Numerical Estimation of HbA1c from Routine Self-Monitoring Data in People with Type 1 and Type 2 Diabetes Mellitus

Abstract

This chapter presents a numerical estimation study of HbA 1c from routine self-monitoring data in people with type 1 and type 2 diabetes mellitus. Models are presented using self-monitoring of blood glucose (SMBG) to estimate HbA1c, the most important and widely accepted marker of glycemic control of people with type 1 or type 2 diabetes. In parallel with the optimization of the estimating functions, the optimal duration of the SMBG data collection period was studied along with the optimal frequency of self-monitoring during that period. The optimal SMBG data collection period is 45 days, and the optimal frequency of SMBG is three reading per day. Two optimal HbA 1c linear functions were developed: Model 1—using only SMBG data, and Model 2—using SMBG data plus an HbA 1c reading taken approximately 6 months prior to the HbA 1c that is being predicted. It was concluded that SMBG data contain valuable information about individuals' glycemic control that, using uncomplicated computation, can be accurately translated into HbA 1c . Because SMBG data are recorded and stored by most contemporary home-monitoring devices, incorporating an appropriate computation and presentation of results would take the utility of such devices well beyond the currently available display of a momentary snapshot of patients' glucose status.