Pessaries ( mechanical devices ) for managing pelvic organ prolapse in women

Background Pelvic organ prolapse is a common problem in women. About 40% of women will experience prolapse in their lifetime, with the proportion expected to rise in line with an ageing population. Women experience a variety of troublesome symptoms as a consequence of prolapse, including a feeling of 'something coming down' into the vagina, pain, urinary symptoms, bowel symptoms and sexual diIiculties. Treatment for prolapse includes surgery, pelvic floor muscle training (PFMT) and vaginal pessaries. Vaginal pessaries are passive mechanical devices designed to support the vagina and hold the prolapsed organs back in the anatomically correct position. The most commonly used pessaries are made from polyvinyl-chloride, polythene, silicone or latex. Pessaries are frequently used by clinicians with high numbers of clinicians oIering a pessary as first-line treatment for prolapse. This is an update of a Cochrane Review first published in 2003 and last published in 2013. Objectives To assess the eIects of pessaries (mechanical devices) for managing pelvic organ prolapse in women; and summarise the principal findings of relevant economic evaluations of this intervention. Search methods We searched the Cochrane Incontinence Specialised Register which contains trials identified from the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, MEDLINE In-Process, MEDLINE Epub Ahead of Print, ClinicalTrials.gov, WHO ICTRP and handsearching of journals and conference proceedings (searched 28 January 2020). We searched the reference lists of relevant articles and contacted the authors of included studies. Selection criteria We included randomised and quasi-randomised controlled trials which included a pessary for pelvic organ prolapse in at least one arm of the study. Pessaries (mechanical devices) for managing pelvic organ prolapse in women (Review) Copyright © 2020 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 1 Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews Data collection and analysis Two review authors independently assessed abstracts, extracted data, assessed risk of bias and carried out GRADE assessments with arbitration from a third review author if necessary. Main results We included four studies involving a total of 478 women with various stages of prolapse, all of which took place in high-income countries. In one trial, only six of the 113 recruited women consented to random assignment to an intervention and no data are available for those six women. We could not perform any meta-analysis because each of the trials addressed a diIerent comparison. None of the trials reported data about perceived resolution of prolapse symptoms or about psychological outcome measures. All studies reported data about perceived improvement of prolapse symptoms. Generally, the trials were at high risk of performance bias, due to lack of blinding, and low risk of selection bias. We downgraded the certainty of evidence for imprecision resulting from the low numbers of women participating in the trials. Pessary versus no treatment: at 12 months' follow-up, we are uncertain about the eIect of pessaries compared with no treatment on perceived improvement of prolapse symptoms (mean diIerence (MD) in questionnaire scores -0.03, 95% confidence interval (CI) -0.61 to 0.55; 27 women; 1 study; very low-certainty evidence), and cure or improvement of sexual problems (MD -0.29, 95% CI -1.67 to 1.09; 27 women; 1 study; very low-certainty evidence). In this comparison we did not find any evidence relating to prolapse-specific quality of life or to the number of women experiencing adverse events (abnormal vaginal bleeding or de novo voiding diIiculty). Pessary versus pelvic floor muscle training (PFMT): at 12 months' follow-up, we are uncertain if there is a diIerence between pessaries and PFMT in terms of women's perceived improvement in prolapse symptoms (MD -9.60, 95% CI -22.53 to 3.33; 137 women; low-certainty evidence), prolapse-specific quality of life (MD -3.30, 95% CI -8.70 to 15.30; 1 study; 116 women; low-certainty evidence), or cure or improvement of sexual problems (MD -2.30, 95% -5.20 to 0.60; 1 study; 48 women; low-certainty evidence). Pessaries may result in a large increase in risk of adverse events compared with PFMT (RR 75.25, 95% CI 4.70 to 1205.45; 1 study; 97 women; low-certainty evidence). Adverse events included increased vaginal discharge, and/or increased urinary incontinence and/or erosion or irritation of the vaginal walls. Pessary plus PFMT versus PFMT alone: at 12 months' follow-up, pessary plus PFMT probably leads to more women perceiving improvement in their prolapse symptoms compared with PFMT alone (RR 2.15, 95% CI 1.58 to 2.94; 1 study; 260 women; moderate-certainty evidence). At 12 months' follow-up, pessary plus PFMT probably improves women's prolapse-specific quality of life compared with PFMT alone (median (interquartile range (IQR)) POPIQ score: pessary plus PFMT 0.3 (0 to 22.2); 132 women; PFMT only 8.9 (0 to 64.9); 128 women; P = 0.02; moderate-certainty evidence). Pessary plus PFMT may slightly increase the risk of abnormal vaginal bleeding compared with PFMT alone (RR 2.18, 95% CI 0.69 to 6.91; 1 study; 260 women; low-certainty evidence). The evidence is uncertain if pessary plus PFMT has any eIect on the risk of de novo voiding diIiculty compared with PFMT alone (RR 1.32, 95% CI 0.54 to 3.19; 1 study; 189 women; lowcertainty evidence). Authors' conclusions We are uncertain if pessaries improve pelvic organ prolapse symptoms for women compared with no treatment or PFMT but pessaries in addition to PFMT probably improve women's pelvic organ prolapse symptoms and prolapse-specific quality of life. However, there may be an increased risk of adverse events with pessaries compared to PFMT. Future trials should recruit adequate numbers of women and measure clinically important outcomes such as prolapse specific quality of life and resolution of prolapse symptoms. The review found two relevant economic evaluations. Of these, one assessed the cost-eIectiveness of pessary treatment, expectant management and surgical procedures, and the other compared pessary treatment to PFMT. P L A I N   L A N G U A G E   S U M M A R Y Pessaries (mechanical devices) for managing pelvic organ prolapse in women Review question How eIective are pessaries (mechanical devices) for managing pelvic organ prolapse in women? Background Pelvic organs, such as the uterus, bladder or bowel, may protrude into the vagina because of weakness in the tissues that normally support them. This protrusion is known as pelvic organ prolapse. Women who experience prolapse report a wide variety of symptoms that aIect their quality of life. The choice of treatment options for pelvic organ prolapse symptoms is dictated largely by patient choice, as the interventions of lifestyle changes, pelvic floor muscle training (PFMT), pessary and surgery are very diIerent with diIerent complication risks. Vaginal pessaries are one treatment option for prolapse that are commonly used to restore the prolapsed organs to their normal position and hence relieve symptoms. A vaginal pessary can be used to treat all four prolapse stages. Pessaries (mechanical devices) for managing pelvic organ prolapse in women (Review) Copyright © 2020 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 2 Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews How up-to-date is this review? We searched for studies published up to 28 January 2020. Study characteristics We found four studies involving 478 women with various stages of prolapse. All four studies were carried out in high-income countries. Key results None of the studies reported whether or not the women's prolapse symptoms were completely resolved. We are uncertain about the eIect of pessaries compared with no treatment on perceived improvement of prolapse symptoms, and cure or improvement of sexual problems. We did not find any evidence relating to the risk of vaginal bleeding or diIiculty urinating in the studies comparing pessaries with no treatment. We are uncertain if pessaries have any eIect on improving women's prolapse symptoms when pessary is compared with PFMT. The evidence is also very uncertain about the eIect of pessaries compared with PFMT on prolapse-specific quality of life and on sexual problems. Pessaries may result in a greater risk of adverse events, including vaginal discharge, increased urinary incontinence and erosion of the vaginal walls, compared with PFMT. Pessary plus PFMT probably leads to more women perceiving improvement in their prolapse symptoms and in their prolapse-specific quality of life compared with PFMT alone. Compared with PFMT alone, pessary plus PFMT may slightly increase the risk of abnormal vaginal bleeding and the risk of having diIiculty urinating, for women who did not have this problem before starting prolapse treatment, but the evidence is very uncertain. Certainty of the evidence We assessed the evidence relating to pessary compared with no treatment as very low certainty because of the low numbers of women participating in the study and because of problems with the way the trial was conducted. We assessed the evidence relating to pessary compared with PFMT, and pessary plus PFMT compared with PFMT alone, as lowto moderate-certainty because of the low numbers of women involved in the studies. Authors' conclusions Each of these trials contain small numbers of women, reducing the certainty of our conclusions. We are uncertain if pessaries improve pelvic organ prolapse symptoms for women compared with no treatment or when compared to another active treatment like PFMT, but pessaries in addition to PFMT probably improve women's symptoms and prolapse-specific quality of life. However, there may be an increased risk of adverse events with pessaries compared to no treatment or PFMT. The review found two economic evaluation studies that compared pessary treatment to alternative interventions (PFMT, expectant management and surgical procedures). Pessaries (mechanical devices) for managing pelvic organ prolapse in women (Review) Copyright © 2020 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 3 Pesaries (m ecnical deices) or m anging plvic ogan prlapse in w om en (Rview ) Coyright © 2020 he Cohrane Colloration. Pulished by John W ley & Sns, Ltd. 4 S U M M A R Y   O F   F I N D I N G S Summary of findings 1.   Any pessary compared to control, waiting list or no active treatment Any pessary compared to standard care Patient or population: pelvic organ prolapse in women Setting: urogynaecological clinic Intervention: any pessary Comparison: standard care Anticipated absolute effects* (95% CI) Outcomes Risk with standard care Risk with any pessary Relative effect (95% CI) No of participants (studies) Certainty of the evidence (GRADE) Comments Perceived resolution of prolapse symptoms: ≥ 12 months Not reported Perceived improvement of prolapse symptoms (higher score = worse symptoms) assessed with: POP score of Australian Pelvic Floor Questionnaire range: 0-10 Follow-up: 12 months The mean score in the comparator group was 0.31 MD 0.03 lower (0.61 lower to 0.55 higher) 27 (1 RCT) ⊕⊝⊝⊝ VERY LOW 1 2 Prolapse-specific quality of life ≥ 12 months Not reported Cure or improvement of sexual problems assessed with: sex score of the Australian Pelvic Floor Questionnaire score: 0-10 (higher score = worse symptoms) Follow-up: 12 months The mean score in the comparator group was 1.45 MD 0.29 lower (1.67 lower to 1.09 higher) 27 (1 RCT) ⊕⊝⊝⊝ VERY LOW 1 2 Psychological outcome measures at ≥ 12 months Not reported Number of women with adverse events (vaginal bleeding, voiding difficulty) Not reported *The risk in the intervention group (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; MD: mean difference; POP: pelvic organ prolapse; RCT: randomised controlled trial. GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Cohrane Lrary Tsted evence. Iform ed desions. Btter hlth. Cohrane Daabase of Sstem tic Rview s Pesaries (m ecnical deices) or m anging plvic ogan prlapse in w om en (Rview ) Coyright © 2020 he Cohrane Colloration. Pulished by John W ley & Sns, Ltd. 5 Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect. 1 Downgraded one level for risk of bias (performance and attrition bias) 2 Downgraded two levels for imprecision (few participants and wide 95% CI crossing the line of no eIect) Summary of findings 2.   