This topic of the poorly labelled sample has been discussed by CPA at some
length, both in the Clinical Biochemistry Specialist Advisory Committee and
in the Joint Advisory Committee.  The guidelines which we found to be
helpful are those produced by the IBMS.  These guidelines are 95% hard line
i.e. they suggest that sample and request form should contain a minimum set
of patient and requester identification before they are accepted for
analysis but, as with all good guidelines, they also contain a let-out
clause indicating that it might be possible to analyse and report on samples
providing the inadequacy of identification is made clear in the report, and
the responsibility for action taken as a result of the report that of the
clinician receiving the report and not of the laboratory.  They also have a
means of attempting to correct the deficit, by contact between a senior
member of the department and the requester, such correction being only by
the requester's written authorisation. (Please forgive me, IBMS, if my
paraphrase is not an accurate representation!).

As with all such situations the easiest way is to be absolutely strict, i.e.
by rigorously deciding not to analyse samples that do not meet the necessary
criteria of identification and insisting that a further sample be taken.
But this is not always possible.  Sometimes a second sample will not give
the diagnostic clinical information that can be obtained from the original
one, either because of rapid metabolic changes or because treatment has been
instituted in the meantime.  Also, it is desirable to minimise trauma, pain
or inconvenience to patients.  And so one is deterred from repeats in babies
or young children, or by asking patients to re-attend (often at considerable
inconvenience) for the sample to be retaken, especially when, as in the case
of dynamic function tests, this repeat may mean a further lengthy stay in
the pathology clinic or a further overnight stay in a hospital bed.  We
should therefore be certain that there is no alternative before demanding a
repeat collection; it is usually the patient and not the errant requester or
phlebotomist that suffers under these circumstances.

My own personal views (and not necessarily those of CPA) are as follows:

1). There are certain areas of laboratory work (e.g. blood bank, cytology
screening) where it is mandatory that all samples must be fully labelled and
documented.
2). It should be a Hospital or Trust decision to allow or not allow any
analysis of incompletely identified samples.
3). There must be a clear standard operating procedure, understood by all
relevant staff, for what should be done when there is unclear or incomplete
labelling of sample or request form.
4). Where there is a mechanism for carrying out and reporting results on an
incompletely identified sample (and some Hospitals or Trusts may decide not
to allow this under any circumstances), there must be no possibility of
incorrect identification additional to that already implicit in hospital
work (fully labelled samples can be taken from the wrong patient or can
contain the wrong information).  Our local policy indicates that we will
carry out investigations in such cases only if the sample and request form
are in the same (individual) envelope and there are no other unlabelled
samples amongst those delivered to the laboratory at the same time
5). The clinician must be warned of the deficient identification and
informed that any action taken on the results is done so on the clinician's
responsibility and no responsibility lies with the laboratory.




%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%