As a clinical microbiologist I occasionally get asked to recommend suitable microbiology tests for a patient, e.g. a returned traveller with a fever, a patient with encephalitis, an immunocompromised patient with CXR changes, etc., etc.
It is always tempting to show off, and display whatever knowledge you have of exotic and peculiar diseases, and give to the requestor an exhaustive (and exhausting) list of investigations to carry out…
There are however a few things to reflect on before constructing such a list:
- Common things are common:- It is important to exclude all the common diagnoses, before considering the more unusual causes of the patient’s symptoms. Returned travellers get flu as well…
- Familiarity leads to competence:- Laboratories are not as good at testing for conditions which they don’t see that often, with the consequent increased likelihood of a false negative or a false positive result. Trust me, you would not want me trying to diagnose your sleeping sickness..
- The laboratory can’t be perfect all the time:- If you request sufficient tests on the one patient, then the odds are you will eventually generate a (false) positive result.
- For each test, think about pre-test probability:- The more exotic your test requests become (“long shots”), the lower the pre-test probability and positive predictive value.
Fishing expeditions need planning and experience. I also prefer a staged approach… “If tests A & B are negative, only then consider tests C & D.”
And whilst on a fishing expedition, don’t forget to treat the patient… There will always be a proportion of patients where you will never get the diagnosis, no matter how hard you try. In the midst of an “investigative frenzy”, don’t forget to cover for the most common and most serious differentials.
No patient was ever cured by investigation alone…
Just to let you know that the Microbiology Matters website has now accumulated 200,000 “visits” since its inception in 2013. It may be some time however before it reaches a million!