Evaluation of a noninvasive blood glucose monitoring device for critically ill patients

Introduction

Critically ill patients frequently experience abnormalities in carbohydrate metabolism and a severe insulin resistance state. Hyperglycemia is a negative predictor of outcome in these patients, as high blood glucose (BG) values are associated with an increased risk of morbidity and mortality. Currently, BG monitoring methods cannot fulfill the need for continuous glucose monitoring in order to safely implement tight glucose control protocols. The purpose of this study is to evaluate the feasibility of the NBM-100 device (OrSense Ltd) for non-invasive (NI) continuous glucose monitoring in critically ill patients.

Materials and methods

The NBM-100 uses a finger-based sensor shaped as a ring, located at the finger's base. Red/near-infrared occlusion spectroscopy detects and analyzes BG and hemoglobin concentrations. The NBM-100 utilizes an enhanced optical signal resulting from a temporary over-systolic occlusion, produced by a finger-based pneumatic cuff. The resulting changes in the optical signal create the sensitivity needed for measuring the glucose concentrations non-invasively.

A study was conducted on six patients (three female, three male, ages 44–88 years) in the ICU of Rabin Medical Center upon receipt of informed consent. The probe of the NBM-100 was placed on the patients' thumb, where it performed NI continuous measurements for up to 24 hours, with readings every 10–15 min. The results obtained from the NBM-100 device were compared with arterial blood samples taken through an arterial line every 30–60 min and analyzed with a blood gas machine (ABL 700; Radiometer, Copenhagen, Denmark).

Results

A prospective analysis based on a uniform model with personal parametric adjustments was performed on the NBM-100 readings, for a total of 80 data points. The calibration phase lasted 1.5–2.5 hours. The reference BG range was 64–247 mg/dl. The resulting median relative absolute error was 11.5%, and the median absolute error was 18 mg/dl. A Clarke error grid analysis showed that 100% of the measurements fell within zones A (79%) and B (21%). In all these clinical settings there was good patient compliance and no adverse effects were identified. One of the trials did not meet a self-consistency criterion for the calibration, and was excluded.

Conclusion

This study indicates the potential use of the non-invasive NBM-100 as a device for continual, accurate, safe, and easy-to-use BG evaluation for ICU. Consequently, it will improve patient care and survival, as well as reducing staff workload. It has the promise for trend analysis, hypoglycemia detection and closed-loop systems enabling automatic glycemic control.