A multisite physician's office laboratory evaluation of an immunological method for the measurement of HbA1c

To evaluate the clinical performance of a new immunological HbA1c method in physicians' office laboratories.

Three physicians' offices participated in the evaluations. The clinicians routinely use HbA1c test results to monitor their patients' long-term blood glucose control. Precision and interlaboratory variability were assessed using three levels of lyophilized controls. Correlation of the method's results to currently available laboratory methods was made. Comparison of finger-stick (capillary) results to venous EDTA whole blood results was made on 134 patients. Physician and laboratory personnel input was evaluated with regard to the clinical utility of the system.

The CVW and CVB were a maximum of 4.5 and 4.4% for the immunoassay system on three levels of control materials at the three sites. Interlaboratory variability among the control means was found to be 4.9-5.4, 8.0-8.3, and 11.7-12.0% HbA1c. Correlation coefficients (r) ranged from 0.95 to 0.99. There was a positive bias by the DCA 2000 compared with the in-house method at site 1. Minimal negative biases were seen by the DCA 2000 with comparative methods used at sites 2 and 3. Median percentage differences with the comparative methods were 12, -1.4, and -5.6%. Comparison of capillary to venous sample results, from the DCA 2000, showed no clinically significant differences. Operator and physician feedback were positive with respect to technical ease in performance of the test and accuracy of results.

Precision was acceptable and interlaboratory variability was low. The immunological method correlated well with manual ion-exchange and automated HPLC methods. The small sample size and good comparison between capillary and venous sample results make fingerstick sampling acceptable. The method provided immediate test results (within 9 min) to the clinicians.