Any pessary compared to pelvic floor muscle training (PFMT) for pelvic organ prolapse in women Any pessary compared to PFMT or pelvic organ prolapse in women Patient or population: pelvic organ prolapse in women Setting: urogynaecological clinic Intervention: any pessary Comparison: PFMT Anticipated absolute effects* (95% CI) Outcomes Risk with PFMT Risk with pessary Relative effect (95% CI) No of participants (studies) Certainty of the evidence (GRADE) Comments Perceived resolution of prolapse symptoms ≥ 12 months Not reported Women's perceived improvement of prolapse symptoms (higher score = worse symptoms) assessed with: POPDI-6 Scale from: 0 to 100 Follow-up: 12 months The mean POPDI-6 score in the PFMT group was 16.4 MD 3.60 lower (8.73 lower to 1.53 higher) 117 (1 RCT) ⊕⊕⊝⊝ LOW 1 Prolapse-specific quality of life (higher score = worse QoL) assessed with: PFIQ-7 Scale from: 0 to 300 Follow-up: 12 months The mean PFIQ-7 score in the PFMT group was 15.8 MD 3.3 higher (8.70 lower to 15.30 higher) 116 (1 RCT) ⊕⊕⊝⊝ LOW 1 Cure or improvement of sexual problems (higher score = better sexual functioning) assessed with: PISQ-12, Scale from: 0 to 48 Follow-up: 12 months The mean PISQ-12 score in the PFMT group was 37.6 MD 2.30 lower (5.20 lower to 0.60 higher) 48 (1 RCT) ⊕⊕⊝⊝ LOW 1 Cohrane Lrary Tsted evence. Iform ed desions. Btter hlth. Cohrane Daabase of Sstem tic Rview s Pesaries (m ecnical deices) or m anging plvic ogan prlapse in w om en (Rview ) Coyright © 2020 he Cohrane Colloration. Pulished by John W ley & Sns, Ltd. 6 Psychological outcome measures at ≥ 12 months Not reported Number of women with adverse events 21/35 women in pessary group and 0/62 women in PFMT group experienced adverse events (14 increased vaginal discharge, 5 increased urinary incontinence, 10 erosion or irritation of the vaginal walls). RR 75.25 (4.70 to 1205.45) 97 (1 RCT) ⊕⊕⊝⊝ LOW 2 Data only available at 24 months' follow-up. *The risk in the intervention group (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; MD: mean difference;PFIQ: Pelvic Floor Impact Questionnaire; PISQ: Prolapse/Urinary Incontinence Sexual Questionnaire; POPDI-6: Pelvic Organ Prolapse Distress Inventory 6; QI: quality of life; RCT: randomised controlled trial; RR: Risk ratio. GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect. 1 Downgraded two levels due to imprecision: few participants and wide 95% CI, consistent with possible benefit and possible harm 2 Downgraded two levels due to imprecision: few participants and few events, and the 95% CI suggests event rates that are implausibly wide Summary of findings 3.   Pessary plus PFMT compared to pelvic floor muscle training (PFMT) alone for pelvic organ prolapse in women Pessary plus PFMT compared to PFMT alone for pelvic organ prolapse in women Patient or population: pelvic organ prolapse in women Setting: urogynaecological clinic Intervention: pessary plus PFMT Comparison: PFMT alone Anticipated absolute effects* (95% CI) Outcomes Risk with PFMT alone Risk with pessary plus PFMT Relative effect (95% CI) No of participants (studies) Certainty of the evidence (GRADE) Comments Perceived resolution of prolapse symptoms ≥ 12 months Not reported Cohrane Lrary Tsted evence. Iform ed desions. Btter hlth. Cohrane Daabase of Sstem tic Rview s Pesaries (m ecnical deices) or m anging plvic ogan prlapse in w om en (Rview ) Coyright © 2020 he Cohrane Colloration. Pulished by John W ley & Sns, Ltd. 7 Perceived improvement of prolapse symptoms Follow-up: 12 months 281 per 1000 605 per 1000 (444 to 827) RR 2.15 (1.58 to 2.94) 260 (1 RCT) ⊕⊕⊕⊝ MODERATE 1 Prolapse-specific quality of life assessed with: POPIQ (higher score = worse quality of life) (scale: 0-100 Follow-up: 12 months Median (IQR) scores: Pessary + PFMT = 0.3 (0–22.2) (132 women); PFMT = 8.9 (0–64.9) (128 women) P = 0.02 260 (1 RCT) ⊕⊕⊕⊝ MODERATE 1 Cure or improvement of sexual problems ≥ 12 months Not reported Psychological outcome measures at ≥ 12 months Not reported Number of women with adverse events abnormal vaginal bleeding Follow-up: 12 months 31 per 1000 68 per 1000 (22 to 216) RR 2.18 (0.69 to 6.91) 260 (1 RCT) ⊕⊕⊝⊝ LOW 2 Number of women with adverse events de novo voiding difficulty Follow-up: 12 months 82 per 1000 109 per 1000 (45 to 263) RR 1.32 (0.54 to 3.19) 189 (1 RCT) ⊕⊕⊝⊝ LOW 2 *The risk in the intervention group (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; ; POPIQ: Pelvic Organ Prolapse Impact Questionnaire; RCT: randomised controlled trial; RR: Risk ratio. GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate; the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited; the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate; the true effect is likely to be substantially different from the estimate of effect. 1 Downgraded one level for imprecision: few participants 2 Downgraded two levels for imprecision: few participants and wide 95% CI spanning possible harm and possible benefit Cohrane Lrary Tsted evence. Iform ed desions. Btter hlth. Cohrane Daabase of Sstem tic Rview s Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews B A C K G R O U N D Description of the condition Pelvic organ prolapse is the descent of some part, or parts, of the female pelvic organs (uterus, bladder and rectum) from their normal anatomical position into the vagina. It aIects about 40% of women aged over 40 (Hendrix 2002), with the prevalence expected to increase as the population ages (Wu 2011). Pelvic organ prolapse includes anterior vaginal wall prolapse (cystocoele, urethrocoele), posterior vaginal wall prolapse (enterocoele, rectocoele, perineal deficiency), and uterine or vaginal vault prolapse. Women can present with prolapse in one or more of these sites. The International Continence Society has standardised the naming of the condition using the POP-Q (Pelvic Organ Prolapse Quantification) evaluation (Bump 1996), but in this review we have also used the descriptive terms above as these are compatible with literature searches. The cause of pelvic organ prolapse is complex and multifactorial. Various risk  factors have been proposed, including pregnancy, childbirth, hereditary factors, denervation or weakness of the pelvic floor, ageing, menopause, and factors associated with chronically raised intra-abdominal pressure (such as obesity, coughing and heavy liTing) (Bump 1998; Dietz 2008; Gyhagen 2015; Jelovsek 2018; MacLennan 2000). Women with prolapse may have a variety of pelvic floor symptoms (Barber 2001; Hagen 2009; Jelovsek 2006; Lone 2011). Women experience symptoms of 'something coming down', pelvic heaviness, a dragging sensation in the vagina, a bulge, lump or protrusion coming down from the vagina and backache. Women can also experience symptoms related to the movement of the pelvic organs and their protrusion into the vagina, which include urinary symptoms, bowel symptoms and sexual diIiculties. Symptoms, and perceived bother of symptoms, may negatively aIect a woman's body image, quality of life and mental health, as well as their ability to perform day to day activities (Jelovsek 2006; Lowder 2011; Pakbaz 2010). Description of the intervention Prolapse treatment may be dependent on a number of factors, including the severity of prolapse, the bother of the associated symptoms, the woman's general health and the woman's treatment preference (Basu 2011; Kapoor 2009). Various treatment options exist for women who have pelvic organ prolapse and these can be split into surgical and conservative options. Approximately 9.5% of women will undergo prolapse surgery in their lifetime (Abdel-Fattah 2011). A US-based study forecast that approximately 9.2 million women will be aIected by pelvic organ prolapse by 2050 (Wu 2009), and about 300,000 women undergo pelvic organ prolapse surgery in the USA every year costing over 1 billion US dollars (the cost year was unclear) (Enemchukwu 2019; Subak 2001). The NHS spent over £45 million on the treatment of prolapse between 2017 and 2018 (Primary Care Women's Health Forum 2019). Vaginal pessaries are passive mechanical devices that are worn internally and designed to support the vaginal walls and apex. An extensive range of pessaries have been described for treating prolapse (Oliver 2011; Poma 2000). These consist mainly of PVC, latex or silicone pessaries, which are shaped devices inserted into and leT inside the vagina to support the prolapsed pelvic organs. Two main groups of pessaries are used: support pessaries (e.g. ring pessary or ring pessary with support) and space filling pessaries (e.g. Gellhorn or cube) (Dwyer 2019; Oliver 2011). Support pessaries are inserted into the vagina and are positioned between the pubic bone and posterior vaginal fornix, providing support to descending organs. They do not prohibit vaginal intercourse and are easier to remove and replace. Space filling pessaries provide support by filling the vaginal space to prevent prolapse descent, creating a suction eIect around the pessary which thereby increases the likelihood of retention. They are not compatible with vaginal intercourse and are more diIicult to remove and replace (Dwyer 2019). Studies suggest that ring pessaries are the most commonly used in practice (Bugge 2013a; CundiI 2000). Recently published NICE guidance suggests considering pessary treatment (alone or with: pelvic floor muscle training (PFMT)) when women have symptomatic prolapse (NICE 2019). Surveys suggest that 87% to 98% of clinicians report using pessaries in their clinical practice (Bugge 2013a; CundiI 2000; Gorti 2009; Pott-Grinstein 2001), while 77% of gynaecologists report using pessaries as a firstline treatment for prolapse (CundiI 2000). Health professionals (87.6%) report that they are most likely to fit pessaries in women aged 60 or older and 98.8% would use a pessary for treatment when a woman is unfit for surgery (Bugge 2013a). Evidence from observational studies suggests that about 76% of women who try a pessary continue to use it for at least four weeks and, of those women who continue pessary use for more than four weeks, 86% continued to use the pessary for over five years (Lone 2011). Furthermore, observational studies have shown that pessaries may improve symptoms for women (Lamers 2011; Manchana 2012). How the intervention might work Pessaries are used in pelvic organ prolapse in order to physically support the vaginal walls and the pelvic organs behind them. The pessary is inserted into the vagina with a view to holding the prolapsed organs, supporting the pelvic structures, and relieving pressure on the bladder and bowel (Dumoulin 2017). Using a pessary in the management of pelvic organ prolapse aims to prevent the prolapse from becoming worse, as well as helping to decrease the frequency or severity of symptoms of prolapse, and averting or delaying the need for surgery (Oliver 2011). Why it is important to do this review The eIectiveness of surgical treatment options for prolapse are reported in other Cochrane Reviews (Baessler 2018; Maher 2016a; Maher 2016b; Maher 2016c; Mowat 2018). Conservative treatment options that are currently used are pelvic floor muscle training (PFMT), vaginal pessaries and lifestyle advice. The eIectiveness of PFMT and lifestyle advice is reported in one review (Hagen 2011a). The remaining conservative option, vaginal pessaries, is the focus of this review. The majority of evidence for the use of pessaries comes from nonrandomised studies (e.g. Brazell 2014; Clemons 2004b; Clemons 2004a; Clemons 2004c; Hanson 2006; Kapoor 2009; Lamers 2011; Lone 2011; Manchana 2012). As such, the eIectiveness of pessaries for managing pelvic organ prolapse still needs to be clearly established. There is a lack of consensus as to the optimal treatment for women. Provided that suIicient numbers of trials of adequate quality have been conducted, the most reliable evidence Pessaries (mechanical devices) for managing pelvic organ prolapse in women (Review) Copyright © 2020 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 8 Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews is likely to come from consideration of randomised controlled trials (RCTs), which is the basis for the current review. The aim is to help identify optimal practice and highlight where there is a need for further research. Given that diIerent costs are associated with the alternative interventions included in this review, it is important to conduct a brief economic commentary (BEC) to find available evidence (if any) on the costs of the interventions relative to their eIectiveness. O B J E C T I V E S To assess the eIects of pessaries (mechanical devices) for managing pelvic organ prolapse in women; and summarise the principal findings of relevant economic evaluations of this intervention. M E T H O D S Criteria for considering studies for this review Types of studies We included randomised trials (RCTs) and quasi-randomised trials in which at least one arm was a pessary for pelvic organ prolapse. We excluded cross-over studies. Types of participants We included studies focusing on adult women who were seeking treatment for pelvic organ prolapse. Participants could have prolapse with any POP-Q staging (Bump 1996), and prolapse of any type: anterior vaginal wall prolapse (cystocoele, urethrocoele); posterior vaginal wall prolapse (enterocoele, rectocoele, perineal deficiency); and uterine or vaginal vault prolapse. We included trials where women were seeking treatment for urinary incontinence (UI) or prolapse if the data for women with prolapse (including