A systematic review on visual scanning behaviour in hemianopia considering task specificity, performance improvement, spontaneous and training-induced adaptations.

Only two studies looked at which scanning behaviour related to search performance and only one at reading, and no study measured how training changes scanning in mobility tasks, so the review's training suggestions rest on a very small base. Half of the articles behind the search and reading findings scored low on the quality appraisal and were kept in anyway, so those results are less reliable. Studies used different definitions and measures of scanning, including different definitions of a fixation, which makes them hard to compare, and most mobility studies did not use a gaze-in-world reference system, which the authors say makes those findings hard to translate to real life. Group averages may hide individual differences: people with hemianopia differ in their other problems after brain injury, such as slower visual processing or reduced working memory, and training built on group averages may not suit an individual. The review examined scanning behaviour, not whether training improves task performance, so it cannot say that any training programme works.

The review was funded by the Visio Foundation in Amsterdam, and four of the authors were also partly supported by the Dutch ZonMw programme Expertisefunctie Zintuigelijk Gehandicapten. The funding and conflicts section reports no other competing interests.

Do not read this as proof that teaching a new scanning pattern improves reading, search or driving in hemianopia. The review examined which scanning behaviours went with better performance in individual studies, not whether training changes performance, and most of the suggested targets come from one or two small studies. The authors themselves say more research is needed before their suggestions are put into training.