Abstract
BACKGROUND: The use of self-monitoring of blood glucose (SMBG) in patients with non-insulin-treated type 2 diabetes is debated. Meta-analyses of randomized clinical trials (RCTs) suggest a small reduction of HbA1c in patients using SMBG, without considering potential confounders, such as SMBG regimen and use of SMBG data to adjust diabetes medications. METHODS: A meta-analysis was performed including RCTs in patients with non-insulin-treated type 2 diabetes, with an intervention of ≥24 weeks and HbA1c as the primary endpoint, to verify the effect of SMBG (vs no monitoring), structured SMBG (vs unstructured), and of SMBG-driven therapy adjustments. RESULTS: In RCTs (n = 8) comparing SMBG with no SMBG (1277 and 1072 patients, respectively), SMBG reduced HbA1c by -0.17% (95% CI -0.25 to -0.09%, P < .003). The reduction in HbA1c was greater in RCTs (n = 3) in which SMBG data were used to adjust diabetes medications (HbA1c decrease: -0.3% [95% CI -0.49 to -0.1%]) than in RCTs (n = 6) where SMBG data were not used for this purpose (HbA1c decrease: -0.1% [95% CI -0.2 to 0.0%]) ( P < .005). In the RCTs comparing structured and unstructured SMBG (757 and 750 patients, respectively), in which structured SMBG data were also used to adjust diabetes medications, the HbA1c difference between groups at study end was -0.27% (95% CI -0.49 to -0.04%, P < .018). CONCLUSIONS: In RCTs performed in non-insulin-treated patients with type 2 diabetes, SMBG is associated with a significant, although small, reduction in HbA1c. HbA1c reduction was greater with structured SMBG and when structured SMBG data were used to adjust diabetes therapy.