Development and Evaluation of a Single Site device for conjoined Glucose Sensing and Insulin Infusion in Diabetes Patients

Goal: Diabetes patients are increasingly using a continuous glucose sensor to monitor blood glucose and an insulin pump connected to an infusion cannula to administer insulin. Applying these devices requires two separate insertion sites, one for the sensor and one for the cannula. Integrating sensor with cannula to perform glucose sensing and insulin infusion through a single insertion site would significantly simplify and improve diabetes treatment by reducing the overall system size and the number of necessary needle pricks. Presently, several research groups are pursuing the development of combined glucose sensing and insulin infusion devices, termed single-port devices, by integrating sensing and infusion technologies created from scratch. Methods: Instead of creating the device from scratch, we utilized already existing technologies and introduced three design concepts of integrating commercial glucose sensors and infusion cannulas. We prototyped and evaluated each concept according to design simplicity, ease of insertion, and sensing accuracy. The best single-port prototype was then used in two clinical trials. The first trial was performed to inform final refinements in device assembly, insertion, and sensor operation techniques. The second trial was performed to ascertain the accuracy of the glucose sensing with the single-port device and to assess the feasibility of using the single-port device in combination with an algorithm to automatically control the blood glucose in diabetes patients. Results: We found that the best single-port device is the one in which a Dexcom sensor is housed inside a Medtronic cannula so that its glucose sensitive part protrudes from the cannula tip. Glucose sensing performed with this single-port device was found to be accurate and reliable - the average mean absolute relative deviation from blood glucose concentrations obtained for the sensor of the device was low (median, 13.0%; interquartile range, 10.5–16.7%; n= 10) and did not differ from that of the additionally worn glucose sensor (versus 13.9%; 11.9–15.3%; P= 0.922). Furthermore, insulin delivery with the single-port device was reliable and safe during home use and, when performed in combination with the control algorithm, was adequate to achieve and maintain near normoglycemia. Conclusion: Results from these studies indicate the feasibility of combining commercial glucose sensing and insulin delivery technologies to realize a functional single-port device. Significance: Our development approach may be generally useful to provide patients with innovative medical devices faster and at reduced costs.