LAUSR

AIM To describe quality management processes and appropriate interpretation with respect to HbA1c point-of-care (POC) testing in a national diabetes and cardiovascular risk screening programme. METHODS We compared HbA1c results from capillary blood, measured by the cobas b 101 (Roche Diagnostics) POC testing system, with results from venous blood measured by accredited laboratory analysers to inform national screening practice and a (separately-reported) randomised controlled trial. Difference plots and regressions were used to aid interpretation around 40 and 50mmol/mol, the cut-offs used to identify "pre-diabetes" and diabetes in New Zealand. RESULTS After initial acceptable tests, subsequent batches delivered POC results that varied from laboratory HbA1c by +6 to -14mmol/mol around the clinical cut-offs. Ten faulty batches of discs were recalled worldwide. POC testing was suspended in one region, as was the planned trial. The manufacturing defect was rectified, accuracy of the new batches was confirmed, and testing resumed. CONCLUSION POC testing must be conducted within stringent quality assurance processes prior to and while in use. Within such a system, POC testing for HbA1c can be sufficiently accurate for screening and diagnosis of diabetes.