An overview and commentary on retrospective, continuous glucose monitoring for the optimisation of care for people with diabetes

Introduction: Normoglycaemia in people with diabetes results in improved outcome. Continuous glucose monitoring provides detailed diagnostic information used to optimise therapy with the goal of achieving normoglycaemia. The objective of this study was to review the published literature evaluating the single device available for blinded, professional clinical use of continuous glucose monitoring; in particular all randomised controlled trials (RCTs) and relevant observational studies. Methods: Published studies (to 31 March 2009) using the Medtronic MiniMed Continuous Glucose Monitoring System (CGMS® System Gold™, Medtronic MiniMed, Inc., Northridge, CA) were identified using appropriate search terms in a series of clinical databases including: Medline, Pubmed, Google Scholar and Scientific Web of Knowledge. Other measures were also taken such as reviewing the reference lists. Findings: In addition to an extensive range of non-randomised studies using the device, seven randomised controlled trials were identified, of which five were in children. Although HbA1c was explicitly stated as the primary endpoint in four studies, the studies were only adequately powered to detect large, between-group minimum differences (range 0.5 to 1.0% HbA1c). Only two studies included subjects with type 2 diabetes. Other endpoints included the frequency of hypoglycaemia and hyperglycaemia. Within-group HbA1c decreases were observed in all but one study. The crude, weighted mean improvement in HbA1c across the studies using the device was 0.6% (range 0% to 0.8%). When compared to a control measure the weighted, mean marginal benefit in HbA1c was 0.2%. No studies recorded how the information was utilised to modify treatment, such as changes in dose titration or treatment regimen. Conclusions: Devices such as the professional CGMS provide detailed diagnostic information. However, any consequent changes in care and outcome can only result from the appropriate application of this information. Although there was a notable improvement in glucose control (HbA1c), existing studies were largely underpowered and did not capture the resulting treatment changes that could lead to improved outcome