References 1 Lai JKC, Lucas RM, Banks E, Ponsonby AL, Ausimmune Investigator Group. Variability in vitamin D assays impairs clinical assessment of vitamin D status. Intern Med J 2012; 42: 43–50. 2 Carter GD. Accuracy of 25-hydroxyvitamin D assays: confronting the issues. Curr Drug Targets 2011; 12: 19–28. 3 Stepman HCM, Vanderroost A, Stöckl D, Thienpont LM. Full-scan mass spectral evidence for 3-epi-25-hydroxyvitamin D3 in serum of infants and adults. Clin Chem Lab Med 2011; 49: 253–6. 4 Shah I, James R, Barker J, Petroczi A and Naughton DP. Misleading measures in vitamin D analysis: a novel LC-MS/MS assay to account for epimers and isobars. Nutr J 2011; 10: 46–54. 5 Binkley N, Krueger DC, Morgan S, Wiebe D. Current status of clinical 25-hydroxyvitamin D measurement: an assessment of between laboratory agreement. Clin Chim Acta 2010; 411: 1976–82. Variability in vitamin D assays impairs clinical assessment of vitamin D status We read with interest the article by Lai et al. 1 that reported variability in serum 25-hydroxy-vitamin D (25(OH)D) results analysed using different methods. The authors reported that compared with the liquid chromatography-tandem mass spectrometry (LCMS/MS) as a gold standard, the Diasorin Liaison method overes- timated substantially the percentage of people with vitamin D deficiency. However, the authors failed to report the performance of their LCMS/MS in any exter- nal quality assurance program (EQA). While 25(OH)D results from immunoassays could be variable, results from LCMS/MS users could also be vastly different. In the last Australian-based EQA (Royal College of Pathologists of Australasia Chemical Pathology Quality Assurance Program), for a target 25(OH)D of 43 nmol/L (sample 36-12), laboratories using mass spectrometry reported results from <27 to 43 nmol/L. The LCMS/MS method- ology is technically demanding, and currently, the cali- brators are not standardised universally, although the recent introduction of the vitamin D standard reference material (SRM 972 and SRM 2972) by the National Insti- tute of Standards and Technology (NIST) and the avail- ability of EQA have reduced interlaboratory variability in 25(OH)D measurements. In addition, while LCMS/MS is probably the most important advance in 25(OH)D methodology, the LCMS/MS methods for 25(OH)D measurement are not perfect yet. It has recently been revealed that LCMS/ MS methods can be interfered by 25(OH)D metabolites, such as C-3 epimer of 25-hydroxy vitamin D [3-epi- 25(OH)D], 2–5 and the biological significance of the metabolites remains unclear. In the October 2011 cycle of the UK-based EQA (the Vitamin D External Quality Assessment Scheme), laboratories using LCMS/MS, the proposed ‘gold standard’ for 25(OH)D analysis overestimated substantially the 25(OH)D concentration in a specimen containing 3-epi-25(OH)D, while most laboratories using immunoassays did not. For sample 405 that contained approximately equal amounts of 25(OH)D 3 and 3-epi-25(OH)D3 (51 nmol/L each), 99% (109/115) of the LCMS/MS users failed to separate out the 3-epimer that coelutes with 25(OH)D3 and resulted a much higher method median of 109 nmol/L as 25(OH)D3. This may be one reason behind the positive bias in some LCMS/MS methods compared with the NIST reference method 5 that is able to separate the 3-epimer. In the study by Lai et al., 1 it was unclear whether their LCMS/MS method was able to separate 3-epi-25(OH)D3, as this was not discussed by the authors. This is important because 3-epi-25(OH)D, thought to be only in neonates, 2 is now commonly found in adults. 3,4 Lensmeyer et al. 4 showed that 212 of 214 specimens from people age ranging from neonate to over 80 years had detectable 3-epi-25(OH)D3. At 25(OH)D3, values of 50–55 nmol/L (20–22 ng/mL), the ratio of 3-epi-25(OH)D3 to 25(OH)D3 varied from 2% to 8.5%. In summary, all the methods for vitamin D-testing, including those using LCMS/MS and immunoassay meth- odologies, are in a state of rapid change and improving for both accuracy and precision. In 2011, many immunoas- says for 25(OH)D, including Diasorin Liaison, have been reformulated or restandardised. The clinical significance of studies that fail to report the performance of their 25(OH)D methods in an EQA is dampened and possibly more so when the ability of the 25(OH)D method to separate 3-epi-25(OH)D is also not considered. Received 23 February 2012; accepted 2 April 2012. doi:10.1111/j.1445-5994.2012.02832.x Z. X. Lu 1,2 and K. A. Sikaris 1 1 Melbourne Pathology and 2 Department of Medicine, Monash University, Melbourne, Australia Letters to the Editor © 2012 The Authors Internal Medicine Journal © 2012 Royal Australasian College of Physicians 960