Table 8 Summary of issues raised and implications for a future definitive trial

Identifying potential participants

a) Via third sector organisations

Not clear how many had been contacted

Very time consuming

Impractical to use research staff

b) Via low vision clinic

Only few expressed interest when Eye Clinic Liaison Officers involved

Would need to support third sector staff more

Debatable whether Eye Clinic is worthwhile route

First contact may be better via GP

Did eligible participants consent?

66/68 (97%) of those eligible consented to enter trial: exceeded target of 50%

Estimate biased upwards since some declined before being screened

Unreliable estimate from this pilot trial

Met recruitment target?

Randomised 64 participants to trial: below target of 80

Recruitment was more difficult than anticipated: need to improve procedures

Compliance with intervention

25/33 (76%) completed at least 9/12 group sessions: exceeded target of 70%

Compliance with exercise class regime is possible

Home exercise duration per week

Median duration 50 min: encouraged to exercise for 120 min

Need to find ways of encouraging participants to exercise at home for longer

Intervention

FaME intervention successfully adapted for VIOP

Need exercises appropriate to ability of individual, providing sufficient challenge

Retention throughout study

59 participants retained to end of study: narrowly missed target of 60

Retention was better than anticipated

Outcome assessments completed

60/64 (94%) provided data at 12 week visit

59/64 (92%) provided data at 24 week visit: exceeded target of 70%

Very few items of missing data on any scale

Once recruited, retention and data collection was very good. No problem with completion if researchers can help, but this is time-consuming

Suitability of candidate outcome measures

Suitable, howeverelements of standard assessments were occasionally inappropriate for VIOP

Other outcome assessments may additionally be included.

Safety issues?

16 serious adverse events, but none deemed due to intervention: less than target of < 10%

No safety issues

Sample size calculation for definitive trial

Calculation very imprecise because of lack of estimates of parameters for SFES-1

Better estimates necessary

Data collection for health economic analysis

Data often missing from resources form

Data collection form needs to be more structured