women with prolapse and UI) were considered, or available separately. Types of interventions We included studies where one arm involved allocation to a pessary aiming to provide support for pelvic organ prolapse. Thus, we excluded studies where the pessary was used for exercise (and not support) or where pessaries were only used to deliver drug treatments (such as oestrogen). Pessaries could be of any type and made of any material. The types of devices included support pessaries (e.g. ring, ring with support) and space-filling devices (e.g. shelf, Gellhorn, cube). We considered the following comparisons and planned to present 'Summary of findings' tables for the first three. • Any pessary versus control, waiting list or no active treatment. • Any pessary versus another treatment (lifestyle interventions, oestrogen treatment, physical interventions such as pelvic floor muscle training, surgery). • Any pessary plus another treatment versus the other treatment alone. • One pessary versus another pessary. Types of outcome measures Primary outcomes • Women's perceived resolution of prolapse symptoms • Women's perceived improvement in prolapse symptoms (assessed using validated symptom questionnaires such as the Pelvic Floor Disability Index (PFDI) (Barber 2001; Barber 2005), or the Pelvic Organ Symptom Score (POP-SS) (Hagen 2009)) Secondary outcomes • Patient-reported satisfaction with pessary treatment • Grade of prolapse (i.e. the compartment that has the most descent, e.g. using Pelvic Organ Prolapse-Quantification (POPQ) classification (Bump 1996)) • Site-specific grading of prolapse (i.e. the descent in each of the specific anterior, apical or posterior compartments, e.g. using POP-Q (Bump 1996)) • Prolapse-specific quality of life (measured with validated instrument, e.g. the Pelvic Floor Impact Questionnaire (PFIQ), Barber 2001; Barber 2005) • Quality of life measured with generic quality of life or health status measures (e.g. SF-36, Ware 1993) • Psychological outcome measures (e.g. Hospital Anxiety and Depression Scale (HADS), Zigmond 1983) • Cure or improvement of bladder problems (including urinary incontinence, occult incontinence and relief of voiding diIiculty) • Cure or improvement of bowel problems (including relief of obstructed defecation) • Cure or improvement of sexual problems (including acceptability of device to both partners) • Number of women with adverse events (vaginal bleeding, voiding diIiculty) Timing of outcome assessment We considered all outcomes at ≥ 12 months. Main outcomes for 'Summary of findings' tables These outcomes have been informed by a recent priority-setting exercise for pessary use (Lough 2018). • Woman's perceived resolution of prolapse symptoms at ≥ 12 months • Woman's perceived improvement in prolapse symptoms assessed using a validated symptom questionnaire at ≥ 12 months • Prolapse-specific quality of life (e.g. PFIQ) at ≥ 12 months • Cure or improvement of sexual problems at ≥ 12 months • Psychological outcome measures at ≥ 12 months • Number of women with adverse events (vaginal bleeding, voiding diIiculty) Search methods for identification of studies We did not impose any language or other restrictions on any of the searches detailed below. Pessaries (mechanical devices) for managing pelvic organ prolapse in women (Review) Copyright © 2020 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 9 Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews Electronic searches Search for clinical e ectiveness studies We identified relevant trials from the Cochrane Incontinence Specialised Register. For more details of the search methods used to build the Specialised Register, please see the Group's webpages where details of the Register's development (from inception) and the most recent searches performed to populate the Register can be found. To summarise, the Register contains trials identified from the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, MEDLINE In-Process, MEDLINE Epub Ahead of Print, ClinicalTrials.gov, WHO ICTRP, Be Part of Research and handsearching of journals and conference proceedings. Many of the trials in the Cochrane Incontinence Specialised Register are also contained in CENTRAL. The terms used to search the Cochrane Incontinence Group Specialised Register are given in Appendix 1. The date of the most recent search of the Register for this review: 28 January 2020. The search methods and strategies used for some of the previous versions of this review are given in Appendix 2. Search for economic evaluations We performed additional searches for the brief economic commentary (BEC). We searched the following. • The NHS Economic Evaluation Database (NHS EED) on the Centre for Reviews and Dissemination (CRD) website (covering from the earliest record in NHS EED, dating from 1968, up to and including 31 December 2014 when their coverage ended). Date of search: 13 February 2020. As NHS EED is no longer actively updated, we performed additional searches of the following databases to identify eligible studies added to these databases from 1 January 2015 onwards (date of search: 13 February 2020): • MEDLINE on OvidSP (covering 1 January 1946 to January February 1 2020); and • Embase (on OvidSP) (covering 1 January 1974 to 2020 Week 06). Details of the searches that were performed can be found in Appendix 3. Searching other resources We searched the reference lists of relevant articles for other possibly relevant trials. We also contacted the authors of included abstracts. Data collection and analysis We conducted data collection and analysis in accordance with methods specified in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019). Selection of studies Two review authors independently screened all potentially eligible abstracts from the search (CB and DG or MD). We nominated a third review author to arbitrate in the event of disagreement (EJA). We excluded any studies which were not relevant at this stage. We then obtained the full-text articles of relevant studies. If there was any uncertainty on the eligibility of the studies based on title and abstract, we obtained the full paper, which was reviewed by the same two review authors. Data extraction and management Two review authors independently undertook data extraction using a predefined data extraction form (from CB, RK and FS). We then made comparisons to ensure accuracy. We resolved any discrepancies by discussion with, or referral to, a third party (EJA). Assessment of risk of bias in included studies We assessed the risk of bias within studies as part of the data extraction. We used the methods outlined in the Cochrane Handbook to assess risk of bias in the following domains (Higgins 2011): random sequence generation (selection bias); allocation concealment (selection bias); blinding of participants and personnel (detection bias); blinding of outcome assessors (performance bias); incomplete outcome data (attrition bias); selective reporting (reporting bias); other bias not considered under other domains. Two review authors assessed the included studies for risk of bias and any discrepancies were resolved by discussion and referred to a third author where necessary. Measures of treatment e<ect For binary outcomes, we have reported risk ratios (RR) and 95% confidence intervals (CI). For continuous outcomes, we reported mean diIerences (MDs) with 95% CIs. Unit of analysis issues The primary analysis was per woman randomised. Had data allowed, we would have analysed each pair of arms in multi-arm studies as a separate comparison. To avoid double-counting where the same study with three treatment arms is included twice in the same meta-analysis, we planned to divide the number of women and events by two in the group that appears twice in the analysis. Dealing with missing data We contacted study authors to obtain missing data in order to be able to conduct intention-to-treat analysis wherever possible. Where randomised data were not available, we have reported data from per protocol analyses as presented in the study publications. We have clearly indicated in the EIects of interventions where we have used per protocol data. Assessment of heterogeneity We intended to combine studies in meta-analysis where clinical and methodological heterogeneity was low; that is, studies had to be broadly similar in terms of participant characteristics, interventions, study design and outcomes in order to be pooled. Had data allowed, we would have assessed statistical heterogeneity in meta-analyses as follows, taking into account the magnitude and direction of eIects as well as the P value and/or confidence interval for I2: • 0% to 40%: heterogeneity might not be important • 30% to 60% might represent moderate heterogeneity • 50% to 90% might represent substantial heterogeneity • 75% to 100% represents considerable heterogeneity Pessaries (mechanical devices) for managing pelvic organ prolapse in women (Review) Copyright © 2020 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 10 Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews Assessment of reporting biases Using the data extraction form, we assessed for data that could have been collected but was not reported. Where there were 10 or more studies in the same meta-analysis and where data allowed, we would have investigated reporting biases (such as publication bias) using funnel plots. If funnel plot asymmetry was suggested by a visual assessment, we would have performed exploratory analyses to investigate it. Data synthesis Had data allowed, we would have used a fixed-eIect metaanalysis for combining data if it was reasonable to assume that the studies were estimating the same underlying treatment eIect (i.e. where trials were examining the same intervention, and the trials’ populations and methods were judged suIiciently similar). If there was clinical heterogeneity suIicient to expect that the underlying treatment eIects diIered between trials, or if substantial statistical heterogeneity was detected, we would have used a random-eIects meta-analysis to produce an overall summary. Subgroup analysis and investigation of heterogeneity Had data allowed, we planned to carry out subgroup analysis according to diIerent prolapse stages or compartments. Sensitivity analysis Had data allowed, we would have conducted sensitivity analyses to examine the eIect of studies with high risk of bias on the results. We would have repeated the main meta-analyses without studies judged to be at high risk and compare those results to the main analyses with all studies included. Incorporating economics evidence Once the search outlined in the Search methods for identification of studies was performed, we developed a brief economic commentary (BEC) to summarise the availability and principal findings of the full economic evaluations that compared pessaries with alternative treatment interventions for pelvic organ prolapse in women (Shemilt 2019). This BEC encompassed full economic evaluations (i.e. cost-eIectiveness analyses, cost-utility analyses and cost-benefit analyses), conducted alongside or based on one or more studies included in the main review of intervention eIects, or model-based economic evaluations. Summary of findings and assessment of the certainty of the evidence We prepared 'Summary of findings' tables using the GRADEpro GDT soTware for the comparisons pre-stated in the Types of interventions. We used the GRADE approach to assess the certainty of evidence related to the outcomes listed in the 'Main outcomes for 'Summary of findings' tables' in the Types of outcome measures (Schünemann 2019). We used the five GRADE considerations (study limitations, consistency of eIect, imprecision, indirectness and publication bias) to assess the certainty of the body of evidence for the prespecified outcomes. We justified all decisions to downgrade the certainty of studies using footnotes. R E S U L T S Description of studies Results of the search We reviewed 182 abstracts, of which 25 full-text reports (from 17 trials) were retrieved and assessed. Twelve reports of eleven trials were subsequently excluded, which are detailed in the Characteristics of excluded studies. Eleven reports of four studies were included in the review. Additionally several studies were identified within the search from their registration on clinical trial databases, and from these we identified a further two studies that, if completed, look likely to be included in future versions of this review. These studies are reported in the Characteristics of ongoing studies. The flow of literature through the assessment process for the update of this review is shown in Figure 1. Pessaries (mechanical devices) for managing pelvic organ prolapse in women (Review) Copyright © 2020 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd. 11 Cochrane Library Trusted evidence. Informed decisions. Better health. Cochrane Database of Systematic Reviews Figure 1.   PRISMA study flow diagram search for clinical e<ectiveness studies


