Page 65 - Glucose Monitoring Devices
P. 65

An extended illustration   63





                  An extended illustration
                  In this section, we discuss the design and results of two previously published studies
                  [18,54,76,92] that illustrate the approach introduced in Accuracy and its conse-
                  quences section. The goals of the project were (1) to assess how different aspects
                  of a BGM system accuracy affect HbA1c, severe hypoglycemia, insulin, and finger-
                  stick use (and costs); and (2) understand how the BGM system accuracy affects
                  long-term financial costs to the patient and the health system. A BGM system
                  includes the meter as well as the strips (both can cause inaccuracy). The project
                  was constrained to type 1 diabetes patients using CSII. The general approach
                  followed is described in Fig. 4.7.
                     The motivation of the study (first level of Fig. 4.7) is to understand how BGM
                  system accuracy contributes to clinical outcomes and ultimately to healthcare costs.
                  In our approach (second level of Fig. 4.7), we needed to simulate patients with
                  different brands of BGM systems, observe how their glycemic control and clinical
                  outcomes were affected, and ultimately assess whether the systems were resulting
                  in increased costs for the health system. To make this possible a set of tools and
                  methods were applied (third level in the figure). First, we created an in silico
                  evaluation platform based on the UVA/Padova simulator [69] where models of mul-
                  tiple BGM systems were coded along with a behavioral model. Second, using the
                  glucose traces resulting from our in silico studies as basis, severe hypoglycemia
                  events per person per year and HbA1c levels were estimated. Simulation outputs
                  were also used to monitor insulin and fingerstick use. Next, the HbA1c Translator
                  was used to produce an estimate of diabetes-related complications due to changes
                  in HbA1c. Finally, using treatment costs for each complication from the United
                  Kingdom, the total costs to the healthcare system were computed. In the following
                  sections, we describe each of the stages.

                  Motivation
                     Meter                       Clinical                  Healthcare
                    Accuracy                    Outcomes                     Costs
                   Approach
                     Patient      Glycemic       Clinical                  Healthcare
                                                            Complications
                     Meter         Control      Outcomes                     Costs
                   Methods & Tools

                       UVA/Padova     HbA1c & Hypo    HbA1c Translator
                        Simulator      Estimatation

                  FIGURE 4.7
                  Approach followed in our illustration.
   60   61   62   63   64   65   66   67   68   69   70