Albumin and calcium reference interval using healthy individuals and a data-mining approach.

BACKGROUND: Harmonization of reference intervals for analytes that have a sound calibration and metrological traceability is a widely recommended practice. The UK Pathology Harmony has recently harmonized reference intervals for calcium and albumin. In this study, we have determined the reference intervals for calcium and albumin on the UK's most commonly used analytical platforms. METHOD: A prospective reference population of healthy individuals was recruited according to the IFCC CRIDL criteria. A second indirect population was collected from 14 primary care setting and measured in laboratories using various analytical platforms and methods (Roche, Abbott, Beckman and Siemens analytical platforms). RESULTS: In total, 299 subjects were recruited; the central 95th centile values for calcium for three out of four analytical platforms were in a close agreement with UK Pathology Harmony reference intervals of 2.2-2.6 mmol/L. Reference intervals of BCG methods from both cohorts and irrespective of analytical platforms were higher for both lower and upper reference limits than those for BCP. In comparison, the indirect study showed an age-related variation. The younger population reference intervals varied by up to 5.7% at the lower reference limit and up to 12% at the upper reference limit compared with Pathology Harmony reference intervals, and the older population showed a variation of up to 14% at both limits. CONCLUSION: While calcium reference intervals can be a subject for harmonization, albumin reference intervals studied showed large variation which is unsupportive of embracing a common reference interval for albumin.

The effect of different analytical platforms and methods on the performance of population-specific adjusted calcium equation.

BACKGROUND: A recent attempt to improve the diagnostic value of adjusted calcium addressed a primary care-specific adjusted calcium equation, but validated the new equation for Roche Cobas, BCG and NM-BAPTA methods only. In this study, we aim to validate a population-specific equation for other methods and platforms. METHOD: We collected retrospective patient data-sets from 15 hospital laboratories using a range of commercially available analytical platforms and methods for calcium and albumin measurements. Raw data-sets were collected and filtered according to Payne's criteria, and separate adjusted calcium equations were derived for hospitalized and primary care patients. RESULTS: Mean albumin and calcium results were significantly higher in primary care populations (P < 0.0001). The prevalence of hypocalcaemia using adjusted calcium ranged between 6% and 44% for inpatient data-sets and was higher in users of BCG methods. The application of community-specific adjustment equation to primary care data-sets reduced the prevalence of hypocalcaemia (mean 1.7%, range 0.8-3.7%). CONCLUSION: We demonstrated that the use of a community-specific calcium adjustment equation to a primary care population reduces both the percentage and the variation of hypocalcaemia between different laboratories.