Use of a reflectance photometer as a diabetes mellitus screening tool under field conditions

Investigate the Accu-Chek II and visual Chemstrip bG methods for non-insulin dependent diabetes mellitus (NIDDM) screening under field conditions with the aid of local providers.

Compare simultaneous random Chemstrip bG, Accu-Chek II and plasma glucose levels using a 6.7 mM cut-off, followed by a 75 gram 2 hour oral glucose tolerance test while screening 1124 people in 15 Alaskan villages with the aid of local personnel during winter field conditions.

The Accu-Chek II and visual Chemstrip bG methods detected 92% and 96%, respectively, of those newly diagnosed with NIDDM whose plasma screen was > or = 6.7 mM. The Accu-Chek II provided correlations of 0.774, 0.850, 0.923, and 0.946 with random, fasting, 1 and 2 hour plasma glucose. The mean Accu-Chek II, Chemstrip bG and plasma levels at which NIDDM was diagnosed were not statistically different. Receiver operator characteristic (ROC) curve analysis revealed optimum screening cut-off values for subsequent new diagnosis of NIDDM to be 8.3, 7.7, and 8.0 mM for visual, Accu-Chek II and hexokinase methods (p = NS).

The Accu-Chek II and Chemstrip bG provided accurate glucose data in random NIDDM screening under field conditions with multiple local providers.