Blood Crystal Testing... Necessary?
by Terry Spear
Last week I posted a question for UK forensic scientists on
Forensic-Science UK list. Mark Webster responded and gave me
permission to share his response. So you will see two things with this
post: (1) my inquiry and (2) Mark Webster's response.
I would be curious what UK forensic scientists think about one of the
standards used by our accrediting body, ASCLD/LAB [American Society of
Crime Laboratory Directors-Laboratory Accreditation Board], which
requires the use of a crystal test (e.g. Takayama) before a scientist
can identify a stain as blood. It was not unusual [at least in the
recent past] to see a stain interpreted as "human blood" if it looked
like blood, gave a positive catalytic test (e.g. phenolphthalein) and
either gave a positive reaction with an immunological test for
human/primate protein or reacted with a human/primate specific probe.
It has been the decision of a number of ASCLD/Lab inspectors that the
appropriate interpretation of this set of analytical results should be
"indications of human blood". The rationale behind this
apparently is that this set of tests is no better than a presumptive
test. This changes, however, when an analyst includes a crystal test in
the analytical scheme. Given the presence of the appropriate crystal,
the analyst can now say that human blood was
The main problem that I have with all of this is that I don't think that
a crystal test is any more specific than a catalytic test. It has been
my experience that most catalytic tests are very good at discriminating
between blood and other things. The main problem with false positive
reactions associated with these tests is not typically vegetable
peroxidases or chemical oxidants but significant amounts of
microorganisms. Given that there is literature ["Microorganisms and
Presumptive Tests for Blood/Journal of Police Science and
Administration, Volume 1, No. 4, 1973] which indicates that the Takayama
test can produce crystals which are indistinguishable from those
produced with blood from the catalase enzyme (found in bacteria), it is
hard to understand why this test should be the magic bullet for calling
a stain "blood". Further, since hemoglobin is found in most
fluids [e.g., semen/saliva/vag. fluid], it can hardly be considered to
be a marker unique to blood.
The result of all of this is that most labs do not want to add a crystal
test to their analytical scheme (largely because most analysts do not
want to waste precious sample) and are therefore required to dilute the
quality of their information by reporting "indications of human
Perhaps I am blissfully ignorant but I am not aware of any situations
where human blood was misidentified as a result of relying upon an
analytical scheme consisting of physical appearance, an appropriate
catalytic test and the detection of human/primate protein or DNA.
CA DOJ - California Criminalistics Institute
Mark Webster's response:
I have over twenty years experience as a forensic biologist in the UK. I
have never carried out a crystal test for haemoglobin in anger, and the
number of instances when I have seen it used by other scientists in
years is zero (I see a lot of other scientists' work).
In the early nineties, I was one of three scientists charged with
the methods used by the different Forensic Science Service labs. Only
laboratory used crystal tests routinely at that time. As far as I know,
this practice has been discontinued even in this laboratory.
It is rare to see an assertion that "human blood was present on the
but UK scientists will state: "Blood was present on the knife. The DNA
profile obtained matches Mr Brown's DNA profile and therefore the blood
could have originated from him" or words to that effect.
If Mr Brown is human, this implies the blood is human!
We all know that a "standard" protocol comprising visual examination,
detection of peroxidase activity and immunological identification of
protein or human-specific DNA profiling does not prove that a stain is
human blood. After all, a stain consisting of animal blood deposited on
surface contaminated with non-blood human body fluid would give a
result using this protocol. Adding a crystal test to this standard
adds no value; the crystal test would also be positive. It simply wastes
part of the sample.
I think most UK scientists adopt a commonsense approach.
If there's a large pool of blood under the body and this yields a DNA
profile matching the deceased's DNA profile, any scientist who gave
evidence that he couldn't be sure that the blood was human ("I only have
indications of the presence of human blood") might be looked upon as
strange by the jury.
On the other hand, if the sample is a barely visible or invisible stain
a soiled garment, in some cases it would be appropriate for a scientist
indicate that the stain might not be human blood - and, more
that the human DNA detected might not have originated from the blood.