Background: Our systematic review and meta-analysis of the benefit of self-monitoring of blood glucose (SMBG) in improving glycemic control in type 2 diabetes was published in 2008. With the few studies that have emerged afterward, we undertook subsequent meta-analysis of the available evidence to update the results.
Methods: Clinical trials of SMBG were identified through electronic searches (MEDLINE, EMBASE, and The Cochrane Library) up to and including June 2009. Studies were included if they met the following inclusion criteria: (1) randomized controlled trial comparing SMBG versus non-SMBG in type 2 diabetes patients not using insulin and (2) hemoglobin A1c (HbA(1c)) reported as an outcome measure. The efficacy was estimated with the mean difference in the changes of HbA(1c) from baseline to final assessment between the SMBG and the non-SMBG groups.
Results: SMBG was effective in reducing HbA(1c) in non-insulin-treated type 2 diabetes (pooled mean difference, -0.24%; 95% confidence interval, -0.34% to -0.14%; P < 0.00001). Glycemic control significantly improved among the subgroup of patients whose baseline HbA(1c) was >or=8%. In contrast, no significant effect of SMBG was detected in patients who had HbA(1c) <8%.
Conclusions: The available evidence suggests the usefulness of SMBG in improving glycemic control in non-insulin-treated type 2 diabetes as demonstrated by the reduction of HbA(1c) levels. In particular, SMBG proved to be useful in the subgroup of patients whose baseline HbA(1c) was >or=8%.