Program

CO31-007

Oral Communication

Recommendations and settings of word prediction software by health-related professionals for people with spinal cord injury: a prospective observational study

Mr Samuel POUPLINa, Dr Nicolas ROCHEa, Prof Djamel BENSMAILa

a CHU Raymond Poincaré

Background: For people with cervical Spinal Cord Injury (SCI), access to computers can be difficult, thus several devices have been developed to facilitate their use. However, text input speed remains very slow compared to users who do not have a disability, even with these devices. Several methods have been developed to increase text input speed, such as Word Prediction Software (WPS). Health-Related Professionals (HRP) often recommend this type of software to people with cervical SCI. WPS can be customized using different settings. It is likely that the settings used will influence the effectiveness of the software on text input speed. However, there is currently a lack of literature regarding professional practices for the setting of WPS as well as the impact for users. The aim of this study is to analyze word prediction software settings used by HRP for people with cervical SCI.

Method: A questionnaire was submitted to HRP who advise tetraplegic people regarding the use of communication devices.

Results:A total of 93 professionals responded to the survey. The most frequently recommended software was Skippy, a commercially available software. HRP rated the importance of the possibility to customise the settings as high. Moreover, they rated some settings as more important than others (p<0.001). However, except for the number of words displayed, each setting was configured by less than 50% of HRP.

Discussion and Conclusion: The results showed that there was a difference between the perception of the importance of some settings and data in the literature regarding the optimization of settings. Moreover, although some parameters were considered as very important, they were rarely specifically configured. Confidence in default settings and lack of information regarding optimal settings seem to be the main reasons for this discordance. This could also explain the disparate results of studies which evaluated the impact of WPS on text input speed in people with cervical SCI.

Keywords : Health-related professionals, cervical spinal cord injury, word prediction software, settings, recommendations.