Agreement of Point-of-Care Capillary Glycated Hemoglobin Levels with Conventional Screening Tests for Diabetes Mellitus in a Canadian First Nations Population.

Abstract

1) How closely do capillary glycated hemoglobin (A1C) levels agree with venous A1C levels? 2) How well do venous A1C levels agree with plasma glucose for diagnosis of diabetes in this population? The Seabird Island mobile diabetes clinic screened people not known to have diabetes by using finger-prick capillary A1C levels with point-of-care analysis according to the Siemens/Bayer DCA 2000 system. Clients then went to a clinical laboratory for confirmatory testing for venous A1C levels, fasting plasma glucose (FPG) and plasma glucose 2 hours after 75 g oral glucose load (2hPG). A reference laboratory compared the DCA 2000 and the clinical laboratory's Roche Integra 800CTS system to the National Glycohemoglobin Standardization Program Diabetes Control and Complications Trial (DCCT) reference. 1) In the reference laboratory, DCA 2000 and Integra 800CTS both agreed very closely with the DCCT standard. In the field, capillary glycated hemoglobin percent (A1C) % was biased, underestimating venous A1C % by a mean of 0.19 (p<0.001). The margin of error of bias-adjusted capillary A1C % was ±0.36 for 95% of the time, compared to ±0.27 for venous A1C%. 2) By linear regression, we found FPG 7.0 mmol/L and 2hPG 11.1 mmol/L predicted mean venous A1C levels very close to 6.5%, with no significant bias. Point-of-care capillary A1C did not perform as well in the field as in the laboratory, but the bias is correctible, and the margin of error is small enough that the test is clinically useful. In this population, venous A1C levels ≥6.5% agree closely with the FPG and 2hPG thresholds to diagnose diabetes; ethnic-specific adjustment of the venous A1C threshold is not necessary.