Data from a randomized controlled trial were used for this analysis. The details of the study are described in . Briefly, the study consisted of type 1 and type 2 individuals on insulin therapy with two arms: (1) the intervention group utilized a continuous glucose monitor (CGM) to assess daily glucose levels and (2) the control group relied on capillary glucose testing. After an initial 14-day masked CGM baseline period, the diabetes management intervention period was 85 days, with a masked 14-day CGM wear at the end of the study for the control group. The control and intervention groups had a total of 25 (10 T1DM, 15 T2DM) and 53 (25 T1DM, 28 T2DM) participants, respectively.
Glycemic Risk Plot
The glycemic risk plot shows the relationship between estimated HbA1c % (eA1c) and minutes below 70 mg/dl (MB70) with superimposed constant-variability contours. While several measures of variability have been reported , variability here is defined as the difference between the median and 10th percentile of glucose values. A gamma distribution is obtained as per  for a range of glucose median and variability values. Dunn et al. used data from the JDRF-CGM trial to generate the gamma distribution parameters for various combinations of median and variability values . The contours were then calculated by holding the variability constant and determining the gamma distribution at various values of median glucose using these parameters. Once the glucose distribution is obtained for a given variability value, the MB70 values can be calculated by multiplying the number of minutes in a day by the probability of glucose < 70 mg/dl defined by the distribution. The median goal is then converted to eA1c using the equation derived from . A single contour for each variability value is then generated by plotting the corresponding eA1c and MB70 pairs. An example of the plot is shown below in Fig. 1.
The plot is divided into four zones, as shown in Fig. 1:
Hyper only: eA1c above 7% with MB70 below 84 min.
In target: eA1c below 7% with MB70 below 84 min.
Hyper and hypo: eA1c above 7% with MB70 above 84 min.
Hypo only: eA1c below 7% with MB70 above 84 min.
The gray contours illustrate how hypoglycemia risk varies with hyperglycemia risk if variability is held constant. The hypoglycemia and hyperglycemia metrics chosen here align with those reported by . A similar plot can be generated with different measures for hypoglycemia risk, hyperglycemia risk, and glycemic variability without changing the underlying relationship.
Each vector on the plot represents data from a single individual with the arrow end representing post-intervention risk and the tail end representing the pre-intervention risk.
Generally, movement towards the in target zone indicates improvement in hyperglycemia and hypoglycemia risk.
The sample vectors shown in Fig. 1 illustrate how glycemic risks are dependent on variability. Examples 2 and 3 show that if variability is not improved, individuals are likely to trade-off one glycemic risk for the other. Example 1 shows that if variability is improved, it is possible to achieve improvement in hypoglycemia and hyperglycemia risk, or at least achieve improvements in one risk while maintaining the other constant. The plot also provides other valuable insights; for example, as the individual moves closer to the in target zone, a larger change in variability is needed to not increase the risk of hypoglycemia as is illustrated by the distance between the contours, which increases near the in target zone.
To verify the predictions of the glycemic risk plot, data were analyzed from a randomized control trial for evaluation of the Navigator QS Continuous Glucose Monitor combined with prototype informatics software used for clinical visits . As described above, the study enrolled individuals with type 1 and type 2 diabetes who were on insulin therapy. Data from the intervention and control groups were pooled for this analysis, and comparisons were made between the baseline and final periods of the study.
The data were analyzed by using two methods: (1) the glycemic risk plot and (2) modified glycemic risk plot. The modified glycemic risk plot (Fig. 1) shows change in MB70 and eA1c between two study periods using a single plot marker. Each point on the plot is one individual, and the size of the plot marker is proportional to the variability change between the study periods, with the color magenta indicating a variability increase and cyan indicating a variability decrease. The data are stratified (annotated using zone numbers) by the starting glycemic condition for any individual based on the zones shown in Fig. 1.
Compliance with Ethics Guidelines
This article does not contain any new studies with human or animal subjects performed by any of the authors.