; with approximately 1500 women in total. There are difficulties in adequately measuring these complex outcomes and more research would be welcome to confirm the provisional conclusions of this review. Larger, double blinded randomised controlled trials with adequate assessment tools are needed because many of the important outcomes are subjective. 
 One of the expected disadvantages of total hysterectomy, an increase in post‐operative vaginal vault prolapse, has not been confirmed in this review and it is possible that the trials were underpowered to adequately assess this outcome. Prolapse may appear years after hysterectomy and more studies with long term follow up are needed to assess whether cervical preservation results in better support of the vaginal vault.

Four of the six studies in this review compared subtotal abdominal with total abdominal hysterectomy. The short term advantages of the laparoscopic approach have been well documented and it has been argued that these benefits may be even more apparent with subtotal laparoscopic hysterectomy. Thus, the comparative benefits of laparoscopic total and subtotal hysterectomy need to be tested in more blinded well designed RCTs.

What's new

Date	Event	Description
6 December 2011	New search has been performed	Six more RCTs identified and added in the 2011 update. Primary outcome list has been restructured to include bowel and sexual activities. This is to reflect the fact that more recent studies report additional outcomes. The number of primary outcomes has been reduced and comparisons structured to assess outcomes within 2 years post‐surgery and greater than 2 years post‐surgery, rather than at multiple time points, to simplify the comparisons.
6 December 2011	New citation required but conclusions have not changed	The addition of 6 studies has not led to a change in our conclusions.

History

Protocol first published: Issue 4, 2004
 Review first published: Issue 2, 2006

Date	Event	Description
6 November 2008	Amended	Converted to new review format.
3 February 2006	New citation required and conclusions have changed	Substantive amendment

Appendices

Appendix 1. Search strategies

CENTRAL l

1 exp Hysterectomy/ (1315) 
 2 hysterectom$.tw. (2040) 
 3 or/1‐2 (2256) 
 4 cerv$.tw. (6692) 
 5 uter$.tw. (3845) 
 6 total$.tw. (81361) 
 7 sub$total$.tw. (346) 
 8 (cerv$ adj5 conserv$).tw. (37) 
 9 supracerv$.tw. (19) 
 10 or/4‐9 (89754) 
 11 3 and 10 (1059) 
 12 limit 11 to yr="2008 ‐Current" (160)

MDSG

Keywords CONTAINS "Hysterectomy" or Title CONTAINS "Hysterectomy" AND Keywords CONTAINS "Hysterectomy, subtotal" or "subtotal" or "total abdominal hysterectomy"  or "total addominal hysterectomy"  or "total hysterectomy"  or "total laparoscopic hysterectomy"  or "supravaginal hysterectomy" or "supracervical hysterectomy" or "Hysterectomy, Vaginal" or "hysterectomy techniques" or Title CONTAINS "Hysterectomy, subtotal" or "subtotal" or "total abdominal hysterectomy"  or "total addominal hysterectomy"  or "total hysterectomy"  or "total laparoscopic hysterectomy"  or "supravaginal hysterectomy" or "supracervical hysterectomy" or "Hysterectomy, Vaginal" or "hysterectomy techniques"

MEDLINE

1 exp Hysterectomy/ (21954) 
 2 hysterectom$.tw. (22723) 
 3 or/1‐2 (32374) 
 4 cerv$.tw. (167335) 
 5 uter$.tw. (132705) 
 6 total$.tw. (1111625) 
 7 sub$total$.tw. (14062) 
 8 (cerv$ adj5 conserv$).tw. (524) 
 9 supracerv$.tw. (304) 
 10 or/4‐9 (1373177) 
 11 3 and 10 (17190) 
 12 randomized controlled trial.pt. (310410) 
 13 controlled clinical trial.pt. (82747) 
 14 randomized.ab. (225974) 
 15 placebo.tw. (133438) 
 16 clinical trials as topic.sh. (155008) 
 17 randomly.ab. (166255) 
 18 trial.ti. (96498) 
 19 (crossover or cross‐over or cross over).tw. (51126) 
 20 or/12‐19 (758822) 
 21 exp animals/ not humans.sh. (3606366) 
 22 20 not 21 (700755) 
 23 11 and 22 (1285) 
 24 (200810$ or 200811$ or 200812$).ed. (206356) 
 25 (2009$ or 2010$ or 2011$).ed. (2319706) 
 26 24 or 25 (2526062) 
 27 23 and 26 (208)

