RESEARCH QUESTION OF THE HTA REPORT

The aims of this investigation are to: assess the benefit of treatment with active vision training versus treatment without active vision training in children and adolescents with deficits in visual developmental in terms of patient-relevant outcomes; determine the costs (intervention costs) and assess the cost-effectiveness of treatment with active vision training versus treatment without active vision training in children and adolescents with deficits in visual development as well as review ethical, social, legal, and organizational aspects associated with the medical intervention.

CONCLUSION OF THE HTA REPORT

A total of 17 randomized trials evaluating active digital vision training in children and adolescents were included in the present health technology assessment (HTA). The identified studies focused on patients with amblyopia (from anisometropia and/or strabismus). To date, no results from randomized trials have been published on other relevant deficits in visual development (refractive anomalies or eye misalignment) or on vision training interventions which are not digitally delivered. Of the 17 included studies, 11 studies (with N = 1138 children and adolescents) evaluated the benefit of dichoptic (binocular) vision training, while 6 studies (with N = 165 children and adolescents) investigated the benefit of monocular vision training. Dichoptic training was compared with (A) no training, (B) sham training, or (C) conventional treatment (occlusion). Treatment was provided at home in all but 1 study, where children and adolescents attended an outpatient clinic. In total, 2 studies evaluated children with amblyopia induced exclusively by anisometropia (refractive amblyopia). The remaining studies included amblyopia from anisometropia and/or strabismus. The age of the children and adolescents in the 11 studies on dichoptic training ranged from 3 to 17 years, and the proportion of those who had received prior treatment ranged from 26% to 96%. In 10 studies, dichoptic training was performed using a video game (e.g. Tetris), while in 1 study, participants watched films or series. The image separation required during dichoptic training was achieved via anaglyph glasses (8 studies), shutter glasses (1 study), or virtual reality glasses (2 studies). Except in the 2 studies using virtual reality glasses, training was conducted on a computer or tablet. In total, 6 studies investigated the benefit of monocular vision training compared with (A) no training or (B) sham training. In 4 studies, monocular training, where the stronger eye is occluded, was performed on an outpatient basis. Two studies evaluated only refractive amblyopia, and 4 studies investigated only 2- to 10-year-old children who had received their initial diagnosis. While monocular training involved playing a video game on a computer or tablet in 5 studies, 1 study asked children and adolescents to trace images which were presented digitally. Three studies used special background stimulation (a moving sinusoidal grating for neural stimulation) during monocular training. The remaining 3 studies did not use this background stimulation. All studies reported data for the outcome “best-corrected visual acuity of the amblyopic eye". (A) Comparing active digital vision training with no training, significant effects were found in favour of active vision training — both dichoptic and monocular. The available study pool fails to show whether the additional occlusion treatment (which some studies used as add-on treatment in both intervention arms) enhances the effect of digital training. (B) When comparing vision training versus sham training, 1 of 6 studies described a statistically significant effect in favour of the investigated intervention. The missing effects in this comparison category could be due to the fact that the respective sham intervention (which was also based on vision training) showed the same effect as the experimental intervention. (C) Only studies on dichoptic training investigated digital vision training in the form of monotherapy versus occlusion treatment alone. Except for 1 study (with a treatment duration of only 2 weeks), no studies showed any statistically significant effects in favour of vision training – on the contrary, the effect estimators were in favour of occlusion treatment. Although isolated studies suggest a statistically significant effect in favour of active vision training compared to no training or sham training for the outcome “best-corrected visual acuity of the amblyopic eye”, the differences measured in these comparisons cannot be assumed to be clinically relevant. Most studies with statistically significant results do not meet a clinical relevance threshold of -0.05 logMAR (improvement by +0.5 lines) for the outcome “best-corrected visual acuity of the amblyopic eye”. Only in 1 study, which compared monocular vision training versus no training, was the clinical relevance threshold exceeded at 1 of 4 measurement time points: The upper and lower limits (95% CIs) of measured mean differences after 9 weeks of treatment were -0.24 logMAR (+2.4 lines) and -0.06 logMAR (+0.6 lines), respectively. In summary, for the outcome “best-corrected visual acuity of the amblyopic eye,” no hint of (greater) benefit of digital dichoptic or monocular vision training was found in any of the comparisons: neither compared to no training nor to sham training or occlusion treatment. For the patient-relevant outcome of binocular vision, none of the studies showed a hint of greater benefit of digital training – neither compared to no training nor to sham training or occlusion treatment. In the present study pool, the number of reported adverse events was rather low, and no study showed any hint of an increased risk of such events – neither with digital training nor in the control group. However, it must be noted that the rather small number of study participants precludes drawing a reliable conclusion with regard to adverse events. No data are available on other patient-relevant outcomes. In Germany, vision training is currently used only as an add-on to occlusion treatment. The average total costs per person for occlusion treatment in the 1 year of treatment are between €606 and €646, which is largely covered by the statutory health insurance (SHI); 2/3 of the costs are for the aids used in the treatment. In addition to these costs, 3 months of treatment with the monocular vision training offered in Germany by the company Caterna Vision GmbH would cost €380. Some health insurance funds cover the costs for the vision training within the framework of selective contracts. Apart from the costs presented for monocular visual training by Caterna Vision GmbH, no further data were available from which a reliable cost estimate for active vision training might be derived. In summary, for vision training to be cost-effective, the additional costs must be associated with either greater benefit (in the form of better treatment outcomes) or savings in conventional treatment (e.g. through reduced treatment times for the same benefit). Since neither of these conditions is currently met, the cost-effectiveness of active vision training in amblyopia cannot be conclusively assessed. Given that no studies were found on other relevant developmental vision impairments or on other non-digitally delivered vision training, extending the economic evaluations to other therapeutic indications or interventions was not deemed useful. When making decisions regarding digital training, the early and risk-associated access to digital media and the increased media consumption associated with such therapy must be considered critically from ethical and social perspectives. In terms of adherence to the prescribed therapy, study data additionally show that problems and treatment discontinuations repeatedly occur in practice. These are often due to a lack of understanding of the video game and/or patients being overchallenged (especially younger children) or underchallenged (especially older children and adolescents) because the ‘health games’ offered are typically video games with a simple design (e.g. Tetris). With regard to the home therapy concept, retaining attention therefore plays a critical role in active vision training. Overall, the available evidence does not allow drawing a final conclusion on the benefit of active vision training in children and adolescents with amblyopia. Alongside the frequent lack of treatment adherence, it is important to note that the studies’ treatment duration was set to only a few weeks, while amblyopia treatment is often required for many years. Furthermore, it was impossible to deduce from the heterogeneous study pool whether children with refractive amblyopia achieve better results than children with amblyopia from strabismus. Such results would be important for the benefit assessment of digital therapies, however, in order to better assess which patients benefit most from such therapy. In addition, further studies are needed to determine whether more interesting (possibly also non-digital) games or vision training interventions lead to better treatment adherence and possibly also show greater benefit with regard to patient-relevant outcomes – especially in older children. Overall, it can be concluded that the outcomes of the new therapies were not superior.