Ionized calcium measurements are influenced by albumin--should ionized calcium be corrected?

Measurement of ionized calcium (CaI) has been reported to be dependent on albumin concentration. We examined the correlation between albumin and CaI measured on different ion selective electrode analyzers and in different groups of patients in a large dataset, extracted from the laboratory information system. In 17,281 outpatients and 16,194 inpatients, significantly positive correlations were found between CaI and albumin, with changes in CaI per 10 g/L change in albumin ranging from 0.007-0.043 mmol/L and 0.017-0.028 mmol/L, respectively. Correlations were found to be significantly different when using different analyzers. In order to examine whether the difference in correlations between the analyzers were really due to different patient populations investigated on the different analyzers, data analyzed on the same type of analyzer from inpatients from four different wards (intensive care unit, medical ward, surgical ward and orthopedic ward) were examined. There was no significant difference in correlations between patients from the four wards. Although, these results points towards technical causes behind the observed differences it cannot be entirely ruled out that clinical diseases or treatment might influence albumin interaction with CaI measurements. Combining all data from both out- and inpatients, a correction formula using a change in CaI of 0.03 mmol/L per 10 g/L change in albumin, was constructed. However, the albumin influence on CaI is only a minor part of the total CaI variation and, in most situations, the relatively small effect of changes in albumin on CaI-results is most likely of no clinical importance.