EMBASE

1 exp Hysterectomy/ (35185) 
 2 hysterectom$.tw. (26174) 
 3 or/1‐2 (41466) 
 4 total$.tw. (1257287) 
 5 complete.tw. (466910) 
 6 4 or 5 (1664702) 
 7 3 and 6 (8546) 
 8 sub$total.tw. (14362) 
 9 supra$cervi$.tw. (391) 
 10 partial$.tw. (455064) 
 11 8 or 9 or 10 (468092) 
 12 7 and 11 (926) 
 13 Clinical Trial/ (810309) 
 14 Randomized Controlled Trial/ (281916) 
 15 exp randomization/ (53159) 
 16 Single Blind Procedure/ (13675) 
 17 Double Blind Procedure/ (99014) 
 18 Crossover Procedure/ (29973) 
 19 Placebo/ (180293) 
 20 Randomi?ed controlled trial$.tw. (61040) 
 21 Rct.tw. (7053) 
 22 random allocation.tw. (1025) 
 23 randomly allocated.tw. (15048) 
 24 allocated randomly.tw. (1674) 
 25 (allocated adj2 random).tw. (682) 
 26 Single blind$.tw. (10777) 
 27 Double blind$.tw. (115701) 
 28 ((treble or triple) adj blind$).tw. (237) 
 29 placebo$.tw. (155788) 
 30 prospective study/ (164952) 
 31 or/13‐30 (1118924) 
 32 case study/ (12385) 
 33 case report.tw. (201855) 
 34 abstract report/ or letter/ (782469) 
 35 or/32‐34 (992855) 
 36 31 not 35 (1086028) 
 37 12 and 36 (146) 
 38 (2010$ or 2011$).em. (1754807) 
 39 37 and 38 (22)

CINAHL

1 exp Hysterectomy/ (1556) 
 2 hysterectom$.tw. (1391) 
 3 or/1‐2 (1989) 
 4 cerv$.tw. (9275) 
 5 uter$.tw. (3266) 
 6 total$.tw. (57757) 
 7 sub$total$.tw. (192) 
 8 (cerv$ adj5 conserv$).tw. (35) 
 9 supracerv$.tw. (26) 
 10 or/4‐9 (69151) 
 11 3 and 10 (653) 
 12 limit 11 to yr="2005 ‐ 2008" (349) 
 13 exp clinical trials/ (66624) 
 14 Clinical trial.pt. (35279) 
 15 (clinic$ adj trial$1).tw. (15159) 
 16 ((singl$ or doubl$ or trebl$ or tripl$) adj (blind$3 or mask$3)).tw. (8936) 
 17 Randomi?ed control$ trial$.tw. (12888) 
 18 Random assignment/ (19554) 
 19 Random$ allocat$.tw. (1364) 
 20 Placebo$.tw. (12335) 
 21 Placebos/ (4737) 
 22 Quantitative studies/ (4303) 
 23 Allocat$ random$.tw. (78) 
 24 or/13‐23 (91550) 
 25 24 and 12 (53) 
 26 from 25 keep 1‐53 (53)

PsycINFO

1 exp Hysterectomy/ (347) 
 2 hysterectom$.tw. (594) 
 3 1 or 2 (614) 
 4 total$.tw. (121233) 
 5 complete.tw. (46339) 
 6 4 or 5 (164343) 
 7 3 and 6 (74) 
 8 sub$total.tw. (156) 
 9 supra$cervi$.tw. (2) 
 10 partial$.tw. (48278) 
 11 8 or 9 or 10 (48425) 
 12 7 and 11 (10)