A B S T R A C T Background
Pelvic organ prolapse is a common problem in women. About 40% of women will experience prolapse in their lifetime, with the proportion expected to rise in line with an ageing population. Women experience a variety of troublesome symptoms as a consequence of prolapse, including a feeling of 'something coming down' into the vagina, pain, urinary symptoms, bowel symptoms and sexual di iculties. Treatment for prolapse includes surgery, pelvic floor muscle training (PFMT) and vaginal pessaries. Vaginal pessaries are passive mechanical devices designed to support the vagina and hold the prolapsed organs back in the anatomically correct position. The most commonly used pessaries are made from polyvinyl-chloride, polythene, silicone or latex. Pessaries are frequently used by clinicians with high numbers of clinicians o ering a pessary as first-line treatment for prolapse. This is an update of a Cochrane Review first published in 2003 and last published in 2013.

Objectives
To assess the e ects of pessaries (mechanical devices) for managing pelvic organ prolapse in women; and summarise the principal findings of relevant economic evaluations of this intervention.

Search methods
We searched the Cochrane Incontinence Specialised Register which contains trials identified from the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, MEDLINE In-Process, MEDLINE Epub Ahead of Print, ClinicalTrials.gov, WHO ICTRP and handsearching of journals and conference proceedings (searched 28 January 2020). We searched the reference lists of relevant articles and contacted the authors of included studies.

Selection criteria
We included randomised and quasi-randomised controlled trials which included a pessary for pelvic organ prolapse in at least one arm of the study.

Library
Trusted evidence. Informed decisions. Better health.

Data collection and analysis
Two review authors independently assessed abstracts, extracted data, assessed risk of bias and carried out GRADE assessments with arbitration from a third review author if necessary.

Main results
We included four studies involving a total of 478 women with various stages of prolapse, all of which took place in high-income countries. In one trial, only six of the 113 recruited women consented to random assignment to an intervention and no data are available for those six women. We could not perform any meta-analysis because each of the trials addressed a di erent comparison. None of the trials reported data about perceived resolution of prolapse symptoms or about psychological outcome measures. All studies reported data about perceived improvement of prolapse symptoms.
Generally, the trials were at high risk of performance bias, due to lack of blinding, and low risk of selection bias. We downgraded the certainty of evidence for imprecision resulting from the low numbers of women participating in the trials.
Pessary versus no treatment: at 12 months' follow-up, we are uncertain about the e ect of pessaries compared with no treatment on perceived improvement of prolapse symptoms (mean di erence (MD) in questionnaire scores -0.03, 95% confidence interval (CI) -0.61 to 0.55; 27 women; 1 study; very low-certainty evidence), and cure or improvement of sexual problems (MD -0.29, 95% CI -1.67 to 1.09; 27 women; 1 study; very low-certainty evidence). In this comparison we did not find any evidence relating to prolapse-specific quality of life or to the number of women experiencing adverse events (abnormal vaginal bleeding or de novo voiding di iculty).
Pessary versus pelvic floor muscle training (PFMT): at 12 months' follow-up, we are uncertain if there is a di erence between pessaries and PFMT in terms of women's perceived improvement in prolapse symptoms (MD -9.60, 95% CI -22.53 to 3.33; 137 women; low-certainty evidence), prolapse-specific quality of life (MD -3.30, 95% CI -8.70 to 15.30; 1 study; 116 women; low-certainty evidence), or cure or improvement of sexual problems (MD -2.30, 95% -5.20 to 0.60; 1 study; 48 women; low-certainty evidence). Pessaries may result in a large increase in risk of adverse events compared with PFMT (RR 75.25, 95% CI 4.70 to 1205.45; 1 study; 97 women; low-certainty evidence). Adverse events included increased vaginal discharge, and/or increased urinary incontinence and/or erosion or irritation of the vaginal walls.

Authors' conclusions
We are uncertain if pessaries improve pelvic organ prolapse symptoms for women compared with no treatment or PFMT but pessaries in addition to PFMT probably improve women's pelvic organ prolapse symptoms and prolapse-specific quality of life. However, there may be an increased risk of adverse events with pessaries compared to PFMT. Future trials should recruit adequate numbers of women and measure clinically important outcomes such as prolapse specific quality of life and resolution of prolapse symptoms.
The review found two relevant economic evaluations. Of these, one assessed the cost-e ectiveness of pessary treatment, expectant management and surgical procedures, and the other compared pessary treatment to PFMT.

Review question
How e ective are pessaries (mechanical devices) for managing pelvic organ prolapse in women?

Background
Pelvic organs, such as the uterus, bladder or bowel, may protrude into the vagina because of weakness in the tissues that normally support them. This protrusion is known as pelvic organ prolapse. Women who experience prolapse report a wide variety of symptoms that a ect their quality of life. The choice of treatment options for pelvic organ prolapse symptoms is dictated largely by patient choice, as the interventions of lifestyle changes, pelvic floor muscle training (PFMT), pessary and surgery are very di erent with di erent complication risks. Vaginal pessaries are one treatment option for prolapse that are commonly used to restore the prolapsed organs to their normal position and hence relieve symptoms. A vaginal pessary can be used to treat all four prolapse stages.

Library
Trusted evidence. Informed decisions. Better health.

How up-to-date is this review?
We searched for studies published up to 28 January 2020.

Study characteristics
We found four studies involving 478 women with various stages of prolapse. All four studies were carried out in high-income countries.

Key results
None of the studies reported whether or not the women's prolapse symptoms were completely resolved. We are uncertain about the e ect of pessaries compared with no treatment on perceived improvement of prolapse symptoms, and cure or improvement of sexual problems. We did not find any evidence relating to the risk of vaginal bleeding or di iculty urinating in the studies comparing pessaries with no treatment.
We are uncertain if pessaries have any e ect on improving women's prolapse symptoms when pessary is compared with PFMT. The evidence is also very uncertain about the e ect of pessaries compared with PFMT on prolapse-specific quality of life and on sexual problems. Pessaries may result in a greater risk of adverse events, including vaginal discharge, increased urinary incontinence and erosion of the vaginal walls, compared with PFMT.
Pessary plus PFMT probably leads to more women perceiving improvement in their prolapse symptoms and in their prolapse-specific quality of life compared with PFMT alone. Compared with PFMT alone, pessary plus PFMT may slightly increase the risk of abnormal vaginal bleeding and the risk of having di iculty urinating, for women who did not have this problem before starting prolapse treatment, but the evidence is very uncertain.

Certainty of the evidence
We assessed the evidence relating to pessary compared with no treatment as very low certainty because of the low numbers of women participating in the study and because of problems with the way the trial was conducted. We assessed the evidence relating to pessary compared with PFMT, and pessary plus PFMT compared with PFMT alone, as low-to moderate-certainty because of the low numbers of women involved in the studies.

