A Novel Approach to Assess Performance of Glucometer: Error Grid Analysis

Abstract Glucose meters are excellent tools for self-monitoring of blood glucose especially in the circumstances where continuous monitoring is mandatory and at decision-making levels. Tight glycemic control protocols are important for preventing ill effects of fluctuating glucose levels. As technology advances, glucometers are getting better in terms of quality of results. The purpose here is to study the quality of glucose meter results in terms of clinical outcomes. The study was conducted in a tertiary care referral hospital. A total of 125 patients were recruited from pediatric wards. Bland-Altman plot, Parke error grid, and surveillance error grid analysis were used for comparing the results of the glucose meters with a standard laboratory method. It was found that there was a significant difference between the results of the 2 methods. Although usually minimal, the glucose meter results deviated from the results of the standard laboratory method. This will affect overall patient care especially in emergency conditions. Sometimes, the risk may be so high that the patient may be mislabeled as hypoglycemic when they actually are hyperglycemic and vice versa. This study is the first of its kind because no similar studies have been reported in the pediatric population. For effective use of glucose meters, the devices should give as accurate results as possible. Results should not only be accurate but also precise, without which critical errors may be possible. We recommend that for any glucose meter there should be regular calibration so that agreement with the reference laboratory method is maintained and effective medical decisions are possible.