Data and analyses

Comparison 1. Subtotal hysterectomy versus total hysterectomy.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Prevalence of stress urinary incontinence within 2 years post surgery	5	955	Odds Ratio (M‐H, Fixed, 95% CI)	1.45 [0.85, 2.47]
1.1 Abdominal surgery	4	826	Odds Ratio (M‐H, Fixed, 95% CI)	1.55 [0.86, 2.78]
1.2 Laparoscopic surgery	1	129	Odds Ratio (M‐H, Fixed, 95% CI)	1.05 [0.29, 3.82]
2 Prevalence of stress urinary incontinence >2 years post surgery	1	178	Odds Ratio (M‐H, Fixed, 95% CI)	1.15 [0.63, 2.08]
2.1 Abdominal surgery	1	178	Odds Ratio (M‐H, Fixed, 95% CI)	1.15 [0.63, 2.08]
2.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
3 Prevalence of incomplete bladder emptying within 2 years post surgery	4	768	Odds Ratio (M‐H, Fixed, 95% CI)	0.94 [0.59, 1.47]
3.1 Abdominal surgery	3	639	Odds Ratio (M‐H, Fixed, 95% CI)	0.89 [0.55, 1.45]
3.2 Laparoscopic surgery	1	129	Odds Ratio (M‐H, Fixed, 95% CI)	1.28 [0.37, 4.44]
4 Prevalence of incomplete bladder emptying >2 years post surgery	1	173	Odds Ratio (M‐H, Fixed, 95% CI)	0.69 [0.37, 1.29]
4.1 Abdominal surgery	1	173	Odds Ratio (M‐H, Fixed, 95% CI)	0.69 [0.37, 1.29]
4.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
5 Prevalence of urinary urgency within 2 years post surgery	2	254	Odds Ratio (M‐H, Fixed, 95% CI)	1.05 [0.47, 2.37]
5.1 Abdominal surgery	1	125	Odds Ratio (M‐H, Fixed, 95% CI)	0.86 [0.25, 2.99]
5.2 Laparoscopic surgery	1	129	Odds Ratio (M‐H, Fixed, 95% CI)	1.23 [0.42, 3.61]
6 Prevalence of urinary urgency >2 years post surgery	1	174	Odds Ratio (M‐H, Fixed, 95% CI)	1.26 [0.68, 2.32]
6.1 Abdominal surgery	1	174	Odds Ratio (M‐H, Fixed, 95% CI)	1.26 [0.68, 2.32]
6.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
7 Prevalence of constipation within 2 years post surgery	2	555	Odds Ratio (M‐H, Fixed, 95% CI)	0.80 [0.49, 1.31]
7.1 Abdominal surgery	2	555	Odds Ratio (M‐H, Fixed, 95% CI)	0.80 [0.49, 1.31]
7.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
8 Prevalence of constipation >2 years post surgery	1	165	Odds Ratio (M‐H, Fixed, 95% CI)	1.52 [0.67, 3.45]
8.1 Abdominal surgery	1	165	Odds Ratio (M‐H, Fixed, 95% CI)	1.52 [0.67, 3.45]
8.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
9 Prevalence of incontinence of stools within 2 years post surgery	1	166	Odds Ratio (M‐H, Fixed, 95% CI)	0.52 [0.05, 5.83]
9.1 Abdominal surgery	1	166	Odds Ratio (M‐H, Fixed, 95% CI)	0.52 [0.05, 5.83]
9.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
10 Prevalence of incontinence of stools >2 years post surgery	1	166	Odds Ratio (M‐H, Fixed, 95% CI)	0.52 [0.05, 5.83]
10.1 Abdominal surgery	1	166	Odds Ratio (M‐H, Fixed, 95% CI)	0.52 [0.05, 5.83]
10.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
11 Satisfaction with sex within 2 years post surgery	2	454	Odds Ratio (M‐H, Fixed, 95% CI)	1.04 [0.68, 1.59]
11.1 Abdominal surgery	2	454	Odds Ratio (M‐H, Fixed, 95% CI)	1.04 [0.68, 1.59]
11.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
12 Satisfaction with sex >2 years post surgery	1	131	Odds Ratio (M‐H, Fixed, 95% CI)	0.95 [0.13, 6.98]
12.1 Abdominal hysterectomy	1	131	Odds Ratio (M‐H, Fixed, 95% CI)	0.95 [0.13, 6.98]
12.2 Laparoscopic hysterectomy	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
13 Satisfaction with sex (cont) within 2 years post surgery	2	192	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.15 [‐0.43, 0.13]
13.1 Abdominal surgery	1	129	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.04 [‐0.39, 0.30]
13.2 Laparoscopic surgery	1	63	Std. Mean Difference (IV, Fixed, 95% CI)	‐0.37 [‐0.87, 0.13]
14 Satisfaction with sex (cont) >2 years post surgery	0	0	Std. Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
14.1 Abdominal surgery	0	0	Std. Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
14.2 Laparoscopic surgery	0	0	Std. Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
15 Prevalence of pain during sex within 2 years post surgery	2	452	Odds Ratio (M‐H, Fixed, 95% CI)	0.87 [0.46, 1.67]
15.1 Abdominal surgery	2	452	Odds Ratio (M‐H, Fixed, 95% CI)	0.87 [0.46, 1.67]
15.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
16 Prevalence of pain during sex >2 years post surgery	1	133	Odds Ratio (M‐H, Fixed, 95% CI)	0.56 [0.25, 1.23]
16.1 Abdominal surgery	1	133	Odds Ratio (M‐H, Fixed, 95% CI)	0.56 [0.25, 1.23]
16.2 Laparoscopic surgery	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
