Important characteristics of a clinical chemistry assay are its
limit of blank (LoB), and its limit of detection (LoD).
The LoB, conceptually the highest reading likely to be obtained from
a zero-concentration sample, is defined operationally by the upper 95%
point of readings obtained from samples that do not contain the analyte.
The LoD, conceptually the lowest level of analyte that can be reliably
determined not to be blank, is defined operationally as true value at
which there is a 95% chance of the reading being above the LoB. These data are from a portion of a LoB/D study
of an assay for a drug used to treat certain cancers.
Twelve pools were used, four of them blanks of different types,
and eight with successively increasing drug levels. The 8 columns
of the data set refer to measurements made using different instruments I
and reagent lots L.