Running Diabetes: The Guardian versus the Enlite sensors: Test #2

aaa

Alright, here is the second part of my sensor experiment. You can find the first part of the test hereFor this second test, I decided to record some additional information regarding the calibrations as well as, like the previous post, taking more pictures of the trends.

One thing to point out is that for this second try you will notice some pink dots below the hour ticker numbers with the Guardian display. At this time I am using the auto mode of the pump, the Basals are adjusted on an ongoing basis depending on the reading and the trend. I am not going to go on details about what it is or what it does or how it works, there are many blogs and websites that talk about it, here is the official website from the manufacturer (Medtronic). But, it was interesting for me to notice, the gaps in basal delivery when the CGMS reading is low sugars or the trend is going down.
As per the initial comparison, both sensors capture pretty well the trending of the sugar during the timelines. There were some occasions in which the readings were off, of the graph were shifted in comparison with one an another. But when I looked at the reading versus the calibration from my meter, I can spot a few additional differences between the two.

Guardian, calibration, glucose readings

Enlite, Guardian, Diabetes

aaa

There were a few times in which the calibration readings versus the CGMS reading was off by a (What I consider) significant amount, given that tolerances for meter and the CGMS are within the +/- 15%.
  • Calibrations 2,3,8,15,20 for Guardian sensor, for a 75% of good reading within the tolerance. And Calibrations 9, 10, 11, 12, 13, 15, 17, 19, 20 for 50% of readings at calibration within the tolerance in the case of the Enlite sensor.
  • The Enlite sensor got worse in comparison to the calibration as the days went by. While the opposite happened with the Guardian, which after the 4th calibration it stayed pretty much within all the way until the last calibration or end of life of the sensor.
  • There were 5 occasions in which the sensors were on opposite sides of the calibration. Meaning that when the Guardian was registering lower than calibration, the Enlite sensor was showing up higher. This was very noticeable in 3 of the 5, calibration 2, 10 and 13. Calibration number 13 is one that it will call for a different call of actions depending on which reading you take as a true reading.
  • The average deviation difference between the two is about 16.34 higher for the Enlite sensor.
  • While the maximum deviation from calibration reading was also higher for the Enlite sensor by 82 points. With a standard deviation difference of around 21 higher for the Enlite, making it a little bit more spread in the reading compared to the calibration readings.

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

Gardian, Enlite, Medtronic, CGMS

aaa

In conclusion, both sensors performed pretty well during the two tests. The trending information and graphing behavior were very similar and enough to make the right calls with adjustments (With the exception mentioned above in one instance). The Guardian has a slightly edge, being closer in accuracy with the calibration readings and reduced variation in standard deviation. If you are using the Enlite, it is great sensor with good reading and no need to update or change to the new one, but if you are in the process of selecting a CGMS and have the option or replacing the one you are using due to insurance, it will be wise to consider the move to the newer technology of the Guardian CGMS and make it a true improvement with auto mode with the insulin pump.

Inserted Guardian 3 and Enlite sensors from Medtronic