Tag Archives: specificity



I have noticed while authorising sputum culture reports, that some people have high thresholds for working up suspect colonies from plate cultures, only proceeding when the potential pathogen is dominant amongst the upper respiratory tract flora that is also inevitably present.

Other people have very low thresholds for working up suspect colonies, painstakingly trying to pick out potential pathogens even when there only a few of the same colony type present in the milieu.

Depending on what your threshold is, it clearly has the potential to produce a different result for the clinician.

But who is right and who is wrong? Is it that sometimes we just try a bit too hard?

My gut instinct is that a few colonies of a respiratory pathogen (e.g. Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis) nestling amongst a mass of mixed upper respiratory tract flora is unlikely to be significant, but I am not aware of any literature that supports such an assertion.

I tend not to get too excited by sputum cultures as a rule… Clinicians don’t hang around waiting for the patient to firstly produce a sputum sample, and then wait another 2 or 3 days whilst the microbiology laboratory processes it. They treat the patient empirically, according to guidelines that are hopefully formulated by laboratory data outlining the expected pathogens and antibiograms in the local area.

Only on rare occasions does a sputum culture result actually change patient management.

If “sputum culture” were a new immunoassay, and therefore subject to FDA approval before being released to the market, it wouldn’t have a prayer of being accepted. The sub-optimal sensitivity and specificity (even with prior filtering of samples using macroscopic appearance and Gram stain) would simply not cut it from a regulatory point of view.

So we will continue to grapple with the vagaries of sputum culture, but I suspect it will be around for many years yet.

But we should not lose any sleep over it, whatever your threshold is…


“The Trade-Off”



Do you culture all urine samples that arrive into your microbiology laboratory?

If so, why?

Is it for reimbursement purposes? I know this is necessary for some parts of the world.

..or is it because you are scared not to?

…or is it because you think it is in the patient’s best interests?

Now that automated urine microscopy analysers are commonplace, some microbiology laboratories selectively culture urine samples based on various microscopy parameters, of which the leucocyte count is a principal criterium.

In addition some urines are cultured based on clinical criteria and regardless of microscopy findings; e.g. pregnancy, immunocompromise, neonates, etc.

In my microbiology laboratory just over 50% of urine samples that arrive into the laboratory find their way onto culture plates. The rest do not get any culture whatsoever.

Do I feel nervous about not culturing half of all the urine samples?

Not in the slightest….

Many “positive” cultures with low numbers of white cells simply represent contamination and can thus lead to overdiagnosis, overinvestigation and overtreatment, not to mention the potential for resistance selection.

It is simply a question of sensitivity v specificity for true infection. By altering the microscopy parameters you can make the urine cultures more sensitive but less specific for true urinary tract infection, or vice versa.

There are no right or wrong thresholds for culture.

And on the ones we don’t culture, we always give the clinicians the option of getting back to us by telephone if they feel culture is strongly indicated despite the microscopy findings.

But they very rarely do…


“Should anaerobic culture be performed on superficial swabs ??”

This is a (version of) a post I published a couple of years ago, and I am still toying with the idea. Any thoughts or experience with this topic would be much appreciated, before I take the plunge!

Anaerobic culture has always been difficult in a diagnostic bacteriology laboratory. In the early days  of bacteriology labs anaerobic culture was very difficult if not impossible, so it just wasn’t done.

Then came along anaerobic jars and anaerobic cabinets where the oxygen could be removed from the environment surrounding the agar plate. This led to an “explosion” in anaerobic culture in many different types of samples.

Now the reality is starting to sink home… We are still not that good at recovering anaerobes. This is not particularly the laboratory’s fault. It is just very difficult to mimic the anaerobic conditions of certain parts of the body right from when the bacteria are taken from the patient until they are growing in an anaerobic environment in the laboratory. 

In addition there are areas of the body such as the GI tract, vagina, oropharynx which contain lots of anaerobes anyway, so culturing anaerobes from these areas means nothing at all in terms of pathogenicity.

So back to the question: Should superficial swabs be cultured for anaerobes?

My gut (excuse the pun) feeling here is that we are not helping the patient by culturing superficial swabs for anaerobes because:

  • Superficial areas are exposed to oxygen, so although anaerobes may exist in deeper areas of a wound, getting a good sample of such with a swab is very difficult.
  •  A clinician’s  decision whether to include anaerobic  cover should very much be decided by the clinical presentation, not the microbiological results. Eg bite wounds, infected diabetic foot ulcers, aspiration pneumonias, dental infections are all clinical conditions where anaerobic cover should generally be routine. Not culturing anaerobes from such specimens should not prompt discontinuation of anaerobic cover.
  • The growth of anaerobes or mixed anaerobes from superficial wound is of dubious value. If isolated, are the anaerobes really causing the problem? My experience is that when anaerobes are found in superficial swab culture, they are often found with other colonising enteric flora, so are of doubtful significance. Other anaerobic isolates may co-exist with other pathogens such as Staphylococcus aureus, and treatment of the Staph aureus alone often sorts out the infection. 

There is no question that anaerobic culture is of more importance when the sample is taken from a sterile site. In these cases it can have a real impact on both the diagnosis and subsequent management of the patient. E.g. Fusobacterium necrophorum in a blood culture suggests the possibility of a Lemierre’s syndrome. Isolating Bacteroides in a blood culture suggests intra-abdominal pathology, often a collection of some sort.

Lots of laboratories still culture for anaerobes from at least some superficial swabs, because that is the way it has always been, or that is what is expected of the lab, or because we do isolate anaerobes from superficial swabs.

However I think that when we decide to culture a superficial swab for anaerobes, we don’t really think about how poor both the sensitivity and specificity of the result is. Although we are trying to be helpful, the truth is that we may not be helping at all….