Thanks Dr Saad

Mani, I attach papers for you to see yourself. I have highlighted in yellow the paragraph of interest. This marker is not futile, as proven by continuous efforts to demonstrate it's value (see the recent paper)

We used this marker in clinical practice in the past, due to poor lab diagnostic tools at that time, as well as restrictions in using broad-spectrum antibiotics without a "solid" justification. We convinced our clinicians that they can trust in it by prospective analysis or nearly one thousand paired serum-csf samples in patients finally diagnosed with bacterial or viral meningitis.

We succeeded to demonstrate CRP utility in clinical practice subject to right time collection of the samples (no later than 5 days from the symptoms onset - in our experience this increased the confidence). Thus we clearly demonstrated the predictive value in two instances: (i) orientation to the etiology of the infection (whether bacterial or non-bacterial), and (ii) useful evolution marker, i.e for rapid or delayed recovery, as well as the probability of sequelae occurrence.

The value of CRP increases if transferrin determination is added.

I say non-viral because in the absence of a good clinical evidence (which was not our case) CRP levels might be slightly increased in conjunction with medical conditions other than viral infection. I have another paper to attach but unfortunately it is only one per intervention, so I will have the third post. I apologize for that!

Metaanalysis here http://www.ncbi.nlm.nih.gov/pubmed/9819187

Romanian published paper here http://www.ncbi.nlm.nih.gov/pubmed/9524672

Bangladesh recent paper here (CRP in children) http://www.banglajol.info/index.php/BJCH/article/view/14275

And interferences to check for correct result interpretation (incliding but not limited to), attached. I am done

Thanks a lot Doinita for your valuable suggestions.

The main questions with these diagnostic indicators, are they useful in clinical practice? Furthermore are they more efficient than the emergency physician?

CRP levels are elevated in case of proven bacterial meningitis (PBM). Thus CSF CRP may be of interest, but the cut-off value is not yet established. Whereas two series of 7 PBM found that a CSF CRP> 100 ng/ml is discriminant, others reported that CSF CRP > 640 is predictive. Furthermore in 20 pediatric PBM, authors highlighted a 0.4 mg/l value. In some series CSF CRP is too insensitive (sensitivity 66%). The research must go on…

Thus, whereas CSF CRP could be one of the diagnostic indicators in case of negative gram-stain smear, I completely agree with the IDSA guidelines and Allan Tunkel’s opinion “these markers should not be used to determine whether an individual patient should receive antimicrobial therapy”.

In clinical practice, prediction models for community acquired bacterial meningitis may be useful, but at least for adults and children greater than 4 years, the presence of at least one sign of severity ( altered consciousness, focal neurologic findings, seizures or shock) and an absolute CSF neutrophils count are predictive of bacterial meningitis. (Accuracy of clinical presentation for differentiating bacterial from viral meningitis in adults: a multivariate approach. a series of our group. Intensive Care Med 2005, full text in my profile and publication).

Of course the dilemmas for the meningitis diagnosis are common, but the main problem for the physician is to define when the suspicion of BM is sufficiently high to start empirical antibiotic treatment, whatever CSF results. The clinical judgment being the principal factor for BM. Nobody will be disapproved for an empirical treatment which will be withdrawn in case “only viral meningitis”, in contrary to…..

CSF Pro CT remains a matter of debate. Does CSF proCT levels add any value to standard is unclear.

CSF proCT was not mentioned in the recent BM review of the lancet 2012 (dilemnas within the dg of acute community acquired BM), in contrast to serum levels.

CSF ProCT impact for the dg of BM was mainly assessed by Retrospective studies.

Serum proCT may be useful for patients with CSF negative gram stain, and may be in case of postoperative status;

We should not forget the 99% sensitivity, the 99% NPV of the BM score as recently reported by Negrovnic et al meta-analysis.

So be careful, before claiming that CSF is the best biomarker for BM