17 Quality of life within 2 years post abdominal surgery (high better)	5		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
17.1 General (abdominal)	3	478	Mean Difference (IV, Fixed, 95% CI)	0.35 [‐0.27, 0.97]
17.2 Physical domain (abdominal)	3	652	Mean Difference (IV, Fixed, 95% CI)	‐0.52 [‐2.18, 1.14]
17.3 Mental domain (abdominal)	4	831	Mean Difference (IV, Fixed, 95% CI)	‐0.61 [‐2.05, 0.82]
18 Quality of life within 2 years post abdominal surgery (low better)	2		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
18.1 General (abdominal)	1	179	Mean Difference (IV, Fixed, 95% CI)	‐1.0 [‐4.92, 2.92]
18.2 Anxiety (abdominal)	1	179	Mean Difference (IV, Fixed, 95% CI)	0.20 [‐2.68, 3.08]
18.3 Depression (abdominal and laparoscopic)	2	242	Mean Difference (IV, Fixed, 95% CI)	‐0.27 [‐1.55, 1.00]
18.4 Psychological domain (laparoscopic)	1	63	Mean Difference (IV, Fixed, 95% CI)	‐2.0 [‐15.66, 11.66]
19 Operating time (mins)	5		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
19.1 Abdominal surgery	4	694	Mean Difference (IV, Fixed, 95% CI)	‐11.26 [‐15.07, ‐7.45]
19.2 Laparoscopic surgery	1	141	Mean Difference (IV, Fixed, 95% CI)	‐5.0 [‐14.80, 4.80]
20 Length of hospital stay (days)	6		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
20.1 Abdominal surgery	5	844	Mean Difference (IV, Fixed, 95% CI)	‐0.17 [‐0.39, 0.04]
20.2 Laparoscopic surgery	1	141	Mean Difference (IV, Fixed, 95% CI)	‐0.20 [‐0.62, 0.22]
21 Return to normal activities (weeks)	2		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
21.1 Abdominal surgery	2	310	Mean Difference (IV, Fixed, 95% CI)	‐0.14 [‐0.53, 0.25]
21.2 Laparoscopic surgery	0	0	Mean Difference (IV, Fixed, 95% CI)	0.0 [0.0, 0.0]
22 Requirement for blood transfusion	5		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
22.1 Abdominal surgery	4	694	Odds Ratio (M‐H, Fixed, 95% CI)	1.24 [0.61, 2.54]
22.2 Laparoscopic surgery	1	141	Odds Ratio (M‐H, Fixed, 95% CI)	1.25 [0.32, 4.86]
23 Blood loss during surgery (mls)	4		Mean Difference (IV, Fixed, 95% CI)	Subtotals only
23.1 Abdominal surgery	3	589	Mean Difference (IV, Fixed, 95% CI)	‐56.63 [‐99.58, ‐13.68]
23.2 Laparoscopic surgery	1	141	Mean Difference (IV, Fixed, 95% CI)	‐36.0 [‐145.35, 73.35]
24 Short term complications (predischarge)	6		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
24.1 Surgical injury	3	549	Odds Ratio (M‐H, Fixed, 95% CI)	0.28 [0.06, 1.36]
24.2 Pelvic haematoma/abscess	3	660	Odds Ratio (M‐H, Fixed, 95% CI)	0.41 [0.13, 1.32]
24.3 Vaginal bleeding	3	660	Odds Ratio (M‐H, Fixed, 95% CI)	0.74 [0.29, 1.91]
24.4 Wound infection	3	733	Odds Ratio (M‐H, Fixed, 95% CI)	0.86 [0.38, 1.95]
24.5 Pyrexia (fever)	5	933	Odds Ratio (M‐H, Fixed, 95% CI)	0.48 [0.31, 0.75]
24.6 Urinary retention	5	933	Odds Ratio (M‐H, Fixed, 95% CI)	0.23 [0.06, 0.81]
24.7 Bowel obstruction/ileus	4	731	Odds Ratio (M‐H, Fixed, 95% CI)	0.59 [0.24, 1.46]
25 Intermediate term complications (after discharge and within 2 years post surgery)	5		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
25.1 Ongoing cyclical bleeding	5	964	Odds Ratio (M‐H, Fixed, 95% CI)	15.97 [6.14, 41.56]
25.2 Persistent pain	5	963	Odds Ratio (M‐H, Fixed, 95% CI)	0.90 [0.58, 1.38]
25.3 Removal of cervical stump	2	457	Odds Ratio (M‐H, Fixed, 95% CI)	5.14 [0.60, 44.35]
25.4 Pelvic prolapse	4	839	Odds Ratio (M‐H, Fixed, 95% CI)	2.74 [0.80, 9.34]
25.5 Gynaecological cancer	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
26 Long term complications (>2 years post surgery)	1		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
26.1 Fistula	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
26.2 Pelvic prolapse	1	127	Odds Ratio (M‐H, Fixed, 95% CI)	0.84 [0.41, 1.70]
26.3 Gynaecological cancer	0	0	Odds Ratio (M‐H, Fixed, 95% CI)	0.0 [0.0, 0.0]
27 Alleviation of pre‐surgery symptoms	2		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
27.1 Back pain	2	266	Odds Ratio (M‐H, Fixed, 95% CI)	1.33 [0.78, 2.27]
27.2 Pelvic pressure	2	266	Odds Ratio (M‐H, Fixed, 95% CI)	0.69 [0.28, 1.68]
27.3 Menstrual abnormalities	1	141	Odds Ratio (M‐H, Fixed, 95% CI)	3.00 [0.12, 74.90]
27.4 Pelvic pain	1	141	Odds Ratio (M‐H, Fixed, 95% CI)	0.86 [0.31, 2.38]
28 Readmission rate (related to surgery)	5		Odds Ratio (M‐H, Fixed, 95% CI)	Subtotals only
28.1 Abdominal surgery	4	869	Odds Ratio (M‐H, Fixed, 95% CI)	1.10 [0.63, 1.91]
28.2 Laparoscopic surgery	1	141	Odds Ratio (M‐H, Fixed, 95% CI)	1.96 [0.68, 5.62]

