When the enumeration of organisms is undertaken, it is important to consider the reliability
of the results obtained, and for this there are three factors that should be considered.
Firstly,
there is the representative nature of the sample being examined with respect to the bulk of
the original water source sampled.
The variation in the number of organisms (counts)
reported for samples obtained from a single water source can be significant, especially over
time as well as between sampling locations.
Secondly, there is the inevitable imprecision
introduced by laboratory procedures when the result is based on an aliquot, or subsample, of
the original sample submitted and not on the entire sample.
Similar imprecision applies when
the count is estimated from a multiple tube method.
A sample bottle brought to the
laboratory should be thoroughly mixed before it is processed.
However, the organisms may
not be distributed evenly or uniformly.
Under the most favourable circumstances, the
organisms will be spread randomly throughout the sample of water.
Hence, by chance,
different aliquots, or subsamples, will contain different numbers of organisms.
This random
variation plays a significant part in all quantitative microbiological
examinations (15).
Thirdly,
there is “uncertainty of measurement” introduced by fluctuations in the performance of staff,
methods, equipment and materials in the laboratory.
It is likely that the variation due to the
representative nature of the sample will be the more dominant (16).
