What Watson has said is correct. Bacteria are identified by percentages, fitting
most likely profiles. If one has a problem organism, that you KNOW is a PURE
CULTURE (this is CRITICAL, and sometimes overlooked - make SURE you're using a one
colony subculture), then one must go to the charts, set up the sugars or other
biochemicals and play the percentages. Aerobic gram positive rods, specifically
the Bacillus species, at least in medical microbiology, are fairly difficult to
identify, and most medical labs won't go to a species identification unless forced
to by the nature of the specimen (blood culture, sterile body fluid). Then again,
it's a matter of ruling out, say, Bacillus anthracis or Bacillus cereus, the two
most problematic species (I don't have the charts in front of me, so I don't
remember offhand what the reactions are that rule these out - it's some combination
of hemolysis and motility, I believe). We used to sign out Bacillus as "Bacillus
species, not anthracis or cereus" for example.
Gram negative rods are identified based on lactose reaction and oxidase reaction,
and biochemicals are set up accordingly. A good basic group of biochemicals
identify most of these (we use a commercial system - Vitek) and then have a backup
commercial system (API) for the problems, with a few other oddball sugars and
biochemicals at our disposal. A lot of times, bugs are identified as _______
species because it's just not possible to speciate, and the doc needs the
sensitivity more than the full identification.
We use the ASM Manual (whatever the current edition is) for problems, but there are
also many other good texts out there. Part of having years of experience in micro
is that you file away the data on the problem bugs that you run into once in
awhile, to utilize that information ten years later when you run into it again. If
there is a problem with commercial systems, it is that one doesn't learn the
biochemical reactions like you do when you set up tubed biochemicals, which is how
I learned and identified bacteria for the first six years of my career. However,
with the volume that our laboratory does, it would be next to impossible to perform
all the identifications in this way anymore - the labor costs would skyrocket.
Now that I'm off my soapbox, I hope this has helped out.
Judy Dilworth, M.T. (ASCP)
Microbiology 26 years
Watson Comehere wrote:
> I see two areas of difficulty that you are encountering.
>> First, by working with dichotomous schemes you are expecting
> a "definitive" answer. Dichotomous schemes
> give yes/no answers -- but sometimes identification requires "judgement"
> as to what the most likely or reasonable or sensible ID is. You just don't get
> a clear cut results.
>> You need to acquire the ability to perform a range of tests and accept
> that in some instances the IDs will be in the form of a probability... or
> your personal judgement about the organism.........
>> Identifying beasties requires a broad look at an array of characteristics
> and sound microbiological judgement. You must develop an ability to accept
> answers in terms of probability.........