Accuracy and Precision of a Point-of-Care HbA1c Test

Abstract

Background: Point-of-care (POC) hemoglobin A1c (HbA1c) testing has advantages over laboratory testing, but some questions have remained regarding the accuracy and precision of these methods. The accuracy and the precision of the POC Afinion™ HbA1c Dx test were investigated. Methods: Samples spanning the assay range were collected from prospectively enrolled subjects at three clinical sites. The accuracy of the POC test using fingerstick and venous whole blood samples was estimated via correlation and bias with respect to values obtained by an NGSP secondary reference laboratory (SRL). The precision of the POC test using fingerstick samples was estimated from duplicate results by calculating the coefficient of variation (CV) and standard deviation (SD), and separated into its components using analysis of variance (ANOVA). The precision of the POC test using venous blood was evaluated from samples run in four replicates on each of three test cartridge lots, twice per day for 10 consecutive days. The SD and CV by study site and overall were calculated. Results: Across the assay range, POC test results from fingerstick and venous whole blood samples were highly correlated with results from the NGSP SRL ( r = .99). The mean bias was −0.021% HbA1c (−0.346% relative) using fingerstick samples and −0.005% HbA1c (−0.093% relative) using venous samples. Imprecision ranged from 0.62% to 1.93% CV for fingerstick samples and 1.11% to 1.69% CV for venous samples. Conclusions: The results indicate that the POC test evaluated here is accurate and precise using both fingerstick and venous whole blood.

Fingerstick Precision and Total Error of a Point-of-Care HbA1c Test

“…Table 3 describes the repeatability and betweenlot components of variance that were calculated using prior duplicate fingerstick measurements in Study A. 5 In Study C, four venous whole blood samples, representing low, threshold, medium, and high HbA1c levels, were tested on the POC test over 20 days. A grand total of 720 measurements for each HbA1c level resulted from testing these samples on each of three cartridge lots and each of three instruments with two runs per day and two measurements per run.…”

“…Table 3 describes the repeatability and between-lot components of variance that were calculated using prior duplicate fingerstick measurements in Study A. 5…”

“…The NGSP awards certification to manufacturers for successfully meeting performance criteria for bias testing. 9 In Study A, conducted previously, 5 a first-replicate POC test value (from fingerstick and venous whole blood samples) was compared against the average of duplicate results from a venous sample tested on the Tosoh glycohemoglobin test on the G8 HPLC analyzer at the NGSP Secondary Reference Laboratory (SRL) at the Diabetes Diagnosis Laboratory, University of Missouri School of Medicine, Columbia MO. Passing-Bablok regression equations were calculated and used to estimate the bias of the POC test at HbA1c values of 5.0% (31.1 mmol/mol), 6.5% (47.5 mmol/mol), 8.0% (63.9 mmol/mol), and 12% (107.7 mmol/mol).…”

“…1 The results of a previous investigation of a POC method, the Afinion™ HbA1c Dx test (Study A) first suggested that this method may offer acceptable accuracy and precision when used by trained laboratory operators. 5 After the completion of Study A, there was interest in evaluating additional components of variance not assessed in that study design. In the study reported here (Study B), a direct evaluation of repeatability, between-operator, and betweeninstrument precision is made using fingerstick (capillary) whole blood samples, the principal sample type used at the POC.…”

“…The results of a previous investigation of a POC method, the Afinion™ HbA1c Dx test (Study A) first suggested that this method may offer acceptable accuracy and precision when used by trained laboratory operators. 5…”