Searches were conducted in CINAHL, Cochrane, EMBASE, MEDLINE OVID, PEDro, and Pubmed databases, without date or language restrictions. The search was conducted between February and March 2022, and updated to include trials until October 2023. Search terms included words related to hysterectomy and to pelvic floor muscle training (see Fig. 1 for the full search strategy for each database).

Fig. 1.

Search strategy for the review.

(0.5MB).

Titles and abstracts were screened to identify relevant trials by two independent reviewers (JMVF/TVC), using Mendeley - Reference Management Software (first search) and Rayyan website (updated search).21 Full copies of peer-reviewed relevant trials were retrieved, and their reference lists were screened to identify potentially additional relevant trials. The method section of the retrieved trials was extracted (TVC) and reviewed by two independent reviewers (JMVF/NFFO) using predetermined inclusion criteria: Design = randomized trials; Participants = women >18 years, after hysterectomy; Intervention = pelvic floor muscle training; Comparison = no intervention/placebo; Outcomes = urinary symptoms, vaginal prolapse, sexual function, pelvic floor muscle strength, or quality of life. Reviewers were blinded to authors, journals, and results. Disagreement or ambiguities were resolved by consensus after discussion with a third reviewer (LRN).

Methodological quality: The quality of included trials was assessed by extracting PEDro scores from Physiotherapy Evidence Database (www.pedro.org.au), which is an 11-item scale designed for rating the methodological quality of randomized trials.

Participants: Trials were included when participants were adults - female sex (from now on referred to as women), according to the Sex and Gender Equity in Research (SAGER) guidelines.22 In addition, to be included, at least 50% of the participants should have undertaken hysterectomy. Characteristics of participants were recorded to assess similarity of trials.

Intervention: Trials were included if the experimental intervention was pelvic floor muscle training, defined as exercises to improve pelvic floor muscle strength, endurance, power, or relaxation.13 The training had to consist of rapid or sustained voluntary contractions, prescribed by a physical therapist or other healthcare professional.23 The experimental intervention could be either home- or center-based, individualized or in groups, and participants could receive in-person or remote supervision.

Comparison: The control intervention could be no intervention/placebo. Characteristics of control were recorded to assess similarity of trials.

Measures: Five outcomes were of interest: urinary symptoms, vaginal prolapse, sexual function, pelvic floor muscle strength, and quality of life. Urinary symptoms may include urinary incontinence, bladder storage or urination symptoms, measured by direct tests (e.g., miccional diary) or questionnaires (e.g., International Consultation on Incontinence Questionnaire).24 Vaginal prolapse had to be measured by direct tests (e.g. Pelvic Organ Prolapse Quantification) or by questionnaires of symptoms (e.g., Pelvic Organ Prolapse/ Urinary Incontinence Sexual Questionnaire).24 Sexual function, which may include disorders of desire, arousal, pain/discomfort, and orgasm, had to be measured by questionnaires (e.g., Female Sexual Function Index, which ranges from 0 to 36 points, and values <21 are indicative of sexual dysfunction).25,26 Pelvic floor muscle strength had to reflect the ability to generate maximum voluntary contraction, measured for example by palpation tests or vaginal manometry.23,27 Quality of life had to be measured by generic (e.g., World Health Organization BREF) or condition-specific questionnaires related to pelvic floor dysfunction (e.g., King's Health Questionnaire).28,29 When multiple condition-specific questionnaires were reported, the questionnaire that evaluated two or more symptoms (e.g., urinary and prolapse symptoms) was used.

Information about the methods and results were extracted by two reviewers (LRN and NFFO).

The post-intervention or change scores were used to obtain the pooled estimate of the effect of the intervention, using a random effects model. A visual inspection of the distribution of effect sizes in the forest plots was performed and the I2 value was calculated to indicate the proportion of variance that was due to heterogeneity.30 Values of I2 greater than 50% are indicative of important heterogeneity.31 The analysis were performed using Review Manager Version 5.3 (Copenhagen, Denmark). The pooled data for each outcome were reported as the mean difference (MD) between groups for continuous data or risk difference (RD) for dichotomous data, and their 95% confidence intervals (95% CI). Where data of trials could not be included in a pooled analysis, the between-group result was reported.

The GRADE system was used to summarize the quality of evidence for each outcome, which ranges from high to very low quality.32 We rated evidence from the high-quality level and downgraded it one point if one of the following prespecified criteria was present: low methodological quality (defined as >50% of trials with PEDro score < 6); inconsistency of estimates after pooling (I2 > 50%), or when assessment was not possible (no pooling); indirectness of participants (when any trial had <80% of included women with hysterectomy or the results from those who undertook hysterectomy were not reported separately); and imprecision (pooling <300 participants for each outcome).33 Two reviewers (NFFO and LRN) assessed the quality of the evidence using the GRADE system, with potential disagreements resolved by consensus.