Authors' conclusions
Each of these trials contain small numbers of women, reducing the certainty of our conclusions. We are uncertain if pessaries improve pelvic organ prolapse symptoms for women compared with no treatment or when compared to another active treatment like PFMT, but pessaries in addition to PFMT probably improve women's symptoms and prolapse-specific quality of life. However, there may be an increased risk of adverse events with pessaries compared to no treatment or PFMT.
The review found two economic evaluation studies that compared pessary treatment to alternative interventions (PFMT, expectant management and surgical procedures). Cochrane Database of Systematic Reviews

Description of the condition
Pelvic organ prolapse is the descent of some part, or parts, of the female pelvic organs (uterus, bladder and rectum) from their normal anatomical position into the vagina. It a ects about 40% of women aged over 40 (Hendrix 2002), with the prevalence expected to increase as the population ages (Wu 2011). Pelvic organ prolapse includes anterior vaginal wall prolapse (cystocoele, urethrocoele), posterior vaginal wall prolapse (enterocoele, rectocoele, perineal deficiency), and uterine or vaginal vault prolapse. Women can present with prolapse in one or more of these sites. The International Continence Society has standardised the naming of the condition using the POP-Q (Pelvic Organ Prolapse Quantification) evaluation (Bump 1996), but in this review we have also used the descriptive terms above as these are compatible with literature searches.
The cause of pelvic organ prolapse is complex and multifactorial. Various risk factors have been proposed, including pregnancy, childbirth, hereditary factors, denervation or weakness of the pelvic floor, ageing, menopause, and factors associated with chronically raised intra-abdominal pressure (such as obesity, coughing and heavy li ing) (Bump 1998;Dietz 2008;Gyhagen 2015;Jelovsek 2018;MacLennan 2000).
Women with prolapse may have a variety of pelvic floor symptoms (Barber 2001;Hagen 2009;Jelovsek 2006;Lone 2011). Women experience symptoms of 'something coming down', pelvic heaviness, a dragging sensation in the vagina, a bulge, lump or protrusion coming down from the vagina and backache. Women can also experience symptoms related to the movement of the pelvic organs and their protrusion into the vagina, which include urinary symptoms, bowel symptoms and sexual di iculties. Symptoms, and perceived bother of symptoms, may negatively a ect a woman's body image, quality of life and mental health, as well as their ability to perform day to day activities (Jelovsek 2006;Lowder 2011;Pakbaz 2010).

Description of the intervention
Prolapse treatment may be dependent on a number of factors, including the severity of prolapse, the bother of the associated symptoms, the woman's general health and the woman's treatment preference (Basu 2011;Kapoor 2009). Various treatment options exist for women who have pelvic organ prolapse and these can be split into surgical and conservative options. Approximately 9.5% of women will undergo prolapse surgery in their lifetime (Abdel-Fattah 2011). A US-based study forecast that approximately 9.2 million women will be a ected by pelvic organ prolapse by 2050 (Wu 2009), and about 300,000 women undergo pelvic organ prolapse surgery in the USA every year costing over 1 billion US dollars (the cost year was unclear) (Enemchukwu 2019; Subak 2001). The NHS spent over £45 million on the treatment of prolapse between 2017 and 2018 (Primary Care Women's Health Forum 2019).
Vaginal pessaries are passive mechanical devices that are worn internally and designed to support the vaginal walls and apex. An extensive range of pessaries have been described for treating prolapse (Oliver 2011;Poma 2000). These consist mainly of PVC, latex or silicone pessaries, which are shaped devices inserted into and le inside the vagina to support the prolapsed pelvic organs. Two main groups of pessaries are used: support pessaries (e.g. ring pessary or ring pessary with support) and space filling pessaries (e.g. Gellhorn or cube) (Dwyer 2019; Oliver 2011). Support pessaries are inserted into the vagina and are positioned between the pubic bone and posterior vaginal fornix, providing support to descending organs. They do not prohibit vaginal intercourse and are easier to remove and replace. Space filling pessaries provide support by filling the vaginal space to prevent prolapse descent, creating a suction e ect around the pessary which thereby increases the likelihood of retention. They are not compatible with vaginal intercourse and are more di icult to remove and replace (Dwyer 2019). Studies suggest that ring pessaries are the most commonly used in practice (Bugge 2013a; Cundi 2000).
Recently published NICE guidance suggests considering pessary treatment (alone or with: pelvic floor muscle training (PFMT)) when women have symptomatic prolapse (NICE 2019). Surveys suggest that 87% to 98% of clinicians report using pessaries in their clinical practice (Bugge 2013a;Cundi 2000;Gorti 2009;Pott-Grinstein 2001), while 77% of gynaecologists report using pessaries as a firstline treatment for prolapse (Cundi 2000). Health professionals (87.6%) report that they are most likely to fit pessaries in women aged 60 or older and 98.8% would use a pessary for treatment when a woman is unfit for surgery (Bugge 2013a). Evidence from observational studies suggests that about 76% of women who try a pessary continue to use it for at least four weeks and, of those women who continue pessary use for more than four weeks, 86% continued to use the pessary for over five years (Lone 2011). Furthermore, observational studies have shown that pessaries may improve symptoms for women (Lamers 2011;Manchana 2012).

How the intervention might work
Pessaries are used in pelvic organ prolapse in order to physically support the vaginal walls and the pelvic organs behind them. The pessary is inserted into the vagina with a view to holding the prolapsed organs, supporting the pelvic structures, and relieving pressure on the bladder and bowel (Dumoulin 2017). Using a pessary in the management of pelvic organ prolapse aims to prevent the prolapse from becoming worse, as well as helping to decrease the frequency or severity of symptoms of prolapse, and averting or delaying the need for surgery (Oliver 2011). The majority of evidence for the use of pessaries comes from nonrandomised studies (e.g. Brazell 2014;Clemons 2004b;Clemons 2004a;Clemons 2004c;Hanson 2006;Kapoor 2009;Lamers 2011;Lone 2011;Manchana 2012). As such, the e ectiveness of pessaries for managing pelvic organ prolapse still needs to be clearly established. There is a lack of consensus as to the optimal treatment for women. Provided that su icient numbers of trials of adequate quality have been conducted, the most reliable evidence Library Trusted evidence. Informed decisions. Better health.

Why it is important to do this review
Cochrane Database of Systematic Reviews is likely to come from consideration of randomised controlled trials (RCTs), which is the basis for the current review. The aim is to help identify optimal practice and highlight where there is a need for further research.
Given that di erent costs are associated with the alternative interventions included in this review, it is important to conduct a brief economic commentary (BEC) to find available evidence (if any) on the costs of the interventions relative to their e ectiveness.

O B J E C T I V E S
To assess the e ects of pessaries (mechanical devices) for managing pelvic organ prolapse in women; and summarise the principal findings of relevant economic evaluations of this intervention.

Types of studies
We included randomised trials (RCTs) and quasi-randomised trials in which at least one arm was a pessary for pelvic organ prolapse. We excluded cross-over studies.

Types of participants
We included studies focusing on adult women who were seeking treatment for pelvic organ prolapse. Participants could have prolapse with any POP-Q staging (Bump 1996), and prolapse of any type: anterior vaginal wall prolapse (cystocoele, urethrocoele); posterior vaginal wall prolapse (enterocoele, rectocoele, perineal deficiency); and uterine or vaginal vault prolapse. We included trials where women were seeking treatment for urinary incontinence (UI) or prolapse if the data for women with prolapse (including women with prolapse and UI) were considered, or available separately.

Types of interventions
We included studies where one arm involved allocation to a pessary aiming to provide support for pelvic organ prolapse. Thus, we excluded studies where the pessary was used for exercise (and not support) or where pessaries were only used to deliver drug treatments (such as oestrogen). Pessaries could be of any type and made of any material. The types of devices included support pessaries (e.g. ring, ring with support) and space-filling devices (e.g. shelf, Gellhorn, cube).
We considered the following comparisons and planned to present 'Summary of findings' tables for the first three.
• Any pessary versus control, waiting list or no active treatment.
• Any pessary versus another treatment (lifestyle interventions, oestrogen treatment, physical interventions such as pelvic floor muscle training, surgery). • Any pessary plus another treatment versus the other treatment alone. • One pessary versus another pessary.

Primary outcomes
• Women's perceived resolution of prolapse symptoms

Timing of outcome assessment
We considered all outcomes at ≥ 12 months.

Main outcomes for 'Summary of findings' tables
These outcomes have been informed by a recent priority-setting exercise for pessary use (Lough 2018).
• Woman's perceived resolution of prolapse symptoms at ≥ 12 months • Woman's perceived improvement in prolapse symptoms assessed using a validated symptom questionnaire at ≥ 12 months • Prolapse-specific quality of life (e.g. PFIQ) at ≥ 12 months • Cure or improvement of sexual problems at ≥ 12 months • Psychological outcome measures at ≥ 12 months • Number of women with adverse events (vaginal bleeding, voiding di iculty)

Search methods for identification of studies
We did not impose any language or other restrictions on any of the searches detailed below.

Library
Trusted evidence. Informed decisions. Better health.

Search for clinical e ectiveness studies
We identified relevant trials from the Cochrane Incontinence Specialised Register. For more details of the search methods used to build the Specialised Register, please see the Group's webpages where details of the Register's development (from inception) and the most recent searches performed to populate the Register can be found. To summarise, the Register contains trials identified from the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, MEDLINE In-Process, MEDLINE Epub Ahead of Print, ClinicalTrials.gov, WHO ICTRP, Be Part of Research and handsearching of journals and conference proceedings. Many of the trials in the Cochrane Incontinence Specialised Register are also contained in CENTRAL.
The terms used to search the Cochrane Incontinence Group Specialised Register are given in Appendix 1. The date of the most recent search of the Register for this review: 28 January 2020.
The search methods and strategies used for some of the previous versions of this review are given in Appendix 2.

Search for economic evaluations
We performed additional searches for the brief economic commentary (BEC). We searched the following.
• The NHS Economic Evaluation Database (NHS EED) on the Centre for Reviews and Dissemination (CRD) website (covering from the earliest record in NHS EED, dating from 1968, up to and including 31 December 2014 when their coverage ended). Date of search: 13 February 2020.
As NHS EED is no longer actively updated, we performed additional searches of the following databases to identify eligible studies added to these databases from 1 January 2015 onwards (date of search: 13 February 2020): • MEDLINE on OvidSP (covering 1 January 1946 to January February 1 2020); and • Embase (on OvidSP) (covering 1 January 1974 to 2020 Week 06).
Details of the searches that were performed can be found in Appendix 3.

Searching other resources
We searched the reference lists of relevant articles for other possibly relevant trials. We also contacted the authors of included abstracts.

Data collection and analysis
We conducted data collection and analysis in accordance with methods specified in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019).

Selection of studies
Two review authors independently screened all potentially eligible abstracts from the search (CB and DG or MD). We nominated a third review author to arbitrate in the event of disagreement (EJA). We excluded any studies which were not relevant at this stage. We then obtained the full-text articles of relevant studies. If there was any uncertainty on the eligibility of the studies based on title and abstract, we obtained the full paper, which was reviewed by the same two review authors.

