POPPY: Individualized Pelvic Floor Muscle Training for Pelvic Organ Prolapse
POPPY was a multicenter, parallel-group randomized trial in 447 women with newly diagnosed, symptomatic stage I to III prolapse that compared five one-to-one individualized pelvic floor muscle training (PFMT) appointments over 16 weeks with a lifestyle advice leaflet alone, with the Pelvic Organ Prolapse Symptom Score (POP-SS) at 12 months as the primary endpoint. POP-SS fell by 3.77 points with PFMT and 2.09 points with the leaflet (adjusted difference 1.52, 95% CI 0.46 to 2.59; P = 0.0053).[1]
- Design and population
- Multicenter, parallel-group, assessor-masked randomized superiority trial. 447 treatment-naive women with newly diagnosed, symptomatic stage I to III prolapse of any compartment, recruited at 25 gynecology outpatient centers in the UK, New Zealand and Australia.
- Arms
- Individualized pelvic floor muscle training (five one-to-one physiotherapy appointments over 16 weeks) plus lifestyle advice leaflet versus lifestyle advice leaflet alone
- Report and follow-up
- 2013 · 12 months from trial entry (blinded POP-Q examination at 6 months)
| Endpoint | Results | Effect | Finding |
|---|---|---|---|
| Mean reduction in POP-SS from baseline at 12 months (0-28 scale); adjusted between-group difference in change | 3.77 (SD 5.62) vs 2.09 (SD 5.39) | difference 1.52; CI 0.46–2.59; P 0.0053 | Met |
| Any further prolapse treatment reported by 12 months (surgery, pessary, physiotherapy referral, estrogen or other) | 24.1% (35/145) vs 49.6% (71/143) | difference 25.5; CI 14.5–36; P <0.0001 | Met |
| Less severe POP-Q stage at 6 months (ordinal odds ratio, blinded examination); improvement by one or more stages | 26.8% (45/168) vs 19.3% (33/171) | OR 1.47; CI 0.97–2.27; P 0.07 | Not met |
| Prolapse surgery reported by 12 months | 11.0% (16/145) vs 9.8% (14/143) | P 0.84 | Descriptive |
Harms: Adverse events reported by participants: 9 events in the intervention group (8 adverse events, 1 unexpected serious adverse event) vs None reported; Discontinued intervention because PFMT was thought to worsen symptoms: 2 of 225 vs Not applicable
Design
Women were new attendees at gynecology outpatient clinics in 25 centers (23 in the UK, one in New Zealand and one in Australia). A gynecologist had to confirm stage I, II or III prolapse of any compartment on the Pelvic Organ Prolapse Quantification (POP-Q) examination, and prolapse had to be the main presenting complaint. Women with previous prolapse treatment, including surgery, were excluded, as were women who were pregnant, less than six months postnatal or unable to attend appointments. Women who needed treatment for vaginal atrophy were eligible after a course of local estrogen.[1]
Allocation was by remote computer with minimization on centre, prolapse stage and whether the woman was considering prolapse surgery. Participants and treating physiotherapists could not be masked. Outcomes were participant-completed postal questionnaires at 6 and 12 months, and the gynecologist who performed the 6-month POP-Q examination was masked to allocation. Centers received POP-Q training with the American Urogynecologic Society training DVD. The stage analyzed in the trial was calculated afterward from the nine recorded POP-Q measurements and on occasion differed from the clinic-assigned stage that determined eligibility.[1]
The primary endpoint was the change in POP-SS from baseline to 12 months. POP-SS has seven items scored 0 (never) to 4 (all of the time) for symptom frequency over the previous four weeks (total 0 to 28). The planned sample was 506 women (253 per group), giving 80% power to detect a 2.5-point difference assuming a standard deviation of 8; the trial randomized 88% of that target. Analysis was by intention to treat with a linear mixed-effects model adjusted for baseline POP-SS and the minimization variables, with multiple imputation as a sensitivity analysis.[1]
Results
Most women (75.6%) had stage II prolapse, 243 of them at or below the hymen, and the most common presentation was combined anterior, posterior and upper-compartment prolapse (45.4%). At 12 months, 295 of 447 women (66%) had returned questionnaires. POP-SS fell by 3.77 points (SD 5.62) with PFMT and 2.09 (SD 5.39) with the leaflet, an adjusted difference of 1.52 points (95% CI 0.46 to 2.59; P = 0.0053). Multiple imputation gave 1.66 (95% CI 0.74 to 2.58). The difference was larger at 6 months (2.84 points, 95% CI 2.05 to 3.63; P < 0.0001). The authors judged the 12-month difference to exceed the minimally important change for the POP-SS. The effect did not differ by prolapse stage, prolapse type, age or motivation for surgery in prespecified subgroup analyses.[1]
