Severe Hypoglycemia Before Landing in a Patient with Type 1 Diabetes Using Sensor-Augmented Pump Therapy

Dear Sir,

Recently, Bertuzzi F et al. published an important article in Diabetes Technology and Therapeutics, which investigated the unintentional delivery of insulin from insulin pumps because of changes in atmospheric pressure. ¹ Using new models of insulin pumps under experimental hyperbaric conditions, the authors showed that pressure variations affected insulin delivery. More specifically, an unintentional increase in insulin delivery occurred with a fall in pressure, and vice versa. This phenomenon appears to be principally related to the change in air bubble volume inside the pump tubing and cartridge induced by pressure modifications, according to Boyle's law. In real life, this phenomenon can occur in situations such as when diving, or during cabin depressurization when taking off for a flight. Whether these abnormalities in insulin delivery may result in clinically relevant issues depends on factors such as insulin sensitivity, current metabolic control, pump settings, previous food intake, and individual skills (among others). However, it is worth advising patients to pay particular attention to their condition in these potentially harmful circumstances.

We would like to report the case of a 56-year-old man with type 1 diabetes (T1D) managed with sensor-augmented pump (SAP) therapy. The subject in question experienced a severe hypoglycemia episode while on a commercial flight, 50 min after the takeoff (5 min before landing). Our patient had been started on SAP therapy with predictive low-glucose suspend advanced feature (Medtronic 640G, Northridge, CA) 10 months previously, following episodes of recurrent, severe hypoglycemia despite use of continuous subcutaneous insulin infusion. Since starting SAP therapy, no new severe hypoglycemia episodes had occurred.

Using specific software, we downloaded information from the pump and sensor, including for the day when severe hypoglycemia occurred. The total insulin dose administered was 26 U/day, 45% of which was basal. The patient performed 4.6 self-monitored capillary blood glucose tests per day and used the bolus advisor in 80% of bolus injections. According to sensor data, the patient spent 11% of the time with glucose values <70 mg/dL and 15% of time with values >180 mg/dL. His most recent HbA_1c reading was 6.5%. With a low glucose threshold of 65 mg/dL, our patient had three suspend before low events per day on average. Before taking off, the patient had administered two manual boluses based on sensor glucose, the first being 1.5 h before takeoff with food: 3.1 units, 30 g of carbohydrates with preceeding glucose sensor values of 266, 267, 270, 270, 259, and 213 mg/dL (every 5 min) and the second bolus being 15 min before takeoff to correct hyperglycemia: 2.1 units, unabsorbed insulin at start of delivery: 2.2 units, with preceeding glucose sensor values of 254, 254, 252, 253, 254, and 259 mg/dL. At that time, the basal rate was 0.6 U/h. Despite activation of the suspend alarm before low feature and low glucose suspend alarm, severe hypoglycemia that required outside help occurred ∼50 min after takeoff. The episode was resolved by oral glucose administration. A comprehensive medical evaluation performed in an emergency room after landing did not find any clinical abnormalities.

Although in this particular case it is not possible to demonstrate that the severe hypoglycemia episode was related directly and exclusively to a transient increase in insulin delivery resulting from ambient pressure changes during the flight, its potential contribution cannot be ruled out. More probably, hypoglycemia was the summative outcome of multiple circumstances, including the administration of two very close manual boluses without taking into account the presence of unabsorbed active insulin. Based on the characteristics of our patient, a higher threshold for low glucose and less tight glycemic control would have been more convenient for this particular case. However, our patient had low concern about hypoglycemia unawareness and its associated risks and low motivation to regain awareness as previously described. ² The management of this subgroup of patients is highly challenging. The device and treatment regimen used by our patient has some potentially clinical implications. This case suggests that the unintentional insulin delivery associated with atmospheric pressure changes may not be prevented by SAP therapy with predictive suspension of insulin delivery driven by low sensor glucose.

Having been used for nearly 40 years, there is no doubt that continuous subcutaneous insulin infusion therapy is an efficient, safe, and flexible treatment for improving both glycemic control and quality of life of patients with T1D. ^3,4 We hope that this assertion will extend to the use of more advanced therapies in the near future, including SAP therapy. However, healthcare professionals should advise their patients to pay particular attention to the possible interference of everyday circumstances in diabetes management, including atmospheric pressure changes during flights. Such recommendations should always be included in educational programs.