The electronic search strategy identified 1347 records. After removing duplicates and screening titles, abstracts, and reference lists, 20 potential trials were listed for full reading. Twelve trials34-45 failed to meet the inclusion criteria (Appendix 1) and the full-text of one trial was not obtained,46 leaving 8 trials to be included.47-55 Two trials47,53 had the same sample and different outcome measures, and two trials50,51 had the same sample and different timing for measures; these trials were, therefore, reported as single trials. Therefore, 6 comparisons from 8 trials were included in the review (Fig. 2). The corresponding authors were contacted by e-mail for additional information, but none replied.

Fig. 2.

PRISMA Flow diagram of included and excluded trials through the review. n = number, RCT = randomized controlled trial.

aTrials may have been excluded for failing to meet one or more than one inclusion criteria.

(0.61MB).

The 6 trials involved 776 participants and investigated the immediate effects of pelvic floor muscle training on urinary symptoms (n = 4),47,48,50,52 vaginal prolapse (n = 3),47,48,50 sexual function (n = 4),49,50,53,54 pelvic floor muscle strength (n = 5),47,48,50,52,53 and quality of life (n = 6).48-50,52-54 Two trials47,51 examined follow-up effects (Table 1).

Methodological quality: The mean PEDro score of the trials was 5.5 (range 5–6) (Table 2). All trials had randomly allocated participants. Most trials reported between-group differences (83%), point estimate and variability data (83%), had similar groups at baseline (67%), <15% dropouts (67%), and reported concealed allocation (50%) and whether an intention-to-treat analysis was undertaken (50%). On the other hand, most trials did not have blinded assessors (67%). No trials blinded participants or therapists, which is difficult or impossible due to the nature of the interventions.

Participants: The mean age of the participants ranged from 46 to 58 years old, and the mean body mass index ranged from 26 to 33 kg/m2. Half of the trials (50%)47,50,53 reported the participants’ level of pelvic floor muscle weakness, and one trial50 included participants with moderate to severe weakness (i.e., Modified Oxford Scale = 0–2). All trials included participants in the postoperative phase after hysterectomy, and three trials48,50,52 reported the percentage of women who had hysterectomy (65−90%).

Intervention: In all trials, the experimental intervention was pelvic floor muscle training in combination with educational exercises or emotional-release management by yoga exercise. Pelvic floor muscle training was delivered either in rehabilitation center/hospital47,48,50,52-54 and/or at home.47,48,49,52 There was some clinical heterogeneity among trials. Only two trials52,54 reported treatment session duration, which varied between 15 and 75 min. All trials reported session frequency: five trials47,48,50,51,53,54 had supervised sessions once per week, and two trials49,52 had seven unsupervised sessions per week. All trials reported program duration, which varied between 4 and 24 wk.

In most trials, the pelvic floor muscle training consisted of 3–6 series and 8–20 repetitions of sustained contractions, supervised by physical therapists.47-50,52,53 Training progression was accomplished by increasing training intensity and/or session duration, or by modifying training positioning. All trials delivered no intervention to the control groups. Three trials48,49,53 provided education to both experimental and control groups.

Outcome measures: Four trials47,48,50,52 examined urinary symptoms: three47,48,52 measured the number of women with urinary symptoms using questionnaires (i.e., Urogenital Distress Inventory and Patient Global Impression of Improvement), and one50 measured number of voids/day using an urinary diary. Three trials47,48,50 examined vaginal prolapse: one50 measured prolapse stage using the Pelvic Organ Prolapse Quantification, and two47,48 measured number of women with prolapse. Four trials49,50,53,54 examined sexual function using questionnaires: two49,50 used the Female Sexual Function Index, one53 used the sexual function domain of the Pelvic Floor Questionnaire, and one trial54 used the Pelvic Organ Prolapse Incontinence Sexual Questionnaire/PISQ-12. Five trials47,48,50,52,53 examined pelvic floor muscle strength: two48,53 used manometry (cmH2O), and three47,50,52 used validated graded scale (i.e., Modified Oxford Scale and Brinks Scale). All trials47-50,52,53 examined quality of life using a condition-specific or a generic questionnaire.