At 12 months, 57.2% (83/145) of the PFMT group and 44.7% (63/141) of the control group reported that their prolapse was better (difference 12.6 percentage points, 95% CI 1.1 to 24.1). At 6 months the proportions were 52% (98/187) and 17% (32/189). By 12 months, 24.1% (35/145) of the PFMT group and 49.6% (71/143) of the control group had received further prolapse treatment (risk difference 25.5 percentage points, 95% CI 14.5 to 36.0). The excess was physiotherapy referral (2/145 versus 38/143), and the groups did not differ in prolapse surgery (11.0% versus 9.8%).[1]
On the blinded 6-month examination, the odds of a less severe POP-Q stage were higher with PFMT but not significantly so (OR 1.47, 95% CI 0.97 to 2.27; P = 0.07); 45 of 168 women (26.8%) in the PFMT group and 33 of 171 (19.3%) in the control group improved by one or more stages.[1]
A record-linkage follow-up included 293 of the 310 POPPY participants resident in Scotland (65.5% of all 447 randomized) and followed hospital records for up to 10 years and 9 months. Any hospital-based treatment for a pelvic floor disorder occurred in 43.6% (65/149) of the PFMT group and 52.8% (76/144) of the control group (adjusted OR 0.61, 95% CI 0.37 to 0.99; P = 0.047), and among the 264 women in the Cox analysis, time to first treatment was longer with PFMT (HR 0.65, 95% CI 0.46 to 0.94). The authors found no clear difference in hospital resource-use costs between groups.[2]
Surgery and treatment delivery
Intervention women attended five one-to-one appointments with a women's health physiotherapist at weeks 0, 2, 6, 11 and 16. The first appointment included an internal pelvic floor muscle assessment (PERFECT scheme) and an individualized home program aiming for ten 10-second maximal holds and up to 50 fast contractions three times daily, recorded in a diary, plus pre-contraction of the pelvic floor against rises in abdominal pressure ("the knack"). The program was adjusted at each visit from examination findings and diary entries. Electromyographic or pressure biofeedback and electrical stimulation were not permitted.[1]
The first appointment took 80 minutes (60 minutes face to face plus 20 minutes of administration) and later appointments 40 minutes. Of the intervention group, 80% (178/222) attended four or five appointments, 70% (156/222) attended all five and 4.5% (10/222) attended none. At 12 months, 78% (115/147) of the PFMT group and 69% (95/138) of the control group reported pelvic floor exercises in the previous four weeks (P = 0.07). Both groups received a leaflet on weight loss, constipation, heavy lifting, coughing and high-impact exercise; it contained no pelvic floor exercise information.[1]
Pessary use by 12 months was 5.5% (8/145) with PFMT and 11.2% (16/143) with the leaflet (P = 0.13). The net cost of the intervention was GBP 131.61 per woman, and GBP 86.59 per one-point POP-SS improvement, in a within-trial NHS analysis at 2009 prices.[1]
Adverse events
Eight adverse events (six vaginal symptoms, one back pain and one abdominal pain) and one unexpected serious adverse event (a skiing injury) were reported, all in the intervention group, and none was judged related to the intervention or to participation. Two of the 30 women in the PFMT group who discontinued the intervention did so because they thought training made their symptoms worse.[1] All-cause mortality during the record-linkage follow-up was 5.4% with PFMT and 4.9% with the leaflet.[2]
Limitations
Participants and physiotherapists knew the allocation, and the primary endpoint was self-reported. The comparator was a leaflet without clinical contact, so physiotherapist attention cannot be separated from the exercise content. Questionnaire return at 12 months was 66%, and 168 of 225 and 171 of 222 women had a 6-month POP-Q examination; the authors state that the trial was not powered for the POP-Q endpoint. A substantial share of the control group obtained physiotherapy referral between 6 and 12 months, which the authors suggest makes the intention-to-treat estimate an underestimate of the training effect. Follow-up for the primary endpoint was 12 months, and all participants were treatment-naive, so the trial does not address training after surgery or alongside a pessary.[1]
The record-linkage report covered only Scottish participants, relied on procedure codes in hospital records, and counted continence procedures and pessary visits among its outcome codes. The authors report that the effect was evident only after adjustment for age, and the upper confidence bound for any treatment was 0.99.[2]
In conversation
POPPY measured success by symptoms with the POP-SS. The authors contrast it with a single-center trial that randomized 109 women to a more intensive program (weekly appointments for three months, then every two weeks for three months) and also reported prolapse stage, in which 19% of the training group and 8% of controls improved.[1] POPPY's stage improvement was 26.8% versus 19.3%, a difference that was not significant.
