Neuro-Infections: The Evidence Behind the Pathway

It is the monsoon, and a 38-year-old agricultural labourer is carried into the casualty of a district hospital. He has had fever and headache for five days, and since this morning he has been confused. His Glasgow Coma Scale is 11, his neck is stiff, and there are no focal signs. The computed tomography scanner is in the next building, and the radiographer is busy. This is an illustrative case, not a real patient.

The Neuro-Infection Pathway tells you what to do for him in the next hour: blood cultures, the first dose of antibiotics, the doses in both national documents, when the tap can wait and the treatment cannot, and how to read the fluid when it comes. What it no longer carries is most of the why. The Indian series, the trial figures and the mechanisms that used to sit in its Doses and Evidence tabs have moved here, so that the tool stays quick at the bedside. By the end of this article you will know why each of those instructions is there, and which of them rests on the firmest ground.

1. The commonest cause is the one the sheet does not treat

The ICMR Standard Treatment Workflow for neuroinfections gives one drug regimen for meningitis, and it is a pyogenic one. ICMR is the Indian Council of Medical Research. Before you reach for that regimen, ask what this man is most likely to have. Two Indian series settle this.

• At Chandigarh, 401 consecutive patients aged 12 and above admitted to the medical emergency with a central nervous system infection: tuberculosis 51.5 per cent, viral meningoencephalitis 13.9 per cent, community acquired bacterial meningitis 9.7 per cent, cryptococcal 6.2 per cent. Mortality 27.9 per cent, and 64.5 per cent where no cause was found (Kumar 2021).
• In a separate Chandigarh cohort of 105 adults with acute febrile encephalopathy and a Glasgow Coma Scale of 10 or less, tuberculous meningitis was 25.7 per cent and pyogenic meningitis 17.1 per cent (Ary 2023).

Notice that the two series use different denominators: the first counts every adult with a central nervous system infection, the second only the sickest, with a Glasgow Coma Scale of 10 or less. They agree on the point that matters. Across four Indian adult series, as the tool gave them, the tuberculous share of central nervous system infection ran from 4.2 to 51.5 per cent, and in the two largest it was 36 and 51.5 per cent. The tool does not name every one of those four series in its reference list, and neither can this article.

“So should I start antituberculosis treatment on everyone?” No. The tool asks you to give the pyogenic regimen at once, because bacterial meningitis kills in hours, and to keep tuberculosis on the list until the fluid has spoken. Look again at that 64.5 per cent: the patients in whom no cause was found did worst of all. The diagnosis you never make is the most dangerous one.

2. The first doses, side by side

The tool prints two Indian national documents next to each other: the workflow of October 2019 and the ICMR antimicrobial guideline of 2022. On a phone it now shows this as two narrow tables. Here is the whole table, as the tool carried it on 4 November 2026. Read the last column slowly: it is where the difference becomes a patient.

“Why does vancomycin matter so much? Ceftriaxone covers the pneumococcus.” Usually it does. Vancomycin is in the regimen for one reason: cephalosporin non-susceptible pneumococcus. The ICMR antimicrobial guideline records why that reason has grown. In meningeal pneumococcal isolates at Vellore, penicillin non-susceptibility rose from 9.5 per cent in 2008 to 42.8 per cent in 2016, and cephalosporin non-susceptibility from 4.7 to 28.5 per cent (ICMR 2022). More than a quarter of those isolates would not have been reliably covered by ceftriaxone alone. So vancomycin is there to cover exactly the organism ceftriaxone may miss, and an underdose of that drug is the worst place to be short.

For our patient, if he weighs 60 kg, the 2022 table gives 900 mg eight hourly. The workflow figure of 500 mg, with no interval printed, could be read as a third of that.

3. Dexamethasone: the timing is the evidence

Why is the steroid given before the antibiotic, and not after? Because that is how the trial that put it into these regimens was done. The European trial that put dexamethasone into these regimens gave 10 mg 15 to 20 minutes before or with the first antibiotic dose, and then every six hours for four days. In 301 adults it reduced unfavourable outcome (relative risk 0.59) and death (relative risk 0.48), and the benefit was concentrated in pneumococcal disease, where unfavourable outcomes fell from 52 to 26 per cent (de Gans 2002).

A relative risk of 0.48 means death was a little under half as likely in the dexamethasone group. Look at where the benefit sat, too: in pneumococcal disease. A dose given hours later, after the antibiotic, is not the treatment the trial tested, and the tool does not pretend it is.

