\r\n\tThe term robustness is in most cases understood in terms of influence of variations of method parameters on results. Our experience suggests, however, that an additional dimension \u2013 robustness in terms of variability of the sample matrix \u2013 is beneficial: different matrices can lead to different matrix effects (either in the narrow or broad sense) \u00a0and thus different responses and different results with the same analyte concentration in the sample.\r\n<\/p>\r\n\r\n
\r\n\tOn the example of blood plasma: depending on the personal variations in metabolism, diet, possible diseases, e.g. the composition (first of all but not limited to the content of proteins, phospholipids and polyunsaturated fatty acids) of blood plasma can vary significantly, even though formally the matrix is the same \u2013 blood plasma [ref 24].\r\n<\/p>\r\n\r\n
\r\n\tThe sample preparation procedure that is suitable for blood plasma of low protein or phospholipid content may give different results for blood plasma with high protein or phospholipid content.\u00a0This effect is closely related to and is occasionally even addressed under\u00a0selectivity<\/a> in validation guidelines.\r\n<\/p>\r\n\r\n \r\n\tHowever, the possible effects of this kind of variability are not limited to the loss of selectivity, but can also influence recovery (and hence trueness), ionization suppression\/enhancement as well as limit of detection (LoD)\/limit of quantitation (LoQ)<\/a>.\u00a0It is thus useful to investigate the effect of the sample matrix variability (in the case of formally identical matrices) more broadly than just for the selectivity.\r\n<\/p>\r\n\r\n\r\n\t\t\t