Guidelines and approvals
- EAU Guidelines on Non-neurogenic Female LUTS (limited update March 2026). Its evidence review describes POPPY's urinary findings: at 6 months, more control women reported urinary incontinence, straining to empty and incomplete emptying, with no difference at 12 months. The guideline advises informing women with prolapse who do not need a vaginal pessary or surgery about the potential relief of lower urinary tract symptoms from PFMT .[3]
- AUGS Best Practice Statement on evaluation and counseling of patients with pelvic organ prolapse (2025). It describes POPPY as a multicenter trial of 447 women in which the training group had a significantly greater reduction in prolapse symptoms, and states that home pelvic floor muscle exercises and pelvic floor physical therapy may be helpful for some women.[4]
Report history
| Published | Report | Cutoff | Headline results | Citation |
|---|---|---|---|---|
| 2013-11-28 | Initial report (12 months) · primary | — | Mean reduction in POP-SS from baseline at 12 months (0-28 scale); adjusted between-group difference in change: 3.77 (SD 5.62) vs 2.09 (SD 5.39); difference 1.52; CI 0.46–2.59; P 0.0053 Met Mean reduction in POP-SS from baseline at 6 months; adjusted between-group difference in change: 3.16 (SD 4.78) vs 0.12 (SD 3.86); difference 2.84; CI 2.05–3.63; P <0.0001 Any further prolapse treatment reported by 12 months (surgery, pessary, physiotherapy referral, estrogen or other): 24.1% (35/145) vs 49.6% (71/143); difference 25.5; CI 14.5–36; P <0.0001 Met Less severe POP-Q stage at 6 months (ordinal odds ratio, blinded examination); improvement by one or more stages: 26.8% (45/168) vs 19.3% (33/171); OR 1.47; CI 0.97–2.27; P 0.07 Not met Prolapse reported as 'better' than at study start, 12 months: 57.2% (83/145) vs 44.7% (63/141); difference 12.6; CI 1.1–24.1; P 0.0336 Prolapse reported as 'better' than at study start, 6 months: 52% (98/187) vs 17% (32/189); P <0.0001 Prolapse surgery reported by 12 months: 11.0% (16/145) vs 9.8% (14/143); P 0.84 Descriptive | Hagen S, Stark D, Glazener C, et al. "Individualised Pelvic Floor Muscle Training in Women With Pelvic Organ Prolapse (POPPY): A Multicentre Randomised Controlled Trial." *Lancet.* 2014;383(9919):796-806. doi:10.1016/S0140-6736(13)61977-7 DOI |
| 2022-06-28 | Record-linkage follow-up to 11 years · secondary | — | Any related secondary-care treatment for a pelvic floor disorder during follow-up (adjusted odds ratio, mixed-effects logistic regression): 43.6% (65/149) vs 52.8% (76/144); OR 0.61; CI 0.37–0.99; P 0.047 Met Time to first related secondary-care treatment (Cox regression, hazard ratio): 264 women and 125 events; adjusted for baseline POP-SS, stage, age group and BMI.; HR 0.65; CI 0.46–0.94; P 0.020 Hospital outpatient and inpatient costs for procedures of interest, 2007 to 2018: The authors found no clear difference in resource-use costs between groups at 11 years; 109 inpatient episodes of care occurred in 94 women (46 PFMT, 48 control).; — Descriptive | Fenocchi L, Best C, Mason H, et al. "Long-term effects and costs of pelvic floor muscle training for prolapse: trial follow-up record-linkage study." *Int Urogynecol J.* 2023;34(1):239-246. doi:10.1007/s00192-022-05272-9 DOI |
References
1. Hagen S, Stark D, Glazener C, et al. "Individualised Pelvic Floor Muscle Training in Women With Pelvic Organ Prolapse (POPPY): A Multicentre Randomised Controlled Trial." Lancet. 2014;383(9919):796-806. doi:10.1016/S0140-6736(13)61977-7
2. Fenocchi L, Best C, Mason H, et al. "Long-term effects and costs of pelvic floor muscle training for prolapse: trial follow-up record-linkage study." Int Urogynecol J. 2023;34(1):239-246. doi:10.1007/s00192-022-05272-9
3. European Association of Urology. EAU Guidelines on Management of Non-neurogenic Female Lower Urinary Tract Symptoms. Limited update March 2026; sections 4.7.4.a.3 and 4.7.4.a.6.
4. Carberry CL, Tulikangas PK, Ridgeway BM, et al. "American Urogynecologic Society Best Practice Statement: Evaluation and Counseling of Patients With Pelvic Organ Prolapse." Urogynecology. 2025;31(5):471-481. doi:10.1097/SPV.0000000000001641