1.1. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 1 Prevalence of stress urinary incontinence within 2 years post surgery.

1.2. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 2 Prevalence of stress urinary incontinence >2 years post surgery.

1.3. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 3 Prevalence of incomplete bladder emptying within 2 years post surgery.

1.4. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 4 Prevalence of incomplete bladder emptying >2 years post surgery.

1.5. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 5 Prevalence of urinary urgency within 2 years post surgery.

1.6. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 6 Prevalence of urinary urgency >2 years post surgery.

1.7. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 7 Prevalence of constipation within 2 years post surgery.

1.8. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 8 Prevalence of constipation >2 years post surgery.

1.9. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 9 Prevalence of incontinence of stools within 2 years post surgery.

1.10. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 10 Prevalence of incontinence of stools >2 years post surgery.

1.11. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 11 Satisfaction with sex within 2 years post surgery.

1.12. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 12 Satisfaction with sex >2 years post surgery.

1.13. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 13 Satisfaction with sex (cont) within 2 years post surgery.

1.15. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 15 Prevalence of pain during sex within 2 years post surgery.

1.16. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 16 Prevalence of pain during sex >2 years post surgery.

1.17. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 17 Quality of life within 2 years post abdominal surgery (high better).

1.18. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 18 Quality of life within 2 years post abdominal surgery (low better).

1.19. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 19 Operating time (mins).

1.20. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 20 Length of hospital stay (days).

1.21. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 21 Return to normal activities (weeks).

1.22. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 22 Requirement for blood transfusion.

1.23. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 23 Blood loss during surgery (mls).

1.24. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 24 Short term complications (predischarge).

1.25. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 25 Intermediate term complications (after discharge and within 2 years post surgery).

1.26. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 26 Long term complications (>2 years post surgery).

1.27. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 27 Alleviation of pre‐surgery symptoms.

1.28. Analysis.

Comparison 1 Subtotal hysterectomy versus total hysterectomy, Outcome 28 Readmission rate (related to surgery).

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Asnafi 2010.

Methods	Randomisation method: Not reported No. of centres: 1 Design: Parallel group Blinding: Not reported but unlikely No. randomised: 150 No. analysed: 150 Power calculation: Yes (150 overall to detect a 20% difference between the groups with 80% power, alpha level of 0.95 and confidence level of 95%) Intention‐to‐treat analysis: Yes, except for sexual functioning/dyspareunia ‐ analyses performed only in women who were sexually active or complained of dyspareunia Source of funding: Babol Medical University, Iran
Participants	Inclusion: Women >35 years; premenopausal; offered abdominal hysterectomy for symptomatic uterine fibroids with confirmation of the lesion or abnormal uterine bleeding without any response to hormone therapy of at least 3 months trial. Exclusion: Age >50 years at screening; positive pregnancy test; genital tract carcinoma; body weight >100kg; diabetes mellitus; candidates for vaginal hysterectomy determined by a gynecologist; unlikely to remain geographically accessible for follow up. Age: 43 and 46 years (mean in each treatment group) Source: From Department of Gynecology in a teaching hospital associated with Babol Medical University in Iran
Interventions	(1) subtotal abdominal hysterectomy (2) total abdominal hysterectomy Follow up: 6 months after surgery
Outcomes	Fever; anaemia; duration of hospitalisation, changes in sexual function
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Stated as "randomly assigned"
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding (performance bias and detection bias) All outcomes	High risk	Lack of blinding could have affected outcomes such as sexual functioning and pain
Incomplete outcome data (attrition bias) All outcomes	Low risk	No reported dropouts
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Unclear risk	Sexual functioning assessed only in subgroups of women ‐ unclear if these subgroups groups were comparable at baseline. Short follow up for assessment of sexual functioning

Ellstrom 2010.

Methods	Randomisation method: Method not described, other than allocation from sealed opaque envelopes No. of centres: 1 Design: Parallel group Blinding: No No. randomised: 132 No. analysed: 104 Exclusions from analysis: Subtotal: declined surgery or operated elsewhere (n=2); salphingoophorectomy (n=2); lost to follow up (n=10) Total: declined surgery or operated elsewhere (n=5); malignancy diagnosed perioperatively (n=1); lost to follow up (n=10) Protocol violations: Subtotal: Change of method due to surgical complications (n=2) Total: n=0 Power calculation: Yes (50‐70 patients per treatment arm required, no other details reported) Intention to treat analysis: Stated as yes, but not true ITT analysis as lost to follow up not included Source of funding: Swedish Medical Research Council (B95‐17X‐11237‐01A) and the Goteborg Medical Society Fund
Participants	Inclusion: Pre‐menopausal patients scheduled for hysterectomy for benign disorders Exclusion: Previous cervical dysplasia; planned oophorectomy; previous symptomatic prolapse Age: 45 years (mean) Source: Patients requiring hysterectomy for benign disorders at the Department of Obstetrics and Gynaecology, Sahlgrenska University Hospital, Gothenburg, Sweden
Interventions	(1) subtotal hysterectomy (2) total hysterectomy For both treatment groups, abdominal hysterectomy was recommended when the diameter of the uterus was >11cm, otherwise vaginal or laparoscopic surgery was planned but the final decision was made by the surgeon. Follow up: 12 months after surgery
Outcomes	Changes in sexual health (measured by the McCoy Female Sexuality Questionnaire) and changes in psychological wellbeing (measured by the Psychological General Well‐being index)
Notes	Lack of power
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	"Randomised in a ratio of 1:1" but method not described
Allocation concealment (selection bias)	Low risk	Sealed opaque envelopes; performed by a study nurse
Blinding (performance bias and detection bias) All outcomes	High risk	Blinding of participants was originally planned but proved impossible. Knowledge of treatment could have affected patients' perceptions of sexual function and health
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	>20% attrition in each group
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Unclear risk	No adjustments made for multiple outcomes

Flory 2006.

