Poggel 2004.

Study characteristics
Methods	Restitutive intervention versus restitutive intervention Design: prospective RCT Stratification: no Randomisation sequence: stated participants were "assigned randomly" but no details given. Correspondence with author did provide any further detail. Comparisons: 2 groups ‐ visual restoration therapy (VRT) with attentional cueing, VRT without attentional cueing Allocation concealment: no ‐ "Patients were not told which type of training they received; a double‐blind control was not possible owing to staff limitations." Blinding: no Power calculation: no Intention‐to‐treat analysis: no. 1 dropout, for whom no data was presented. Stated: "Patient 5 did not participate in the training procedure owing to time constraints and is not listed here." Other recruitment details: not stated Patient and public involvement: not stated
Participants	Total study population: 20 recruited (data available for 19) Withdrawals: 1 ("Patient 5 did not participate in the training procedure owing to time constraints and is not listed here.") Method of diagnosing VFD: high‐resolution perimetry and standard Tübingen Automated Perimetry Characteristics of population: participant details are listed in Table 6. Type and severity of visual problems: participant details are listed in Table 7. Inclusion criteria: homonymous visual field defects after postgenicular lesions, age 18 to 75 years. Visual field size, assessed by high‐resolution campimetry, had to be stable; that is, patients were included only if visual field size increased or decreased by < 2% over at least 4 weeks before the study started". Exclusion criteria: optic nerve, retinal or other ophthalmic disorders; mental deficits; impaired attention and neglect; psychiatric disorders; epilepsy and diseases with obvious visual or cognitive effect Baseline comparison of treatment groups: not stated This study included patients with visual field defects only (no visual neglect). Method of diagnosing visual perceptual problems: not stated
Interventions	Group 1: visual restitution therapy (VRT) with attentional cueing (n = 10) Intervention: computer‐based stimulation of visual field, with cueing Intervention type: restitution Materials: computer‐based vision restoration therapy (Nova Vision, Magdeburg, Germany). Procedures: a high‐resolution computer‐based campimetric test (high‐resolution perimetry [HRP]; Nova Vision, Magdeburg, Germany) was used to assess visual field size and to determine ARVs.Training stimuli appeared on a dark computer screen, each target increasing in brightness in four steps from dark gray (30 cd/m2) to bright white (96 cd/m2) over 2000 milliseconds. Stimulus size, fixation control, mode of response, and viewing distance were identical to those used for HRP. For vision restoration therapy with attentional cueing, the training stimulus was preceded by a large (12° x 12° visual angle) dim gray cue frame enclosing a predetermined segment in the upper visual field that included parts of the intact and blind fields as well as ARVs. The frame appeared for 200 milliseconds and was followed at a randomised interval (mean stimulus onset asynchrony: 1000 milliseconds, range 750 to 1250 milliseconds) by a training stimulus presented in the area that had been stimulated by the cue frame (the attention field). In each training session, patients received 500 training stimuli. In the experimental group, approximately one‐third of the trials, that is, those where the target was to be presented in the attention field, were preceded by a cue. Targets were presented at random locations within the upper and the lower visual fields. Depending on the percentage of stimuli detected, the duration of each training session was approximately 30 to 35 minutes." Provided by: not stated.Delivery: training completed at home using computer (Table 5). Regimen: "Patients performed the training in six training units, each lasting approximately 1 month. A unit consisted of 56 sessions, with two sessions per day." Tailoring: yes: "Each patient's ARV was determined based on five HRP tests. The training area was then adjusted to the visual field border of that individual." Modification: not stated. Adherence: not stated ("Data from each session were saved on a disk") Group 2: visual restitution therapy (VRT) with no attentional cueing (n = 9) Intervention: computer‐based stimulation of visual field Intervention type: restitution Materials: see above. Procedures: see above.Provided by: see above. Delivery: see above. (Table 5). Regimen: see above. Tailoring: see above. Modification: see above. Adherence: see above
Outcomes	See Table 8 Change in size of visual field: Tubinger automated perimetry (TAP) High resolution perimetry (HRP) Time points when outcomes were assessed: at the end of each training unit (HRP), and after the 6th training unit (HRP and TAP)
Notes	Training and measurement were essentially the same thing. Authors referred to Kasten 2001 as proof that control was not needed. They did not address his conclusions on effects of training when recruiting from an already trained group ‐ see Jobke 2009 above. Outcome data was largely presented for all patients combined (control and experimental groups combined)
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation concealment (selection bias)	Unclear risk	"Patients were not told which type of training they received; a double‐blind control was not possible owing to staff limitations" ‐ the paper did not state what methods (if any) were used to conceal the allocation from those enrolling and assigning participants and correspondence with the author did not provide any further information.
Blinding (performance bias and detection bias) All outcomes	High risk	"Patients were not told which type of training they received; a double‐blind control was not possible owing to staff limitations".
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	1 participant did not participate, none of his data were included in final analysis. The co‐author was not totally clear on the reason for non‐participation, but believed it to be another stroke.
Other bias	High risk	Researchers had financial interest in intervention being investigated: "BA Sabel and E Kasten are inventors. BA Sabel is a consultant and shareholder of NovaVision Inc, and has equity in excess of USD 10,000".