This is the common mistake, and it is an easy one to make in a busy casualty: the antibiotic goes in, and the steroid is remembered at the next round. The tool keeps the 2022 dose, the interval and the timing in its Doses tab, and the WHO rule that goes with it: stop the steroid if the fluid does not look bacterial.

4. Reading the fluid

When the tap is done, this is the table the tool reads it against. The bacterial and viral columns are the World Health Organization’s table 3.1 of 2025. The tubercular column is not from that table: the figures are from Indian series named in the tool’s references (Sahu 2026; Chaurasia 2023). On a phone the tool now shows it as two narrow tables; here it is whole.

“Lymphocytes, high protein, low sugar. Isn’t that always tuberculosis?” Not in India. Scrub typhus meningitis sits between the columns, and it is the one that gets called tuberculous. The tool keeps the figures that separate them, from two Indian series, in its Doses tab, because they change what you do tonight.

5. The eschar nobody looked for

Scrub typhus appears nowhere on the workflow. That matters more than it sounds. In the Gorakhpur outbreak investigation of 2016, of 407 hospitalised patients with acute encephalitis syndrome and a cerebrospinal fluid pleocytosis, 65.4 per cent had scrub typhus and 10.3 per cent had Japanese encephalitis. Case fatality was lower in the scrub typhus group, 15.2 against 36.2 per cent, which is what a treatable cause looks like in a mortality table (Mittal 2018). In adults in Puducherry the attributable fraction was 11 per cent. The Puducherry paper is in the tool’s source notes but not on its reference list.

An attributable fraction is the share of the illness that can be put down to that one cause. Eleven per cent of adult acute encephalitis is a lot of patients who will recover on a cheap tablet. So turn our patient over, look in the axilla, the groin and under the waistband for a black, painless eschar, and remember that doxycycline is in the 2022 empirical regimen for acute encephalitis syndrome even though it is absent from the workflow.

6. Why a quinine bolus is so dangerous

If his malaria test comes back positive for falciparum, the workflow’s secondary care column says quinine 20 mg/kg bolus. India’s national malaria guideline says, in capital letters, never give a bolus (NVBDCP 2014). The tool prints both, and tells you to follow the guideline. Why the capitals?

The mechanism is worth holding: quinine is a potent stimulus to insulin release, and a comatose patient with falciparum malaria is already hypoglycaemic more often than not. A bolus adds hypotension to that. So a fast injection pushes a patient who is already short of glucose and already at risk of shock further down both slopes at once. The loading dose goes in over four hours, in dextrose. Artesunate is the first choice in any case.

7. The tap that was not done

Every source the tool read agrees that the lumbar puncture should happen and that most patients need no scan first. So why does it so often not happen? Practice diverges from that in Indian hospitals for reasons that are largely structural: the scanner is downstairs and busy, consent takes time, and nobody is blamed for the tap that was not done. The cost lands later, as a sterile culture and a patient on 14 days of a cephalosporin for a disease nobody named.

Think back to the Chandigarh figure: where no cause was found, nearly two in three died. A tap done after the first dose of antibiotics will still show the cells, the protein, the sugar and the adenosine deaminase, and it can still be sent for the cartridge test for tuberculosis. A tap not done shows nothing.

Back to the casualty

Our patient has two sets of blood cultures and a malaria test sent. He gets dexamethasone and then, within the hour and without waiting for the scan, ceftriaxone 2 g twelve hourly, vancomycin at 15 mg/kg eight hourly, and doxycycline, as the 2022 guideline gives for acute encephalitis syndrome, because scrub typhus has not been ruled out. Acyclovir is added, because his confusion makes encephalitis possible. The scan is done when it can be, and the tap after it if the scan is clear. When the fluid comes back, it is read against all three columns, and tuberculosis and scrub typhus stay on the list until the results take them off. Each of those steps is a line in the tool. Now you know what stands behind each line.

Try this

A 75 kg woman with suspected bacterial meningitis and a possible encephalitis is written up for vancomycin 500 mg and acyclovir 500 mg, both eight hourly. Using the 2022 doses the tool prints, what should she get? Illustrative case.

Answer. Vancomycin 15 mg/kg, to a maximum of 2 g a dose, eight hourly: about 1,125 mg eight hourly, more than twice what was written. Acyclovir 10 mg/kg eight hourly: 750 mg. At 75 kg, 500 mg is two thirds of the dose, which is exactly the gap the tool warns about. And if her fluid turns out not to look bacterial, her dexamethasone stops.