Data extraction and management
Two review authors independently undertook data extraction using a predefined data extraction form (from CB, RK and FS). We then made comparisons to ensure accuracy. We resolved any discrepancies by discussion with, or referral to, a third party (EJA).

Assessment of risk of bias in included studies
We assessed the risk of bias within studies as part of the data extraction. We used the methods outlined in the Cochrane Handbook to assess risk of bias in the following domains (Higgins 2011): random sequence generation (selection bias); allocation concealment (selection bias); blinding of participants and personnel (detection bias); blinding of outcome assessors (performance bias); incomplete outcome data (attrition bias); selective reporting (reporting bias); other bias not considered under other domains. Two review authors assessed the included studies for risk of bias and any discrepancies were resolved by discussion and referred to a third author where necessary.

Measures of treatment e ect
For binary outcomes, we have reported risk ratios (RR) and 95% confidence intervals (CI). For continuous outcomes, we reported mean di erences (MDs) with 95% CIs.

Unit of analysis issues
The primary analysis was per woman randomised. Had data allowed, we would have analysed each pair of arms in multi-arm studies as a separate comparison. To avoid double-counting where the same study with three treatment arms is included twice in the same meta-analysis, we planned to divide the number of women and events by two in the group that appears twice in the analysis.

Dealing with missing data
We contacted study authors to obtain missing data in order to be able to conduct intention-to-treat analysis wherever possible. Where randomised data were not available, we have reported data from per protocol analyses as presented in the study publications. We have clearly indicated in the E ects of interventions where we have used per protocol data.

Assessment of heterogeneity
We intended to combine studies in meta-analysis where clinical and methodological heterogeneity was low; that is, studies had to be broadly similar in terms of participant characteristics, interventions, study design and outcomes in order to be pooled. Had data allowed, we would have assessed statistical heterogeneity in meta-analyses as follows, taking into account the magnitude and direction of e ects as well as the P value and/or confidence interval for I : • 0% to 40%: heterogeneity might not be important • 30% to 60% might represent moderate heterogeneity • 50% to 90% might represent substantial heterogeneity • 75% to 100% represents considerable heterogeneity

Assessment of reporting biases
Using the data extraction form, we assessed for data that could have been collected but was not reported. Where there were 10 or more studies in the same meta-analysis and where data allowed, we would have investigated reporting biases (such as publication bias) using funnel plots. If funnel plot asymmetry was suggested by a visual assessment, we would have performed exploratory analyses to investigate it.

Data synthesis
Had data allowed, we would have used a fixed-e ect metaanalysis for combining data if it was reasonable to assume that the studies were estimating the same underlying treatment e ect (i.e. where trials were examining the same intervention, and the trials' populations and methods were judged su iciently similar). If there was clinical heterogeneity su icient to expect that the underlying treatment e ects di ered between trials, or if substantial statistical heterogeneity was detected, we would have used a random-e ects meta-analysis to produce an overall summary.

Subgroup analysis and investigation of heterogeneity
Had data allowed, we planned to carry out subgroup analysis according to di erent prolapse stages or compartments.

Sensitivity analysis
Had data allowed, we would have conducted sensitivity analyses to examine the e ect of studies with high risk of bias on the results. We would have repeated the main meta-analyses without studies judged to be at high risk and compare those results to the main analyses with all studies included.

Incorporating economics evidence
Once the search outlined in the Search methods for identification of studies was performed, we developed a brief economic commentary (BEC) to summarise the availability and principal findings of the full economic evaluations that compared pessaries with alternative treatment interventions for pelvic organ prolapse in women (Shemilt 2019). This BEC encompassed full economic evaluations (i.e. cost-e ectiveness analyses, cost-utility analyses and cost-benefit analyses), conducted alongside or based on one or more studies included in the main review of intervention e ects, or model-based economic evaluations.

Summary of findings and assessment of the certainty of the evidence
We prepared 'Summary of findings' tables using the GRADEpro GDT so ware for the comparisons pre-stated in the Types of interventions.
We used the GRADE approach to assess the certainty of evidence related to the outcomes listed in the 'Main outcomes for 'Summary of findings' tables' in the Types of outcome measures (Schünemann 2019). We used the five GRADE considerations (study limitations, consistency of e ect, imprecision, indirectness and publication bias) to assess the certainty of the body of evidence for the prespecified outcomes. We justified all decisions to downgrade the certainty of studies using footnotes.

Library
Trusted evidence. Informed decisions. Better health.

Cochrane Database of Systematic Reviews
This is the second review update. In the first version of the review, no relevant randomised controlled trials (RCTs) were identified (Adams 2004). In the first update (Bugge 2013b), we included one cross-over study (Cundi 2007), which we have now excluded because the cross-over study design means that we cannot tell what the true e ects of the pessary are.
Searches for economic evaluations to inform the development of the brief economic commentary (BEC) produced 479 unique records a er removal of duplicates. We retrieved the full-text of seven articles that appeared to meet our eligibility criteria but found that three of these did not include the correct population and were therefore excluded; the remaining four reports of two economic evaluations were included in this BEC. The flow of literature through the assessment process is shown in Figure 2.

Library
Trusted evidence. Informed decisions. Better health.

Included studies
Four trials met the inclusion criteria. Full details are included in the Characteristics of included studies.

Design
All four included trials were RCTs. In one trial only six women were randomised before patient preference took priority and data were not reported per randomised group (Coolen 2018). The data were therefore not usable in the review.

Participants
The trials included 478 women and the mean age ranged from 30.4 (Baessler 2019) to 65.6 years (Panman 2016). The trials included women with: stage I to III POP-Q symptomatic prolapse (Cheung 2016); stage II or greater symptomatic prolapse (Coolen 2018; Baessler 2019); and prolapse at or beyond the hymen (Panman 2016). One trial included women who were in the immediate postpartum period (Baessler 2019).

Length of follow-up
Follow-up in included trials ranged from six weeks (Baessler 2019) to two years (Panman 2016).

Outcomes
None of the included trials reported the primary outcome of resolution of prolapse symptoms. Three trials reported the primary outcome of perceived improvement of prolapse symptoms (Baessler 2019; Cheung 2016; Panman 2016).
The secondary outcomes reported in the trials were as follows. None of the included studies reported data on patient-reported satisfaction with pessary treatment or psychological outcome measures.

Excluded studies
Eleven studies were excluded (see the Characteristics of excluded studies). Of these, two were excluded because it was not possible to separate the data for the women with prolapse from the data for women with urinary incontinence (Meriwether 2015; Taege 2017). A further seven trials were excluded because they investigated pessary plus another treatment compared with pessary alone (Chou 2013

Risk of bias in included studies
See Figure 3 and Figure 4 for visual representations of the risk of bias in the included trials. Risk of bias for each included study is further detailed in the Characteristics of included studies.

Library
Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews Cochrane Database of Systematic Reviews

Figure 4. 'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias): All outcomes Blinding of outcome assessment (detection bias): All outcomes Incomplete outcome data (attrition bias): All outcomes Selective reporting (reporting bias) Other bias 0% 25% 50% 75% 100% Low risk of bias Unclear risk of bias High risk of bias

Random sequence generation
Three trials had a low risk of bias for random sequence generation as they utilised remote computerised systems for randomisation (Baessler 2019; Cheung 2016; Panman 2016). Coolen 2018 had a high risk of bias as only 6/113 women in the study were randomised. For the remaining women in this study, the patients' treatment preference took priority.

Allocation concealment
All four trials had a low risk of bias for allocation concealment, as they used remote systems and in blocks of variable size, hence blinding the researchers to the possible treatment allocation (Baessler 2019; Cheung 2016; Coolen 2018; Panman 2016).

Blinding of participants and personnel
All trials were unable to blind participants given the nature of the interventions in use (pessaries, PFMT and surgery) (Baessler 2019; Cheung 2016; Coolen 2018; Panman 2016). We therefore judged them to be at high risk of performance bias since knowledge of their assignment could influence their subjective assessment of their prolapse symptoms.

Blinding of outcome assessment
We assessed two trials as low risk of detection bias (Baessler 2019; Cheung 2016) because assessors were blinded to group allocation. One trial was judged to be at an unclear risk of bias for this domain (Coolen 2018). One trial was judged at high risk of detection bias because the outcome assessors were not blinded to group allocation (Panman 2016). In all cases where it was reported, attempts were made to blind the assessors to previous measurements.

Incomplete outcome data
One trial was assessed as high risk of attrition bias because of the substantially higher withdrawal rate in the control group (Baessler 2019). Two trials were assessed to be at a low risk of attrition bias because there was no di erential attrition and the trialists conducted intention-to-treat analysis (Cheung 2016; Panman 2016). In one trial, the risk of attrition bias was unclear due to a lack of information (Coolen 2018).

Selective reporting
Three trials were assessed as having a low risk of reporting bias as all outcomes were reported across all the published material (Baessler 2019; Cheung 2016 Panman 2016). The fourth trial did not provide su icient information to judge the risk of reporting bias and so was judged to be at unclear risk of bias (Coolen 2018).

Other potential sources of bias
There were no indications of any other kind of bias in the included trials, so all four were judged to be at low risk of bias (Baessler 2019; Cheung 2016; Coolen 2018; Panman 2016).

E ects of interventions
See: Summary of findings 1 Any pessary compared to control, waiting list or no active treatment; Summary of findings 2 Any pessary compared to pelvic floor muscle training (PFMT) for pelvic organ prolapse in women; Summary of findings 3 Pessary plus PFMT compared to pelvic floor muscle training (PFMT) alone for pelvic organ prolapse in women

Any pessary versus control, waiting list or no active treatment
One trial compared a ring pessary versus standard care (defined as no treatment for six weeks postpartum) for women who had stage II prolapse one to three days postpartum (Baessler 2019). In this trial, 13 women were randomised to pessary care and 21 to standard care.

Women's perceived resolution of prolapse symptoms
Not reported.

Women's perceived improvement in prolapse symptoms
Baessler 2019 measured improvement in prolapse symptoms using the Australian Pelvic Floor Questionnaire (Baessler 2009). We are uncertain if pessaries have any e ect on women's perceived improvement in prolapse symptoms compared to standard care (mean di erence (MD) -0.03, 95% confidence interval (CI) -0.61 to 0.55; 1 study; 27 women; very low-certainty evidence; Summary of findings 1; Analysis 1.1).

Library
Trusted evidence. Informed decisions. Better health.

