To determine whether the Primus high-pressure liquid chromatography (HPLC) is suited to alternate site testing (AST) for HbA1c in a hospital diabetes outpatient clinic.
Patients were attending the clinic for routine management of their diabetes. A number of diabetic patients with uremia (n = 11) were also investigated. HbA1c levels were measured in the outpatient setting by the Primus HPLC and in a more limited study the DCA-2000 instrument using the new 6-min assay cartridge. HbA1c measurements were also performed with Pierce affinity minicolumns and a Bio-Rad Variant HPLC.
The Primus HPLC assay had low imprecision of 2.3, 1.6, and 1.0% for HbA1c levels of 4.7, 7.3, and 11.1%, respectively, and was not prone to interference by carbamylated hemoglobin as found for the ion-exchange Variant HPLC method. Method comparison studies showed that the bias and proportional error between the Pierce affinity minicolumn procedure (standardized with respect to an external quality control program) and the Primus HPLC (Y) was −0.4 and 1.2% respectively (n = 32). Similarly the bias and proportional error between the Primus and DCA-2000 methods was 0.7 and −2.5%. The Primus was shown to give falsely elevated HbA1c concentrations if the time between sequential injections was > 28 min.
The Primus HPLC has a decided advantage over specialty AST instruments, like the DCA-2000, in not only meeting AST requirements but also allowing rapid automated batch processing of all laboratory HbA1c samples.