Word prediction software in people with cervical spinal cord injury: A literature review

Abstract

Individuals with cervical spinal cord injury (ICSCI) have difficulty accessing computers through conventional methods. The purpose of this paper was to provide a critical narrative review on types of computer access devices and word prediction software for ICSCI. A search was conducted in several electronic databases. There is limited data about Test Input Speed (TIS) in ICSCI. One study showed that TIS in ICSCI was significantly lower [mean = 11 words per minute (wpm)] ( P = 0.001) than TIS of abled-bodied individuals (19 wpm). In a sample of ICSCI, there was a significant difference between voice recognition system (mean = 18 wpm), a standard keyboard (mean = 8 wpm) and an onscreen keyboard (mean = 3 wpm) ( P < 0.001). However, the literature suggested that the impact of Word Prediction Software (WPS) for increasing TIS remains questionable. Indeed, there is limited scientific documentation of the influence of the parameters of WPS on TIS. Regarding the number of words displayed in the prediction list, this parameter did not increase TIS. For the parameter “frequency of use”, use of WPS with the activation of frequency of use increased TIS in participants with high-level tetraplegia. For participants with low-level tetraplegia, the use of WPS with frequency of use activated only decreased the number of errors. Each professional could provide clear and accurate information on the benefits of WPS for people with cervical SCI. WPS must be set regarding the needs of people with cervical SCI. If not, the use of these software's may go against the expected benefits.