Diagnosing Central Nervous System (CNS) Tuberculosis in the laboratory is not easy. There needs to be a balance between rejecting unnecessary and inappropriate investigations for CNS TB and at the same time making sure no true cases are missed.
Here are a few tips:
Investigations for TB in CSF should always be requested only after CSF protein and cell parameters are known. (Of course these parameters can occasionally be normal in CNS TB but knowledge of them changes the positive and negative predictive value of any future testing)
Investigations for TB PCR in CSF should always be requested by a consultant clinician. (on the assumption that a consultant has the knowledge base and experience to best select the patients where testing is appropriate, not always the case however). I have seen numerous instances of junior doctors requesting TB PCR on 0.5 ml CSF on a patient with a couple of days of headache.
Investigations for TB PCR in CSF should always be approved by a clinical microbiologist. If there is no such filter, there is the risk of testing volumes running riot and inappropiate testing occurring. TB PCR is specific but only has a sensitivity of between 50 and 70% at diagnosing CNS tuberculosis. It is important to make requestors aware that a negative result is virtually meaningless. Auditing TB PCR requesting volume and indications is always a good audit project for students.
Insist on adequate CSF volume for testing. The sensitivity of both culture and PCR tests for TB in CSF increases with increasing volume submitted. Different labs have different volume requirements for these tests. For optimal testing, somewhere around 5 ml CSF are needed for these tests, which is not an insignificant volume. That is not to say that smaller volumes should always be rejected. The clinician needs to be made very aware however that doing TB culture or PCR on tiny amounts of CSF may be of extremely limited value.
Educate the requestors. With written protocols, presentations etc TB investigations in CSF is a problem that must be faced by many clinical microbiology departments throughout the world. Better to be pro-active than reactive about what is acceptable and what is not.
For more detail on the diagnosis of CNS tuberculosis see these guidelines, which are well written and still reasonably up to date. They also highlight the importance of clinical suspicion, comprehensive imaging and empirical treatment when laboratory diagnosis is proving elusive.
In the developed world, big laboratories will receive many hundreds of requests for TB culture and TB PCR in CSF, but will likely be able to count the number of probable or confirmed cases on the fingers of one hand. Close co-operation between the laboratory and the clinicians is required, as it is not a diagnosis you want to miss…