Self-monitoring of glucose levels for people with type 2 diabetes.

To determine the most effective way of monitoring glucose levels as an indicator of glycaemic control in people with type 2 diabetes in the community, we conducted a criteria-based review of randomized controlled trials and systematic reviews of randomized controlled trials that studied the efficacy of various glucose monitoring strategies. We searched the Cochrane Library, Medline, Embase, CINAHL and BNIPlus databases for relevant studies. The journals 'Diabetes', 'Diabetic Medicine', 'Diabetologica', 'Evidence-Based Medicine' and 'Evidence-Based Nursing' were hand searched. The outcome of interest was glycaemic control, as measured by glycated haemoglobin (HbA(1c)). A total of 642 titles were identified from the search; three studies answered the question criteria and only one study met all the quality criteria. The study that met the criteria was a systematic review of four trials measuring the efficacy of self-monitoring of glucose levels. The reduction in HbA(1c) in those who monitored glucose levels was estimated to be -0.25% (95% Cl -0.61 - +0.10). This result shows a small improvement, but it is not statistically significant. A meta-analysis was also performed on three studies (n=278) comparing HbA(1c) in subjects who performed blood glucose monitoring with those who performed urine monitoring. The reduction in HbA(1c) when monitoring blood glucose rather than urine glucose was -0.03% (95% Cl -0.52 - +0.47). This result is not statistically significant. The efficacy of blood and urine glucose monitoring testing, for people with type 2 diabetes, in improving glycaemic control as measured by HbA(1c) levels is still questionable. A rigorous randomized controlled trial is needed to establish these answers although there is no evidence of harm. Clinical protocols that make recommendations for glucose monitoring strategies for people with type 2 diabetes should acknowledge that the evidence is weak. There is no basis to recommend one method above another.