As diabetes becomes more prevalent, the cost to treat and monitor diabetes rises. In the United States the direct medical costs for a person with diabetes is 2.3 times higher than a person without diabetes, totaling $116 billion dollars in 2007.1 Blood glucose test strips are in the top five of total expenditures in Canada and was the largest single component of diabetes management in the United Kingdom.2,3 It is unclear if there is a benefit of self-monitoring blood glucose (SMBG) in patients with type two diabetes mellitus (T2DM) not being treated with insulin.4,5 Furthermore, it is unknown if the benefits of SMBG in these patients outweigh the additional costs.2,6
Cameron and colleagues conducted a systematic review to determine if SMBG in patients with T2DM and not on insulin was cost-effective.7 The review included seven randomized controlled trials (RCT) that investigated the effect of SMBG in noninsulin treated patients versus no SMBG. These trials included a total of 2270 patients with a mean age range of 50-66, male participants ranged 21-66%, and the patients had a previous history of diabetes for 3-11 years. The investigators used data from the systematic review and United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model to forecast long-term health complications for T2DM and perform a cost-effectiveness analysis of whether SMBG is cost-effective in noninsulin treated T2DM patients.2 The analysis was done in the perspective of the Canadian ministry of health and data on cost per test strip and utilization was obtained from the Ontario Public Drug Program and Long-term Care.2 Researchers used the testing frequencies from the seven RCT to calculate the average number of test strips used per day (approximately 9 test strips per week).2 Quality-adjusted life-years and costs were inflated or discounted according to the guidelines of the Canadian Agency for Drugs and Technologies in Health.2
The primary outcome of the review was quality-adjusted life-years using the US catalogue of EuroQol 5-dimension (EQ-5D) scores. Noninsulin-treated T2DM patients without a history of diabetes related complications (as defined by UKPDS) were assumed to have an EQ-5D score of 0.753. The UKPDS model uses hemoglobin A1C (HbA1C) as a surrogate marker to project long-term complications associated with T2DM. A mean decrease in HbA1C of 0.25% was found in the self-monitoring group. This decrease in HbA1C correlated to a 0.024 increase in quality-adjusted life-years in the self-monitoring group. The self-monitoring group had an additional cost of $2,711 to gain 0.024 quality-adjust life-years. The researchers found that self-monitoring would cost $113,643 to gain one quality-adjusted life-year. The authors concluded that the benefits of self-monitoring blood glucose in noninsulin treated T2DM does not offset the costs of blood glucose test strips. Therefore, self-monitoring is not an effective or efficient use of health care resources because the clinical benefits of self-monitoring and the related savings do not outweigh the cost of the blood glucose test strips.