Table 1 Summary of findings against 14 methodological issues for feasibility research

1. Did the feasibility/pilot study allow a sample size calculation for the main trial?

Achieved: a definitive trial would require recruitment of between 232 (for a target difference of four units in ICIQ-FLUTS score at 6 months) and 922 (for a target difference of two units) patients, to achieve a sample of 130 to 516 primary outcome responses.

Observed standard deviation for difference in ICIQ-FLUTS = 7.

Observed retention rate of 63 %.

2. What factors influenced eligibility and what proportion of those patients approached were eligible?

37 % of those approached and assessed screened positive.

Of 771 women identified and approached, 284 were (37 %) were deemed eligible.

Main reasons for ineligibility (see Table 2) were as follows:

• Patient has not undergone a course of pelvic floor training

• Urge incontinence

• Patient did not attend clinic

3. Was recruitment successful?

Recruitment was slower than anticipated

The mean rate of recruitment per open site month at original sites was 1.9, with a recruitment rate per open site month at the additional full recruiting site of 2.5.

In 8 months, the PICs did not identify any potentially eligible patients

4. Did eligible participants consent?

Largely achieved, with a recruitment total of 222, representing 93 % of the target of 240.

Of the 284 women who screened positive, 222 (78 %) consented to randomisation

5. Were participants successfully randomised and did randomisation yield equality in groups?

Achieved

Overall, 110 women were randomised to the control arm and 112 to the intervention arm.

Baseline comparability of the two groups was adequate (Tables 3 and 4)

6. Were blinding procedures adequate?

Not applicable - non-blinded design

7. Did participants adhere to the intervention?

Largely achieved

In the control group, one woman was found to be ineligible post-randomisation and was withdrawn. The remaining 109 received no IUT.

In the intervention group, 102 of the 112 women randomised (91 %) received IUT. 2 withdrew because they were unhappy with their allocation, 1 did not attend for cystometry, 3 withdrew for other reasons, and 4 did not receive IUT.

8. Was the intervention acceptable to participants?

Mixed findings; 59 eligible women did not wish to be randomised. Qualitative interviews provided insights into preferences of those consenting to randomisation.

Although most eligible women were willing to be randomised, some had a previously undeclared preference for avoiding IUT and expressed relief at being allocated to the control group. These data will be reported separately elsewhere.

We did not succeed in interviewing any women who did not consent to the trial.

9. Was it possible to calculate intervention costs and duration?

We have demonstrated that meaningful and usable data were collected using the instruments we designed for this purpose, and it is feasible to calculate intervention costs based on data collected and from reference costs information. These data will be reported separately elsewhere.

Questionnaires and CRF pages designed to collect relevant information for costs calculation performed reasonably well, with a good response rate and low level of missing data. The average total cost per patient is £1815.26 (SD: 210.39) for ‘IUT’ arm and £1775.37 (SD: 455.38) for ‘no IUT’ arm, based on complete cases only. The mean duration of ‘IUT’ was 40 minutes (SD: 11.028)

10. Were outcome assessments completed?

Mixed picture with poorer completion of outcome assessments at 6 months.

75 % of women returned baseline questionnaires and 63 % returned 6-month questionnaires. Of those who returned questionnaires, some returned incomplete or blank questionnaires

Bladder diaries and pad use data were poorly completed.

Baseline full completion rates:

ICIQ-FLUTS: 98 %

ICIQ-UI SF: 99 %

ICIQ-LUTSqol: 95 %

UDI overall score: 84 %

6-month full completion rates:

ICIQ-FLUTS: 90 %

ICIQ-UI SF: 91 %

ICIQ-LUTSqol: 87 %

UDI overall score: 81 %

11. Were outcomes measured those that were the most important?

Some evidence from patient interviews that women were less likely to return questionnaires if they were satisfied with the results of treatment

Lower response rates for instruments towards the end of questionnaire booklet

12. Was retention to the study good?

Rates of loss to follow-up were significant

75 % of women had face-to-face or telephone follow-up after surgical treatment, but only 56 % (63 % of those circulated) returned follow-up questionnaires at 6 months.

13. Were the logistics of running a multicentre trial assessed?

Achieved. Some centres performed better than others, and PICs were not fruitful.

The need to build in adequate time for obtaining global and local approvals was identified

14. Did all components of the protocol work together?

Achieved - components had good synergy

No significant differences identified with trial processes or researchers’ abilities to implement them