Clinical Evaluation of a Photonic Crystal Sensor for Glucose Monitoring in Urine

Self-monitoring of blood glucose is not only an effective method to treat diabetes, but also a painful invasive process to patients. To develop a noninvasive technique, we evaluated the clinical performance of a polymerized crystalline colloidal array (PCCA) sensor for glucose monitoring in urine. A two-dimensional crystalline colloidal array (CCA) was assembled using polystyrene particles and further embedded into 3-acrylamidophenylboronic acid functionalized hydrogel. 101 urine samples were tested, and the glucose concentrations was calculated by the particle spacing of PCCA. All the urine samples were also analyzed by automated urine analyzers. Pearson correlation and linear regression analyses showed a linear relationship between the PCCA results with a range of 0.4-53.3 mmol L⁻¹ and the Cobas® 8000 references (r=0.914; [PCCA]=0.9281 [HK] + 2.0585 mmol L⁻¹). Cohen's kappa (κ) statistic with 0.65 (sig < 0.05) showed the PCCA results had moderate agreement with references. Also, Bland-Altman plot showed that AX-4030 sensor's systematic error was higher than PCCA sensor (mean=7.4813 > 0.5271 mmol L⁻¹). The PCCA urine glucose biosensor has the advantages of fast fabrication, low cost, easy operation and high sensitivity, which provides a promising technique for noninvasive point-of-care glucose monitoring.