Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorJody W.ZylkeMD, Contributing EditorIndividualAuthor
To the Editor: In their study of hair mineral analysis by commercial laboratories, Dr Seidel and colleagues1 erroneously claim to have examined the "reliability" of hair mineral analysis. We also believe that the authors have misinterpreted the hypothesis they set out to test.
The reliability of an analytical procedure cannot be assessed by supplying a portion of the same sample to several laboratories. The reliability of data can be determined only from calculation of the coefficient of variation of repeated analyses of portions of the same sample within the same laboratory. Only then could the accuracy and reliability of the data from different laboratories be compared to identify those that are most accurate and reliable. Such a calculation is impossible from the design of this study. Furthermore, the lack of any other objective data concerning the mineral and trace element content of the submitted sample precludes any conclusions about the accuracy of any of the test results. The authors seem to advocate accuracy through consensus and then express dissatisfaction with the degree of consensus they obtained.
Kaminski, Jr MV, Glade MJ. Accuracy of Hair Mineral Analysis. JAMA. 2001;285(12):1576-1578. doi:10.1001/jama.285.12.1576