Methods	Randomisation method: Computer generated block randomisation No. of centres: 1 Design: Parallel group Blinding: No No. randomised: 80 No. analysed: 63 Dropout at the end of follow up: 9/40 in subtotal group (2 after randomisation; 2 moved/wrong phone; 3 not interested; 4 other reasons); 8/40 in total group (1 after randomisation; 3 moved/wrong phone; 2 not interested; 3 other reasons) Power calculation: Yes (32 per treatment arm for moderate effect size (difference of 0.5 SD) gave 80% power, with alpha=0.05) Intention‐to‐treat analysis: No Source of funding: Canadian Foundation for Womens Health Institute of Health Research
Participants	Inclusion: 18‐55 years old; pre‐menopausal; fluent in French language Exclusion: Prior oophorectomy; prior uterine prolapse; prior chemotherapy; prior neoplasia in the uterus/cervix Age: 44 years (mean) Source: From surgeons/gynaecologists and local media announcement, study undertaken at Department of Obstetrics and Gynecology, University of Montreal
Interventions	(1) subtotal laparoscopic hysterectomy (2) total laparoscopic assisted vaginal hysterectomy Follow up: 6 ‐ 7 months after surgery
Outcomes	Sexual drive; sexual arousal; orgasm; sexual behaviour; overall sexual functioning; pain (Likert scale and MPQ); depression and other psychological symptoms (BDI and BSI); body image (SSS and BES); psychosocial functioning
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated block randomisation
Allocation concealment (selection bias)	Low risk	Treatment assignment concealed in consecutively numbered sealed envelopes, opened by surgeons at the time of surgery
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Not reported
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	>20% dropout but balanced between groups
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Unclear risk	Groups not balanced at baseline (87% of women in subtotal group and 66% of women in total group had fibroids)

Gimbel 2003.

Methods	Randomisation method: Restricted, computer generated block randomisation No of centres: 11 Design: parallel group Blinding: No No. randomised: 319 No. analysed: 277 Dropout at end of follow up: 15% in subtotal group; 11% in total group Power calculation: yes Intention to treat analysis: Authors claimed both 'regular' ITT and per protocol analysis but 13% of randomised participants excluded from analysis Source of funding: Numerous trial groups/organisations and hospitals
Participants	Inclusion: Women who are scheduled for hysterectomy for benign disease Exclusion: Laparoscopic/vaginal hysterectomy; dysplasia (cervical); uterine prolapse; malignant disease; diabetes; participation in other research projects; unable to read/write Danish; former urological operation; cervix problems; psychological problems; poor mental function; neurological disease; chronic alcoholism. Age: 47 years (mean) Source: Departments of Obstetrics and Gynaecology in Denmark
Interventions	(1) subtotal hysterectomy (2) total abdominal hysterectomy Follow up 1 year
Outcomes	Primary: Perceived urinary incontinence Secondary: Quality of life (SF36); constipation; prolapse; satisfaction with sexual life; pelvic pain; vaginal bleeding; postoperative complications; dyspareunia
Notes	A later publication compared the effects of interventions on sexual function (Zobbe 2003) and another later publication (Gimbel 2005) compared the effects of the interventions on a more detailed specification or urinary symptoms (stress, urge and mixed incontinence and incomplete bladder emptying)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated randomisation procedure
Allocation concealment (selection bias)	Low risk	Central allocation
Blinding (performance bias and detection bias) All outcomes	High risk	Authors acknowledged the "lack of blinding"
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	>10% lost to follow up ‐ no reasons given
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Multiple sources of funding including Organon but unlikely bias because of independent data monitoring. No baseline imbalance between groups.

Gorlero 2008.

Methods	Randomisation method: computer generated numbers Number of centres: 1 Design: parallel group No. randomised: 117 No. analysed: 105 Dropout at end of follow up: 12/117 (10.3%) ‐ reasons not given Power calculation: no Intention to treat analysis: no Source of funding: not stated
Participants	Inclusion: Women requiring an abdominal hysterectomy for a benign indication Exclusion: 2nd or 3rd degree uterine prolapse; age >75 years; malignancy; BMI>29; previous pelvic surgery; endometriosis or history of chronic pelvic pain; abnormal cervical smears; psychiatric disorders Age: subtotal hyst (mean 46 years); total hyst (mean 49 years) Source: Department of San Martino Hospital and University of Genoa in Genoa, Italy (Jan 2003 to December 2005)
Interventions	(1) subtotal hysterectomy (2) total hysterectomy Follow up: 1 year
Outcomes	Primary: Womens' satisfaction (evaluated by answers to a questionnaire on sexual activity, body image and health status) Secondary: Occurrence of surgical complications; postoperative recovery
Notes	Study measures 'satisfaction' by women's responses to questions on sexual activity, body image and quality of life
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated numbers
Allocation concealment (selection bias)	Low risk	Sealed opaque envelopes opened immediately before surgical incision
Blinding (performance bias and detection bias) All outcomes	High risk	Stated as not blinded
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	No reasons given for incomplete data and no information on distribution between groups
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Groups comparable at baseline and no other potential bias identified

Learman 2003.

