Experience from Implementing Telemetric In-Hospital Continuous Glucose Monitoring During the COVID-19 Pandemic

Diabetes increases the severity of the Coronavirus disease COVID-19 and hyperglycemia at admission for COVID-19 is associated with worse outcome. ¹ Klonoff et al ² showed that treatment of hyperglycemia in COVID-19 admission is associated with improved outcomes. To mornitor glucose levels closely but minimize contact with quarantined COVID-19 patients, some hospitals—including ours—are using in-hospital Continues Glucose Monitors (CGMs).

We write this letter to help clinicians better implement in-hospital CGM. We are currently conducting a RCT on telemetric in-hospital CGM in quarantined patients with COVID-19 infection or other infections at Nordsjællands Hospital, Denmark (604 beds). Participants are randomized to either standard care finger-prick + blinded CGM or only CGM (Dexcom G6), ³ Our primary endpoint is time in range for glucose. CGM data are transmitted to a smartphone in the patient’s room and then to a tablet in the nurse’s office. To date 87 patients have enrolled.

Occasionally the CGM fails to complete the “warmup” period of 2 hours for unknown reasons. It is not always possible to troubleshoot which part of the system is failing. A support team should—at least around the implementation—be available, preferably 24/7. We have mounted new sensors, but then the monitor that did not work. This is time consuming and expensive. Sensor-life may be sorter than the expected 10 days.

We have experienced repetitive errors in the CGM-to-smartphone-to-tablet transmission. This may be due to too far distance between the sensor and the smartphone or temporary problems with intranet coverage. Occasionally the signal is unsteady even within 6 meters of the patient if the smartphone is put into a drawer or closet. App updates and “crashes” in the Dexcom follow app have occurred. We have also experienced patients who—became delirious and demounted their CGM, primarily at night.

Some healthcare workers (HCWs) prefer to measure capillary glucose values by finger-prick method because they don’t trust CGM technology. HCWs need to know that (1) There is a time delay from BG to ISF values, ⁴ and (2) accuracy can vary between different glucose meters and CGMs.

Hospitalized patients frequently move from one room to another and risk losing their smartphones. Furthermore, a Dexcom G6 is not approved for MR/CT scans and must be removed before scanning. The monitor can be reused on the same patient with a new sensor. Despite oral and written instructions and alerts about this, we have experienced four patients going into CT scan wearing their CGM who then needed a replacement (two because of malfunction and two because of a risk of disturbed data.)

The cost of a CGM system far exceeds that of finger-prick equipment which potentially limits the use. Moreover, with an average duration of hospital stay of only 2.4 days (at our hospital) sensors may be wasted if patients are discharged within hours of admission.

Therefore, implementing in-hospital CGM faces many challenges and the current CGM-setup seems immature for in-hospital use. ⁵ In the future, hospital based CGM systems may be implemented with an endocrinologist or a diabetes team to optimize diabetes care, thereby potentially improving outcomes.