Take-home points

• In an Indian adult with fever and confusion, tuberculous meningitis is as likely as pyogenic meningitis, and the national sheet treats only the second.
• Blood cultures, then the first dose, then the tap when it is safe. The scan never delays the antibiotic.
• Vancomycin is there for the resistant pneumococcus. Dose it by weight, eight hourly.
• Dexamethasone goes in before or with the first antibiotic dose, not later.
• Look for the eschar, think of scrub typhus, and never give quinine as a bolus.

Declaration: the author built and maintains the tool. It is free to use. The figures in this article are those the tool carried in its Doses and Evidence tabs on 4 November 2026, each from the source named beside it; where the tool names no source on its reference list, the article says so. The Neuro-Infection Pathway is still in administrator preview, pending clinical review.

References

1. Indian Council of Medical Research, Department of Health Research. Standard treatment workflow: neuroinfections (ICD-10-G03.9). New Delhi: Ministry of Health and Family Welfare; 2019.
2. Indian Council of Medical Research. Treatment guidelines for antimicrobial use in common syndromes. Chapter 8: CNS infections. New Delhi: ICMR; 2022.
3. World Health Organization. WHO guidelines on meningitis diagnosis, treatment and care. Geneva: World Health Organization; 2025.
4. Tunkel AR, Hartman BJ, Kaplan SL, Kaufman BA, Roos KL, Scheld WM, et al. Practice guidelines for the management of bacterial meningitis. Clin Infect Dis. 2004;39(9):1267-84.
5. Tunkel AR, Glaser CA, Bloch KC, Sejvar JJ, Marra CM, Roos KL, et al. The management of encephalitis: clinical practice guidelines by the Infectious Diseases Society of America. Clin Infect Dis. 2008;47(3):303-27.
6. de Gans J, van de Beek D. Dexamethasone in adults with bacterial meningitis. N Engl J Med. 2002;347(20):1549-56.
7. Sharma SK, Ryan H, Khaparde S, Sachdeva KS, Singh AD, Mohan A, et al. Index-TB guidelines: guidelines on extrapulmonary tuberculosis for India. Indian J Med Res. 2017;145(4):448-63.
8. National Vector Borne Disease Control Programme, National Institute of Malaria Research. Guidelines for diagnosis and treatment of malaria in India. 3rd ed. New Delhi: Ministry of Health and Family Welfare; 2014.
9. Kumar D, Pannu AK, Dhibar DP, Singh R, Kumari S. The epidemiology and clinical spectrum of infections of the central nervous system in adults in north India. Trop Doct. 2021;51(1):48-57.
10. Ary KA, Singh H, Suri V, Sharma K, Biswal M, Singh MP, et al. Changing clinical profile and predictors of mortality in patients of acute febrile encephalopathy from north India. J Glob Infect Dis. 2023;15(3):101-7.
11. Mittal M, Bondre V, Murhekar M, Deval H, Rose W, Verghese VP, et al. Acute encephalitis syndrome in Gorakhpur, Uttar Pradesh, 2016: clinical and laboratory findings. Pediatr Infect Dis J. 2018;37(11):1101-6.
12. Sharma K, Sharma M, Shree R, Modi M, Goyal M, Narang D, et al. Xpert MTB/RIF ultra for the diagnosis of tuberculous meningitis: a diagnostic accuracy study from India. Tuberculosis (Edinb). 2020;125:101990.
13. Chaurasia AK, Gupta P, Mathur MK, Nagi V. Study of the role of cerebrospinal fluid C-reactive protein and adenosine deaminase to differentiate various types of meningitis. J Assoc Physicians India. 2023;71(5):11-12.
14. Valappil AV, Thiruvoth S, Peedikayil JM, Raghunath P, Thekkedath M. Differential diagnosis of scrub typhus meningitis from tuberculous meningitis using clinical and laboratory features. Clin Neurol Neurosurg. 2017;163:76-80.
15. Sahu SK, Dhamija RK, Maheshwari S, Panda AK, DungDung AA, Khurana S, et al. Clinical profile and prognostic determinants in central nervous system tuberculosis: an observational study from a tertiary neurology centre in India. Ann Indian Acad Neurol. 2026;29(2):217-21.
16. Marais S, Thwaites G, Schoeman JF, Torok ME, Misra UK, Prasad K, et al. Tuberculous meningitis: a uniform case definition for use in clinical research. Lancet Infect Dis. 2010;10(11):803-12. [Cited for the case definition only. The scoring is not implemented in this tool]