Methods	Randomisation method: Computer generated random numbers sequence in blocks with sealed numbered opaque envelopes Number of centres: 4 Design: parallel group No randomised: 135 No analysed: 135 Drop out at end of follow up: 10% for subtotal hyst and 4% for total hyst Power calculation: yes Intention to treat analysis: Yes (for some outcomes) Source of funding: AHRQ Stratified by clinical centre
Participants	Inclusion: Pre‐menopausal women with symptomatic fibroids who have decided to undergo abdominal hysterectomy OR pre‐menopausal women who have abnormal bleeding and a minimum 3 month trial of hormonal management who want hysterectomy; if >/= 45 yrs, FSH </= 30 mIU/mL and negative biopsy within 6 months for hyperplasia/cancer Exclusion: Age >50 years; positive pregnancy test; desire for future childbearing; genital tract cancer (known or suspected); cervical dysplasia or carcinoma in situ; complex or atypical endometrial hyperplasia; candidate for vaginal hysterectomy; not geographically accessible for 4 yrs. Age: 41.8 (mean) Source: University gynaecological clinics affiliated with 4 universities in USA
Interventions	(1) subtotal hysterectomy (2) total abdominal hysterectomy Follow up: 2 yrs
Outcomes	Primary: Surgical complications and clinical outcomes: reduction in symptoms; hospital readmissions; rate of complications; degree of symptom improvement; activity limitation Secondary: Sexual function and health related quality of life
Notes	A later publication compared the effects of the interventions on sexual function and quality of life
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated random number sequence, stratified by centre, in blocks
Allocation concealment (selection bias)	Low risk	Sealed numbered opaque envelopes
Blinding (performance bias and detection bias) All outcomes	High risk	Unblinded
Incomplete outcome data (attrition bias) All outcomes	Low risk	Loss to follow up but reasons clearly specified
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes reported
Other bias	Low risk	No baseline imbalance, funding by AHRQ

Morelli 2007.

Methods	Randomisation method: Computer generated No. of centres: 1 Design: Parallel group No. randomised: 141 No. analysed: 129 (primary outcome at 24 months); 141 for surgical outcomes Dropout at end of follow up: Subtotal group: 8/71 (11.3%) ‐ 1 death, 7 lost to follow up. Total group: 4/70 (5.7%) ‐ 1 death, 3 lost to follow up Power calculation: Not reported Intention to treat analysis: No for primary outcomes but surgical outcomes were ITT. Source of funding: Not reported
Participants	Inclusion: Age >30 years; pre‐menopausal; abnormal uterine bleeding with previous hormonal treatment for at least 3 months and diagnosis confirmed by echo or hysteroscopy OR symptomatic uterine leiomyomas (bleeding, compression etc) with diagnosis confirmed by echo or hysteroscopy OR patients >45 years with FSH ≤30 mIU/ml and negative endometrial biopsy for hyperplasia or carcinoma. Exclusion: Pregnancy; age >50 years; planned pregnancy; diagnosed or suspected genital cancer; dysplasia; endometrial hyperplasia; candidate for vaginal hysterectomy Age: Mean 42 years Source: Not reported ‐ all patients identified through a vaginal screening program in Catanzaro, Italy
Interventions	(1) subtotal laparoscopic hysterectomy (2) total laparoscopic hysterectomy (both using standard surgery procedures) Follow up: 24 months after surgery
Outcomes	Surgical outcomes: operation time, blood loss, other operative complications; readmission to hospital during follow up; irregular bleeding; pelvic pain; pelvic compression; lumbar pain; urinary urgency; sensation of incomplete emptying of bladder; stress incontinence
Notes	Publication translated from Italian into English by Lorenzo Moja of the Italian Cochrane Centre
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated
Allocation concealment (selection bias)	Unclear risk	Not reported
Blinding (performance bias and detection bias) All outcomes	Unclear risk	Not reported but unlikely
Incomplete outcome data (attrition bias) All outcomes	Low risk	Reasons clearly specified for dropouts before the conclusion of the trial at 24 months
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Groups balanced at baseline. No other possible bias identified.

Persson 2010.

Methods	Randomisation method: Random numbers table with block randomisation according to centre No. of centres: 8 Design: Parallel group No. randomised: 200 No. analysed: 178 Dropout at end of follow up: Subtotal group: 5/104 withdrew consent prior to surgery, 4/104 withdrew consent during study period and 1/104 missing diary. Total group: 3/96 withdrew consent prior to surgery, 2/96 intraoperative finding of cancer, 1/96 converted to subtotal hysterectomy, 1/96 protocol violation, 5/96 withdrew consent during study period, 1/96 missing diary Power calculation: Yes: difference in PGWB score of 8 points Intention to treat analysis: Stated as intention to treat but 10% of subtotal and 13% of total group not included in the analyses Source of funding: Medical Research Council of south‐east Sweden and County Council of Ostergotland and Linkoping University
Participants	Inclusion: Planned hysterectomy for benign gynaecological condition, proficiency in Swedish, preservation of at least one ovary Exclusion: Malignancy in genital organs, previous or present cervical dysplasia, rapidly growing fibroids where malignancy could not be ruled out, preoperative treatment with GnRH analogues, post‐menopausal women without hormone replacement therapy, severe psychiatric disorders Age: Mean 46 years Source: Patients identified from seven hospitals and one private gynaecological clinic in Sweden ‐ admitted for hysterectomy because of benign gynaecological conditions
Interventions	(1) subtotal abdominal hysterectomy (2) total abdominal hysterectomy (both techniques according to surgeon discretion) Follow up: 12 months after surgery
Outcomes	Primary: General psychological wellbeing Secondary: Post‐operative complications (including stress incontinence), surgical and clinical outcomes during surgery
Notes	Two publications
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Random numbers table with block randomisation according to centre
Allocation concealment (selection bias)	Low risk	"Opaque envelopes numbered sequentially in accordance with random table, opened consecutively"
Blinding (performance bias and detection bias) All outcomes	High risk	"Women informed about their assignment prior to surgery"
Incomplete outcome data (attrition bias) All outcomes	Low risk	"Missing outcome data balanced in numbers across intervention groups with similar reasons for missing data across groups"
Selective reporting (reporting bias)	Low risk	All prespecified outcomes reported
Other bias	Low risk	Comparison groups balanced at baseline. No pharmaceutical funding.

