The Libre has revealed some surprising results. On the Lantus, the 10th-90th percentile breadth of readings is significantly wider than on Levemir (there are more higher and lower readings, and the higher ones are significantly higher.
From what I can see from the data, I seem to have less variance and more consistency across the day. This is the set of results on Lantus:
And this is the set for the last month of Levemir:
For me what is interesting is the bandwidth of the results has clearly narrowed, and early morning Hypos have reduced. There is a noticeable increase compared to Lantus at the 10am timeslot, and this is as a result of the overnight Basal running out with Levemir, which I’ve clearly not fully tamed yet. Bringing that in line will be hugely helpful.
The other thing of note is the shape of the 10th-90th percentile band versus the median. With the Levemir, the 10th to 90th (10/90) clearly tracks the median, whereas with the Lantus, there is a noticeable difference the pattern. This might relate to the variable absorption I was seeing with the Lantus.
The other thing I note is that the Libre is now estimating a normal Hba1C, which I’d be very pleased with. Just need to get the variance under control…