Twenty nine studies were included in the review; 9 cross-sectional studies (n=6,191 patients, range 84 to 2,855); 9 longitudinal studies (n=29,900 patients, range 38 to 23,412); and 11 RCTs (n=1,759 patients, range 10 to 689).
Results from the cross-sectional studies were inconclusive; two reported self monitoring to be correlated with lower glycosylated haemoglobin (HbA1c), three with higher HbA1c, and four showed no correlation.
Results from the longitudinal studies were also mixed; four reported no association and four a lower HbA1c with self-monitoring.
Of the RCTs, six reported improved glucose control, two reported no effect of a monitoring intervention, one showed improved glycaemic control with the use of a booklet, and two reported reductions in HbA1c with monitoring of urine or blood glucose, but no difference between these two methods.
Mediators and moderators of self-monitoring were discussed.