Patient-reported satisfaction with pessary treatment
Not reported.

Grade of prolapse
Not reported.

Site-specific grading of prolapse
Baessler 2019 measured site-specific grading of prolapse using the POP-Q. A er the pessary has been removed for a period of time, we are uncertain if it has any e ect on the anatomical measurement of prolapse compared with standard care (anterior compartment: MD 0.10, 95% CI -0.73 to 0.93; 1 study, 27 women; Analysis 1.2; posterior compartment (0.10, 95% CI -0.49 to 0.69; 1 study, 27 women; Analysis 1.3).

Prolapse-specific quality of life
Not reported.

Quality of life measured with generic quality of life or health status measures
Not reported.

Psychological outcome measures
Not reported.

Cure or improvement of bladder problems
Baessler 2019 measured bladder problems using the bladder score of the Australian Pelvic Floor Questionnaire. We are uncertain if pessaries have any e ect on bladder problems compared with standard care (MD -0.15, 95% CI -0.55 to 0.25; 1 study, 27 women; Analysis 1.4).

Cure or improvement of bowel problems
Baessler 2019 measured bladder problems using the bowel score of the Australian Pelvic Floor Questionnaire. We are uncertain if pessaries have any e ect on bowel problems compared with standard care (MD 0.00, 95% CI -0.60 to 0.60;1 study, 27 women; Analysis 1.5).

Cure or improvement of sexual problems
Baessler 2019 measured bladder problems using the sex score of the Australian Pelvic Floor Questionnaire. We are uncertain if pessaries have any e ect on curing or improving sexual problems compared with standard care (MD -0.29, 95% CI -1.67 to 1.09; 1 study; 27 women; very low-certainty evidence; Summary of findings 1; Analysis 1.6).

Number of women with adverse events (vaginal bleeding, voiding di iculty)
The adverse events of vaginal bleeding and voiding di iculty were not reported.

Any pessary versus another treatment
One trial compared a pessary with pelvic floor muscle training (PFMT) (Panman 2016). In this trial, 82 women received a pessary and 80 women received PFMT. Another trial compared a pessary with surgery (Coolen 2018). However, of the 113 women recruited to the trial only six consented to the randomisation process (two were randomised to pessary and four to surgery). The data were presented for the whole study population as a prospective cohort and therefore were not usable for this review.

Women's perceived resolution of prolapse symptoms
Not reported.

Women's perceived improvement in prolapse symptoms
Panman 2016 measured improvement in prolapse symptoms using the Pelvic Floor Disability Index (PFDI-20) (Barber 2005), and also the subscale of the PFDI measure on specific prolapse symptoms (Pelvic Organ Prolapse Distress Inventory 6 (POPDI-6)). Panman 2016 reported the number of women self-reporting 'better' or 'much better' symptoms. At 24 months, we are uncertain if pessaries have any e ect on women's perceived improvement in prolapse symptoms compared with PFMT because the certainty of evidence is low and the 95% CI spans possible harm and possible benefit (risk ratio (RR) 1.01, 95% CI 0.64 to 1.59; 1 study; 142 women; Analysis 2.5).

Patient-reported satisfaction with pessary treatment
Not reported.

Grade of prolapse
Not reported.

Site-specific grading of prolapse
Not reported.

Number of women with adverse events (vaginal bleeding, voiding di iculty)
Data on complications in one study were only available as part of the per protocol analysis (Panman 2016). Pessaries may substantially increase the risk of adverse events compared with PFMT (RR 75.25, 95% CI 4.70 to 1205.45; 1 study; 97 women; lowcertainty evidence; Summary of findings 2; Analysis 2.18). There were no reported complications for the PFMT group. Of the 35 women who persisted with a pessary until 24 months, 60% (n = 21) reported pessary-related side e ects. These were: increased vaginal discharge (n = 14); increased urinary incontinence (n = 5); and irritation or erosions of the vaginal wall (n = 10).
Panman 2016 also reported reasons for pessary device removal. For 34 women, the initial pessary fit was unsuccessful, with a further seven women then also experiencing unsuccessful fit. Twelve women discontinued pessary treatment for the following reasons: increased incontinence (n = 4); vaginal erosion (n = 2); pessary treatment too burdensome (n = 2); no e ect of pessary treatment (n = 1); discomfort during intercourse (n = 1); and other reasons (n = 3).

Any pessary plus another treatment versus the other treatment alone
One trial compared pelvic floor muscle training (PFMT) plus a pessary with PFMT alone (Cheung 2016). A total of 276 women were randomised in this trial: 137 received PFMT treatment and 139 received PFMT treatment plus a vaginal pessary.

Women's perceived resolution of prolapse symptoms
Not reported.

Women's perceived improvement in prolapse symptoms
At 12 months, pessary plus PFMT probably leads to more women perceiving improvement in their prolapse symptoms compared with PFMT alone (RR 2.15, 95% CI 1.58 to 2.94; 1 study; 260 women; moderate-certainty evidence; Summary of findings 3; Analysis 3.1). With pessary plus PFMT, 32.3% more women will probably feel that their symptoms have improved compared with pessary alone (16.3% more to 54.6% more).
Cheung 2016 also measured perceived improvement in prolapse symptoms using the PFDI questionnaire (and the POPDI subscale -prolapse-specific symptoms; Barber 2001). In the intention-totreat ( ITT) analysis, there was a di erence at 12 months for the POPDI subscale of the PFDI for women undertaking PFMT plus pessary (median score 32.1, interquartile range (IQR) 12.5 to 78.6; 139 women) compared to those in the PFMT-only group (median score 49.4, IQR 21.4 to 95.2; 137 women; P = 0.04), which further suggests that a pessary may improve prolapse symptoms when added to PFMT.

Patient-reported satisfaction with pessary treatment
Not reported.

Grade of prolapse
Not reported.

Site-specific grade of prolapse
Not reported.

Prolapse-specific quality of life
In Cheung 2016, at 12 months the women's prolapse-specific quality of life was measured using the POPIQ subscale of the PFIQ (higher score = worse quality of life; Barber 2001). On POPIQ score, pessary plus PFMT probably improves women's prolapse-specific quality of life (median 0.3, IQR 0 to 22.2; 132 women, in comparison to the PFMT-only group (median 8.9, IQR 0 to 64.9, 128 women; P = 0.02; moderate-certainty evidence; Summary of findings 3). In the 12-month ITT analysis, Cheung 2016 also measured urinary symptoms using the UDI and associated quality of life using the Urinary Impact Questionnaire (UIQ).

Cure or improvement of bowel problems
In the 12-month ITT analysis, Cheung 2016 measured bowel symptoms using the CRADI and associated quality of life using the CRAIQ.

Cure or improvement of sexual problems
Not reported.

Number of women with adverse events (vaginal bleeding, voiding di iculty)
Pessary plus PFMT may slightly increase the risk of abnormal vaginal bleeding compared with PFMT alone (RR 2.18, 95% CI 0.69 to 6.91; 1 study; 260 women; low-certainty evidence; Summary of findings 3; Analysis 3.5). Pessary plus PFMT may lead to 3.7% more women having abnormal vaginal bleeding compared with PFMT alone (1% fewer to 18% more).
We are less certain if pessary plus PFMT has any e ect on the risk of de novo voiding di iculty because the 95% CIs are wide and span possible harm and possible benefit (RR 1.32, 95% CI 0.54 to 3.19; 1 study; 189 women; low-certainty evidence; Analysis 3.6; Summary of findings 3).

One pessary versus another pessary
No trials were identified for this comparison.

Summary of main results
Four trials involving a total of 478 women were included in this updated review. None of the included trials reported our primary outcome of women's perception of cure of prolapse symptoms. Due to the varied nature of the comparisons, it was not possible to pool data.
The evidence is very uncertain about the e ect of pessaries compared with standard care on perceived improvement of prolapse symptoms, prolapse-specific quality of life, and cure or improvement of sexual problems (Summary of findings 1). We did not find any evidence about the risk of adverse events compared with standard care (no treatment).
We are uncertain whether pessaries have any e ect on women's perceived improvement in prolapse symptoms compared with pelvic floor muscle training (PFMT). We are also uncertain whether pessaries have any e ect on prolapse-specific quality of life measured with the Pelvic Floor Impact Questionnaire-7 (PFIQ-7), or on cure or improvement of sexual problems measured with the Pelvic Organ Prolapse/Urinary Incontinence Sexual Questionnaire (PISQ-12) compared with PFMT (low-certainty evidence). However, pessaries may result in a large increase in risk of adverse events compared with PFMT (low-certainty evidence). Adverse events included increased vaginal discharge, increased urinary incontinence and erosion or irritation of the vaginal walls (Summary of findings 2).
Pessary plus PFMT probably leads to more women perceiving improvement in their prolapse symptoms and in their prolapsespecific quality of life compared with PFMT alone (moderatecertainty evidence). However, pessary plus PFMT may slightly increase the risk of abnormal vaginal bleeding compared with PFMT alone (low-certainty evidence). We are less certain if pessary plus PFMT has any e ect on the risk of de novo voiding dysfunction because the 95% CIs are wide and span possible harm and possible benefit (Summary of findings 3).

Overall completeness and applicability of evidence
We identified no studies comparing di erent types of pessaries.
The studies included here are small and do not report several of our outcomes of interest. In particular, we identified no evidence regarding the e ect of pessaries on women's perceived resolution of prolapse symptoms. However, the women included in the studies are largely representative of the wider population of women with pelvic organ prolapse in that they represent women with prolapse of di erent types, stages and across a range of ages.
Although the evidence base about vaginal pessary use for pelvic organ prolapse in women is growing, the available trials o er limited answers to the primary question regarding the e ectiveness of pessaries in resolving or improving women's prolapse symptoms. In addition, we do not have adequate evidence to be certain regarding the e ect of pessaries on quality of life, or their complications and adverse e ects.

Quality of the evidence
Overall, we judged the studies to have a low risk of bias in most domains. Given that it was not possible to blind participants to their treatment allocation in any of the studies, and given that our primary outcome is subjectively reported, we judged the studies to have a high risk of performance bias. Due to imbalanced attrition and missing outcome data, we also judged some studies to have a high risk of attrition bias.
With regard to the outcomes presented in the 'Summary of findings' tables, we made decisions to downgrade the certainty of evidence because of the extent of risk of bias. Additionally, we downgraded all outcomes for imprecision due to the small sample sizes in the studies.

Library
Trusted evidence. Informed decisions. Better health.

