A pilot randomized controlled trial comparing effectiveness of prism glasses, visual search training and standard care in hemianopia

ObjectivePilot trial to compare prism therapy and visual search training, for homonymous hemianopia, to standard care (information only). MethodsProspective, multicentre, parallel, single-blind, three-arm RCT across fifteen UK acute stroke units. ParticipantsStroke survivors with homonymous hemianopia. InterventionsArm a (Fresnel prisms) for minimum 2hours, 5days per week over 6weeks. Arm b (visual search training) for minimum 30minutes, 5days per week over 6weeks. Arm c (standard careinformation only). Inclusion criteriaAdult stroke survivors (>18years), stable hemianopia, visual acuity better than 0.5 logMAR, refractive error within 5 dioptres, ability to read/understand English and provide consent. OutcomesPrimary outcomes were change in visual field area from baseline to 26weeks and calculation of sample size for a definitive trial. Secondary measures included Rivermead Mobility Index, Visual Function Questionnaire 25/10, Nottingham Extended Activities of Daily Living, Euro Qual, Short Form-12 questionnaires and Radner reading ability. Measures were post-randomization at baseline and 6, 12 and 26weeks. RandomizationRandomization block lists stratified by site and partial/complete hemianopia. BlindingAllocations disclosed to patients. Primary outcome assessor blind to treatment allocation. ResultsEighty-seven patients were recruited: 27Fresnel prisms, 30visual search training and 30standard care; 69% male; mean age 69years (SD 12). At 26weeks, full results for 24, 24 and 22 patients, respectively, were compared to baseline. Sample size calculation for a definitive trial determined as 269 participants per arm for a 200 degree(2) visual field area change at 90% power. Non-significant relative change in area of visual field was 5%, 8% and 3.5%, respectively, for the three groups. Visual Function Questionnaire responses improved significantly from baseline to 26weeks with visual search training (60 [SD 19] to 68.4 [SD 20]) compared to Fresnel prisms (68.5 [SD 16.4] to 68.2 [18.4]: 7% difference) and standard care (63.7 [SD 19.4] to 59.8 [SD 22.7]: 10% difference), P=.05. Related adverse events were common with Fresnel prisms (69.2%; typically headaches). ConclusionsNo significant change occurred for area of visual field area across arms over follow-up. Visual search training had significant improvement in vision-related quality of life. Prism therapy produced adverse events in 69%. Visual search training results warrant further investigation.