There are several well-recognized variants in assay profile for antibiotics, vitamins, and amino
acids, namely :
(a) Calibration of assay,
(b) Precision of assay,
(c) Accuracy of assay, and
(d) Evaluation of assay performance.
The various aspects of assay profile stated above shall now be treated briefly in the sections that
follows :
Calibration of Assay
Irrespective of the method adopted for the microbial assay it is absolutely necessary to work out
a proper calibration in case the ultimate result is necessarily expected in terms of the absolute units viz.,
mg.L– 1.
Calibrator Solutions — The calibrator solutions are essentially prepared either from a pure
sample of the drug to be assayed or a sample of known potency.
Importantly, there are certain drug substances that are hygroscopic in nature ; and, therefore,
their inherent potency may be expressed as :
(a) ‘as-is’ potency — which refers to — ‘the potency of the powder without drying’,* and
(b) ‘dried potency’— which refers to — ‘the potency after drying to constant weight under
specified/defined experimental parameters’.
Importantly, in as-is potency, the drug should be stored in such a manner that it may not lose or
absorb water ; whereas, in dried potency the drug should always be dried first before weighing.
Thus, once an appropriate ‘standard materials’ is actually accomplished, the calibrator solutions**
usually covering a suitable range of concentrations should be prepared accordingly. However,
the actual number and concentration range of the collaborators shall solely depend on the specific
type of assay being carried out. Likewise, the matrix*** wherein the calibrators are dissolved duly is
also quite vital and important, unless it may be shown otherwise, must be very much akin to the respective
matrix of the samples.
Note : (1) It should be absolutely important when carying out the assay of drugs present in ‘serum’, due
to the fact that protein-binding may invariably influence the ultimate results of microbiological
assay predominantly.
(2) No assay can give rise to fairly accurate results unless and until the suitable ‘calibrator solutions’
(i.e., calibrators) precisely prepared in an appropriate matrix.
Precision of Assay
Precision refers to – ‘agreement amongst the repeated measurements’.
Alternatively, precision is an exact measure of reproducibility, and is duly estimated by replicating
a single sample a number of times thereby determining :
mean result (X) ,
standard deviation (SD), and
coefficient of variation (SD/ X × 100).
Intra-Assay Precision—usually refers to the precision within a single-run exclusively.
Inter-Assay Precision—normally refers to the precision between two or more runs.
Degree of Precision—required in a specific instance essentially will determine two cardinal factors,
namely :
number replicates actually needed for each calibrator, and
number plus concentration range of calibrators.
Accuracy of Assay
Accuracy may be defined as — ‘a measure of the correctness of data as these correspond to
the true value’.
Considering that the calibrator solutions were prepared correctly from the suitable ‘drug’, the
resulting accuracy of a specific result shall exclusively depend upon two important aspects, namely :
precision of assay, and
specificity of assay.
Poor Specificity is encountered usually in the following three instances, such as :
samples comprising of endogenous interfering materials,
presence of other antibacterial agents, and
active metabolites of the ‘drug’ being assayed.
Positive Bias i.e., if the other drugs or drug metabolites are present simultaneously, accuracy
of assay shall be expressed predominantly as a positive bias.*
Negative Bias i.e., if there are antagonists present in an appreciable quantum, accuracy of assay
will be expressed mostly as a negative bias.
Note : In fact, inaccuracy caused due to apparent poor precision will invariably exhibit absolutely
‘no bias’, and that caused on account of either under–or over-potent calibrators will exhibit
positive and negative bias respectively.
Evaluation of Assay Performance
It has been duly proved and established that while assessing the performance characteristics of an
altogether newly developed assay, both intra–and inter–assay precision duly spread over the entire
range of expected concentrations must be estimated precisely.
Important Points : These are as stated under :
(1) It is extremely important to check the accuracy with the help of the ‘spiked samples’** very
much spread over the entire range of concentrations used in the assay.
(2) Assaying ‘drug substances’ in biological fluids e.g., urine, blood, serum, sputum, cerebrospinal
fluid (CSF) etc.
(3) Samples withdrawn from individual subjects who have been duly administered with the drug
either enterally*** or parenterally**** by virtue of the fact that in vitro metabolites may only be
apparent in these instances.
(4) Such substances that might have an inherent tendency to interfere in the assay should be
thoroughly checked for there possible interference either alone or in the presence of the ‘drug substance’
being assayed.
(5) In an ideal situation, preferentially a relatively large number of samples must be assayed
both by the ‘new method’ and the ‘reference method’ individually, and the subsequent results obtained
may be meticulously by linear regression ; and thus the ensuing correlation coefficient of the
said two methods determined.
(6) Routinely employed methods may be tackled with ‘internal controls’* almost in every run ;
and, therefore, the laboratories that are actively engaged in the assay of clinical specimens must take
part in an external quality control programme religiously.