Eleven RCTs (15 articles; 1,506 patients, range 34 to 476) were included in the review. Follow-up durations ranged from immediately post-intervention up to 12 months. Two RCTs were at low risk of bias for all criteria; the other RCTs had unclear risk of bias for one or more criterion.
One RCT of patients with diabetes reported that patients who received computer-based interventions showed significant improvements in glycosylated haemoglobin levels at three months compared to standard education.
Five RCTs compared computer-based interventions to interventions with no patient self-management element (no intervention, usual care or assessment only). All five RCTs reported benefit in a behavioural or clinical outcome.
The remaining five RCTs compared computer-based interventions combined with standard patient self-management versus standard patients self-management alone. Only one RCT reported improvements in certain behaviours in adult atopic asthmatics. The remaining four RCTs reported no significant differences between treatment groups in patients with varying health conditions (fully reported in the review).
Four RCTs assessed acceptability of the computer-based interventions; there was no evidence of high attrition which the authors took as a proxy measure of intervention acceptability. Other findings were reported in the review.