Two trials were rejected due to poor methodology. Six RCTs were deemed to be of good methodological quality and were included in the review (433 participants). The average CASP quality score was 90%. Four trials scored 100% on the CASP tool and the other two RCTs did not report on the procedure for randomisation, power analysis or intention-to-treat analysis.
Computerised cognitive-behavioural therapy was statistically superior to control post-treatment for sleep onset latency (-0.55, 95% CI -0.80 to -0.30; Ι²=0%; four trials), number of awakenings (-0.45, 95% CI -0.70 to -0.20; Ι²=51%; four trials), sleep efficiency (0.40, 95% CI 0.15 to 0.64; Ι²=63%; four trials), sleep quality (0.41, 95% CI 0.16 to 0.65; Ι²=45%; four trials) and Insomnia Severity Index (-0.86, 95% CI -1.18 to -0.53; Ι²=0%; two trials).
No statistically significant differences were found between groups for wake time after sleep onset (-0.18, 95% CI -0.43 to 0.06; Ι²=55%; four trials), total sleep time (0.22, 95% CI -0.03 to 0.46; Ι²=0%; four trials) and time in bed (-0.25, 95% CI -0.57 to 0.07; Ι²=0%; three trials).
Numbers needed to treat across four trials for the outcome of Insomnia Severity Index ranged from 1.44 to 5.36 (average of 3.59). Treatment adherence was 78% based on those who completed computerised cognitive-behavioural therapy.