Urinary symptoms: The effects of pelvic floor muscle training on urinary symptoms were examined by pooling post-intervention data from three trials involving 429 participants,47,48,52 which reported number of women with urinary symptoms. The mean risk difference was −0.1, which suggests 10% reduction in the probability of experiencing urinary symptoms after pelvic floor muscle training; however, there was substantial heterogeneity among trials and this estimate was imprecise (95% CI: −0.3, 0.1; I2 = 73%) (Fig. 3). The quality of the evidence was rated as low (downgraded due to inconsistency of estimates and indirectness of participants). One additional trial50 reported that pelvic floor muscle training might reduce the number of voids/day (MD −1; 95% CI: −2.7, 0.6; n = 49).

Fig. 3.

(a) Risk difference (95% CI) of pelvic floor muscle training versus no intervention (n = 429) for urinary symptoms (number of women with urinary symptoms); (b) Risk difference (95% CI) of pelvic floor muscle training versus no intervention (n = 387) for vaginal prolapse (number of women with vaginal prolapse); (c) Mean difference (95% CI) of pelvic floor muscle training versus no intervention (n = 272) for sexual function (2–36 points); d) Mean difference (95% CI) of pelvic floor muscle training versus no intervention (n = 382) for pelvic floor muscle strength (cmH2O); e) Standardized Mean difference (95% CI) of pelvic floor muscle training versus no intervention (n = 297) for quality of life (0–108).

(0.78MB).

Vaginal prolapse: The effects of pelvic floor muscle training on vaginal prolapse were examined by two trials,47,48 involving 387 participants. The mean risk difference was 0 (95% CI: −0.1, 0.1; I2 = 37%), which suggests no or negligible benefits on prolapse after pelvic floor muscle training (Fig. 3). The quality of the evidence was rated as moderate (downgraded due to indirectness of participants). One trial50 did not provide usable data for pooling/calculating estimates.

Sexual function: The effects of pelvic floor muscle training on sexual function were examined by pooling post-intervention data from two trials,49,50 involving 272 participants. Pelvic floor muscle training improved sexual function by 5 points (95% CI: 4, 6; I2 = 0) on the Female Sexual Function Index (2 to 36 points) (Fig. 3). The quality of the evidence was rated as moderate (downgraded due to imprecision). One trial,53 which provided change score data on the domain of sexual function of the Pelvic Floor Questionnaire, also reported benefits in favor of pelvic floor muscle training (MD −3 points out of 10; 95% CI: – 6, −1; n = 24). One trial,54 which did not provide usable data for pooling or calculating estimates, reported no between group-differences after intervention.

Pelvic floor muscle strength: The effects of pelvic floor muscle training on strength were examined by pooling change score data from three trials,47,48,54 involving 382 participants. The best estimate suggested that pelvic floor muscle training improves strength by SMD 0.5; however, there was substantial heterogeneity among trials and this estimate was very imprecise (95% CI: −0.4, 1.3; I2 = 81 %) (Fig. 3). The quality of the evidence was rated as low (downgraded due to inconsistency of estimates and indirectness of participants). One trial,50 which provided post-intervention scores on the Modified Oxford Scale, reported no benefits on strength after pelvic floor muscle training (MD 0.1; 95% CI: −0.7, 0.8; n = 47). One trial52 did not provide usable data for pooling/calculating estimates.

Quality of life: The effects of pelvic floor muscle training on quality of life were examined by pooling post-intervention data from three trials,49,50,54 involving 297 participants. The standardized mean difference was 0.5 (95% CI: −0.1, 0.9; I2 = 70%) (Figure 3), which provides no clear evidence in favor of pelvic floor muscle training. The quality of the evidence was rated as very low (downgraded due to low methodological quality, inconsistency of estimates, and imprecision). Three additional trials47,48,52 also reported no benefits on quality of life after pelvic floor muscle training but only one48 provided data for calculating estimates, based on change scores (MD 0; 95% CI: −0.1, 0.1; n = 43).

Two trials examined the maintenance of benefits six51 and twenty-four47 months after the intervention period, but the results could not be combined in meta-analyses due to different type and timing of measurements. One trial,51 involving 49 participants, provided imprecise results for urinary symptoms (MD −0.8 voids/day; 95% CI: −2.6, 0.9), sexual function (MD 4 points out of 36; 95% CI: −4, 13), pelvic muscle strength (MD −0.3 points out 5; 95% CI: −1.0, 0.4), and quality of life (MD −0.6 points out of 100; 95% CI: −1.7, 2.8). The second trial,47 involving 308 participants, indicated no long-term difference between-groups for urinary symptoms (RD 0; 95% CI: −0.1, 0.1), vaginal prolapse (RD 0;95% CI: −0.1, 0.1), and pelvic muscle strength (MD 0 points; 95% CI: −0.4, 0.5).