Thakar 2002.

Methods	Randomisation method: Computer generated numbers and sealed opaque envelopes opened after surgical incision made. No of centres: 2 Design: parallel group Blinding: double (participants and investigator for 1 year of trial) No randomised: 279 No analysed: 279 (only for peri‐operative outcomes) Dropout at end of follow up: 8% in subtotal group; 14% in total group Power calculation for sample size: yes Intention to treat analysis: yes for some outcomes, but some data not available for analysis of primary outcomes Source of funding: Responsive Funding Program, Research and Development; NHS Executive; London.
Participants	Inclusion: Women offered abdominal hysterectomy for a benign indication Exclusion: >60 years; suspected carcinoma; body weight >100 kg; previous pelvic surgery; known endometriosis; abnormal cervical smears; symptomatic uterine prolapse; symptomatic urinary incontinence Age: 43‐44 (mean) Source: 2 London hospitals in the UK (Jan 1996 to Apr 2000)
Interventions	(1) subtotal hysterectomy (2) total abdominal hysterectomy Follow up: 1 yr
Outcomes	Primary: Bowel, bladder and sexual function Secondary: Postoperative complications; intra‐operative outcomes and complications; readmission rate; changes in psychological outcomes and health status/quality of life
Notes	A later publication (Thakar 2004) compared the effects of the interventions on health status/quality of life and psychological outcomes and another later publication (Thakar 2005) compared the effects of the interventions on longer follow up (7 to 11 years after surgery).
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer generated numbers
Allocation concealment (selection bias)	Low risk	Sealed opaque envelopes only opened after surgical incision
Blinding (performance bias and detection bias) All outcomes	Low risk	Participants and investigators blinded for 1 year
Incomplete outcome data (attrition bias) All outcomes	Low risk	Clear explanations given for missing data but analysis at 9 years undertaken on 65% of original study group
Selective reporting (reporting bias)	Low risk	All pre‐specified outcomes reported
Other bias	Low risk	No baseline imbalance, funding by research program

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Einarsson 2010	Not RCT ‐ does not mention randomisation to groups
Gimbel 2007	Meta‐analysis, not RCT
Kives 2010	Guideline on subtotal hysterectomy, not RCT
Lalos 1986	Did not measure one or more of the primary outcomes for the review
Lyons 1993	Not randomised
Robert 2008	Meta‐analysis, not RCT
Showstack 2004	Resource use for total and supracervical hysterectomy was compared. These outcomes are not relevant to the review

Characteristics of studies awaiting assessment [ordered by study ID]

Asgari 2009.

Methods	RCT
Participants	Patients who were candidates for hysterectomy with benign disease with no contraindications for laparoscopic surgery; recruited from Arash Hospital from March 2007 to April 2009. N=45; 20 for subtotal and 25 for total hysterectomy
Interventions	(1) subtotal laparoscopic hysterectomy; (2) total laparoscopic hysterectomy
Outcomes	Duration of surgery, blood transfusion, length of hospital stay, post‐operative pain, time to return to normal activities, sexual function, dyspareunia, cyclic bleeding, cervical prolapse, intra and post‐operative complications
Notes	Awaiting translation

Ghanbari 2007.

Methods	Single blinded RCT
Participants	N=50; 25 randomised to subtotal abdominal hysterectomy and 25 randomised to total abdominal hysterectomy
Interventions	(1) subtotal abdominal hysterectomy; (2) total abdominal hysterectomy
Outcomes	Duration of surgery, volume of bleeding, duration of hospital stay, operative complications, dyspareunia, sexual satisfaction, ongoing bleeding
Notes	Awaiting translation

Differences between protocol and review

For the 2011 update, as a result of peer review, the list of outcomes was simplified and analysed according to only two periods of follow up: within two years and greater than two years.

Contributions of authors

Valeria Ivanova devised the idea for the review and wrote the protocol after discussion with Professor Cindy Farquhar. She selected trials for inclusion and extracted data.

Anne Lethaby commented on the protocol and finalised the protocol after peer review. She performed searches for trials, selected trials for inclusion, extracted and entered data, assessed all studies for risk of bias and wrote the remaining sections of the review. She also led the 2011 update of the review.

Asima Mukhopadhyay selected trials for inclusion, extracted data and assessed studies for risk of bias and commented on the text of the review for the 2011 update. She also provided clinical input and wrote the discussion.

Raj Naik provided clinical input during the selection of primary and secondary outcomes for the 2011 update and also commented on the text of the review.

Neil Johnson commented on the text of the original publication of the review.

Sources of support

Internal sources

• Department of Obstetrics and Gynaecology, University of Auckland, New Zealand.

External sources

• No sources of support supplied

Declarations of interest

None known

New search for studies and content updated (no change to conclusions)