Potential biases in the review process
We made every attempt to reduce bias in the review as far as possible, with broad inclusion criteria and a comprehensive search strategy not limited by language or publication type. We also ensured data extraction and 'Risk of bias' assessments were carried out independently by two review authors.

Brief economic commentary
In addition to the main review, we searched for economic evaluations that compared pessaries with either surgical procedures or conservative management, or both. From searches conducted in February 2020, we identified two economic evaluation studies (Hullfish 2011 and Panman 2016). We did not subject these identified economic evaluations to critical appraisal and we do not attempt to draw any firm or general conclusions regarding the relative costs or e iciency of pessaries in treating pelvic organ prolapse.
Hullfish 2011 conducted a model-based cost-utility analysis (CUA) comparing expectant management, placement of a pessary and surgical procedures (vaginal reconstructive surgery (VRS), traditional/open abdominal sacrocolpopexy, and robotic-assisted abdominal sacrocolpopexy) for post-hysterectomy pelvic organ prolapse. The study was conducted in the USA. A Markov model was developed to link potential health states and probabilistic events that follow from each treatment strategy. The model inputs include event probabilities, costs and utilities for outcomes at "a baseline case of a healthy 65 year old, post-hysterectomy female with ≥ stage III apical prolapse of the vagina, who wishes to preserve coital function". Event probabilities were mainly obtained from a review of published literature but expert knowledge was used to derive missing estimates. Due to the unavailability of health-related quality of life utility scores for pelvic organ prolapse, the authors used relative utility values that were developed based on clinical researchers' opinions. The model structure allows individuals to exist in one of seven health states during any given month over a 12-month period. These health states include pelvic organ prolapse with no complications, with presenting complications (voiding dysfunction); pessary with no complications, pessary with complication (vaginal erosion); repaired pelvic organ prolapse without late/postoperative complications (urinary tract infection); and repaired pelvic organ prolapse with major late complications (i.e. reoperation). Only direct medical costs of each treatment alternative were estimated from a number of sources for the analysis. The currency and price year are US $ and 2007 respectively. At baseline, only pessary and VRS were found to be cost-e ective: pessary use yielded about 10.4 quality-adjusted months at a cost of $10,000 per patient and the VRS alternative obtained 11.4 quality-adjusted months at a cost of approximately $15,000 per patient.
In a deterministic sensitivity analysis, expectant management with VRS was e icient when baseline probability of pelvic organ prolapse complications was decreased. Pessary alternative was no longer e icient a er the probability of pelvic organ prolapse complications was reduced to 0.12 and the probability of surgery following pessary use was increased to 0.17. Monte Carlo simulation results indicated that pessary use was an optimal alternative below $5600 'Willingness to Pay' (WTP) per qualityadjusted life month threshold and VRS was the optimal strategy above that threshold. The authors concluded that their model demonstrated that pessary use and VRS are both cost-e ective alternatives for treating hysterectomy vaginal prolapse given the context of their study.
Panman 2016 conducted both a CUA and a cost-e ectiveness analysis (CEA) to compare pessary treatment with pelvic floor muscle training (PFMT) using RCT data (three follow-ups within two years) from 20 primary care practices in the Netherlands (Wiegersma 2014). Participants included all women aged at least 55 years registered with their primary care practice. The outcome used in the CEA was distress measured by Pelvic Floor Distress Inventory-20 (PFDI-20; Barber 2005), while quality-adjusted lifeyears (QALYs) were measured in the CUA based on EQ-5D utility values using the UK population tari (Dolan 1995). Only direct medical costs were included in the analysis, with a time horizon of two years. The main costs of the intervention included pessaries and pessary-related visits, costs of physical therapy, consultations with GPs and medical specialists, use of absorbent pads, medication and costs of operative procedures. Both costs and health benefits were recorded and estimated on an individual participant basis, and then the mean di erences of e ects and costs between the two treatment groups were estimated. The costs were valued in Euros at 2014 price level and later converted to US dollars based on the November 24 2015 exchange rate (1 Euro = 1.0657 dollars). Results of the analysis were summarised in incremental cost-e ectiveness ratio (ICER) and incremental cost-utility ratio (ICUR) for the CEA and CUA respectively. However, the focus is on the results of the cost-e ectiveness planes which, as indicated below, show the likelihood that an intervention was less costly but more e ective.
The costs per person were $309 in the pessary group and $437 in the PFMT group. The mean di erence was $128 (95% CI 27 to 236), with costs dependent on the primary treatment. Costs for pessaries and pessary-related visits were higher in the pessary group ($202 per person compared with $0.5 in the PFMT group), whereas costs for physical therapy were higher in the PFMT group ($324 per person for the PFMT group compared with $2 in the pessary group).
In the CUA, both treatment groups lost QALYs over the study period, with slightly less lost in the pessary group (0.024 in the pessary group and 0.065 in the PFMT group). In total, 71% of the 5000 replications in the bootstrap simulation indicated better outcomes and lower costs for the pessary group.
The authors stated that there was no di erence in change in pelvic floor symptoms between pessary treatment and PFMT in women aged 55 years or older recruited by screening. Although pessary treatment led to a statistically significant greater improvement in typical prolapse symptoms compared with PFMT, the di erence was small and the clinical relevance doubtful.
The authors reported that their study was the first to investigate the cost-e ectiveness of pessary treatment and PFMT from such a trial and that there were no significant di erences between pessary treatment and PFMT in the reduction of pelvic floor symptoms but women with apex prolapse symptoms benefited more from pessary treatment than from PFMT. Pessary treatment was preferable to PFMT in the cost-e ectiveness analysis, with PFMT being more expensive but no more e ective.
The interventions compared in one of the economic evaluation studies match interventions compared in Summary of findings Cochrane Database of Systematic Reviews 2 (Panman 2016). It would therefore be useful to compare the findings of the economic evaluation with data in Summary of findings 2.
End users of this review will need to assess the extent to which results of identified economic evaluations may be applicable (or transferable) to their own setting.

Agreements and disagreements with other studies or reviews
The previous version of this review (Bugge 2013b), the most recent review for the International Continence Society monograph (Dumoulin 2017), and the current review have all reached similar conclusions. All conclude that there is insu icient evidence on which to guide clinical practice surrounding prolapse management using vaginal pessaries.

Implications for practice
The evidence from randomised controlled trials (RCTs) on which to base the treatment of women with pelvic organ prolapse using vaginal pessaries remains very limited. The research base is growing but more trials are needed in order for data to be pooled in a way that will allow for greater certainty in recommendations for treatment. Clinical sta may wish to read this review alongside the surgical Cochrane Reviews ( In the postpartum period, one small trial compared pessary versus standard care (no treatment for six weeks), and across the available comparisons the certainty of evidence was very low, making recommendations for practice in the postpartum period di icult. Panman 2016 compared pessary versus pelvic floor muscle training (PFMT), and it was uncertain if pessary had an e ect on women's perceived improvement in prolapse symptoms in comparison to PFMT. Both PFMT and pessary are active treatments and, as such, it is not known if improvements were greater than if no treatment had been given. Cheung 2016 added a pessary to PFMT and probably led to more women perceiving an improvement in prolapse symptoms. It may be that practitioners consider the added benefit of a pessary for women with prolapse who are receiving PFMT.
There was some evidence of an increase in adverse events with pessary use. However, the clinical significance of these adverse events is unknown but may be argued to be low. Specifically, increased vaginal bleeding and discharge may be adverse events a woman is willing to accept if she experiences a decrease in prolapse symptoms. De novo stress urinary incontinence (SUI) may be a more problematic adverse event for women, but the evidence surrounding the e ects of pessary on de novo SUI are not certain. While some evidence is suggestive of an increase in adverse events for women who use a pessary for prolapse, the extent to which these events are bothersome for women requires further investigation before recommendations can be made for practice.

Implications for research
Adequately-powered trials measuring clinically-important outcomes are urgently needed to guide international practice. The comparison between pessary and no active treatment is perhaps di icult as there is observational evidence suggesting benefit of pessary and other conservative treatments, making a no active treatment group possibly unacceptable to women or policy makers (Hagen 2011a; Lamers 2011).
Across the other comparisons more research is needed, specifically: pessary versus another active treatment; pessary as an adjunct to other treatments; and comparison between pessary types. It is important that future trials measure women's perceived resolution of prolapse symptoms. Qualitative research about the bothersomeness of adverse events of pessary use would also support the understanding of adverse event reporting in the trials. As the evidence base evolves, it will also be helpful to explore the e ectiveness of pessary for di erent stages and types of prolapse.
This review identified only two relevant economic evaluations, which highlights both the lack of economic evidence and the need to conduct more economic evaluations on prolapse interventions that includes pessaries.

A C K N O W L E D G E M E N T S
We are grateful to Chantale Dumoulin and Durhane Wong-Rieger for their valuable comments on dra s of the review, along with a health economist who provided feedback on the brief economic commentary. We thank Sheila Wallace for assistance with searches for this review and Muhammad Imran Omar, Lindsey Elstub and Eugenie Johnson for advice on undertaking and reporting the review.
We acknowledge the contributions of Suzanne Hagen, who was an author on the original review, and Fiona Reid, who was an author on the first revision of the review.

Abdel-Fattah 2011
Abdel-Fattah M, Familusi A, Fielding S, Ford J, Bhattacharya S. Primary and repeat surgical treatment for female pelvic organ prolapse and incontinence in parous women in the UK: a register linkage study. BMJ Open 2011;1:e000206.

Baessler 2009
Baessler Trial stopped early due to poor recruitment and 'occasional pessary problems which were solved with pessary removal'

Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk PC-generated randomisation list put into opaque envelopes. Preference was permitted but data are available for randomised group only Allocation concealment (selection bias)

Low risk Opaque envelopes opened after consent and examination for levator avulsion
Blinding of participants and personnel (performance bias) All outcomes High risk Not possible to blind participant or caregiver.
Blinding of outcome assessment (detection bias) All outcomes Low risk Assessors blinded to group allocation.
Incomplete outcome data (attrition bias) All outcomes High risk Some differential loss to follow-up (1/13 in pessary group, 6/21 in standard care group).
Selective reporting (reporting bias) Low risk Data provided on measured outcomes.

Other bias Low risk None
Baessler 2019 (Continued) Study characteristics   The reference lists of relevant articles were searched for other possibly relevant trials.
We did not impose any language or other restrictions on any of the searches.

Library
Trusted evidence. Informed decisions. Better health.