Validation of an adjusted calcium formula using the Roche calcium (NMBATPA) and albumin (BCG) methods at Groote Schuur Hospital

Master Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title

University of Cape Town

ABSTRACT: Validation of an adjusted calcium formula using the Roche calcium (NMBATPA) and albumin (BCG) methods at Groote Schuur Hospital Introduction: Most laboratories continue to adjust serum total calcium (tCa) concentration for serum albumin as a surrogate marker of calcium status, despite the availability of ionised calcium (iCa) measurement. Current recommendations by the Association for Clinical Biochemistry and Laboratory Medicine (ACB) advocate that laboratories should use formulae specific for their tCa and albumin methods and analytical platforms. The National Health Laboratory Service at Groote Schuur Hospital (GSH) undertook to investigate this recommendation. An adjusted calcium (aCa) formula specific for the Roche serum tCa and albumin methods was derived from 3131 patients. We investigated the validity and clinical utility of this locally derived aCa formula. Methods: The tCa, albumin and iCa were analysed in blood from 162 inpatients and outpatients at GSH. Corrected calcium (cCa) was calculated using the Payne cCa formula, and aCa was calculated with the new aCa formula. Patients were classified as hypo-, normoor hypercalcaemic using iCa, tCa, cCa and aCa measurements. Cohen's kappa statistic, loglinear and logistic regression models and interclass and concordance correlation coefficients were used to assess agreement between tCa, Payne cCa and aCa against iCa (gold standard). Agreement was further assessed according to renal status and albumin concentrations. Results: The aCa demonstrated good correlation with iCa, but its performance was not significantly better than tCa or Payne cCa in correctly classifying calcium status. Furthermore, albumin concentration was demonstrated to predict the performance of the calcium status classification by the aCa and cCa formulae, irrespective of renal status. Conclusion: The laboratory-specific aCa formula did not perform significantly better than tCa and the Payne cCa formula. This implies that aCa does not add value over tCa where iCa measurements are not readily available.