Meningitis & Encephalitis Diagnosis
Meningitis & Encephalitis Diagnosis
The assessment here runs in a deliberate order, and the order is what prevents the two errors this condition invites: giving the treatment too late because a scan had to happen first , and taking the fluid through raised pressure because the puncture was the next step on the list . Samples come first. Imaging comes where it is indicated. Treatment does not wait for either. The lumbar puncture is the test that separates the causes, and it is worth being precise about what it settles and what it leaves open, because the gap between the two is where most of the mistakes happen — including the commonest one of all, which is a fluid reported as negative after antibiotics have already been given, and an illness then filed as unexplained rather than as unanswered.
The question that comes before anything else
There is a question that has to be answered before any of the machinery is brought to bear, and it is not which organism this is. It is which layer is involved: the membranes, the brain substance, or both.
The reason it comes first is that it decides what is requested and what is started before the answer is in. A person with clear meningitis needs the fluid, and needs it promptly. A person with encephalitis needs imaging of the brain substance and an electroencephalogram, needs antiviral cover while the question is open, and needs their fluid interpreted knowing that in a disease of the brain substance the fluid may be entirely normal. Treating both as the same workup produces the same failure in two directions: an encephalitis investigated only by puncture, and a meningitis imaged while the organism goes unnamed.
The question is answered at the bedside rather than in a laboratory, and it is answered from three things: the level of consciousness, the behaviour and speech, and whether there has been a seizure. A febrile, headachey, miserable person who is entirely themselves is one question. A febrile person who is not themselves is another, and the difference between them is larger than any number on this page.
The second question, asked alongside rather than after, is whether the presentation carries any of the features that make a lumbar puncture unsafe at that moment. That is the subject of the next section, and it is asked before the needle rather than at the time of it.
The lumbar puncture, and when the scan comes first
A lumbar puncture measures the pressure at which the system is running and takes a sample of the fluid for chemistry, cells, and whatever the question turns out to require. It is the single most informative test in this condition, and it is also the one that has a rule attached to it.
The rule is this. Where there are focal neurological signs, a recent seizure, a fall in the level of consciousness, or swelling of the optic discs, imaging comes first and the puncture is deferred. Deferred, not cancelled: the question is still there, and it is asked as soon as it is safe to ask it. Where none of those is present, the puncture comes first, because imaging cannot name the organism and the fluid can. Neither order is a reason to delay antibacterial therapy, which is the sentence that gets forgotten most often in the middle of an argument about sequence.
What is taken is worth knowing, because the list is longer than families are usually told and the extra tubes cannot be added afterwards. Opening pressure. Cell count with a differential — which type of cell is there, not merely how many. Protein. Glucose, taken against a blood glucose drawn at the same time, because the ratio is the point rather than the number. Gram stain and culture. And material held for a polymerase chain reaction panel, for herpes simplex virus, for tuberculosis, and for cryptococcus, according to the question.
Two practical things complete it. Blood cultures are drawn before the first dose wherever that does not delay the dose, and they are worth more than families realise because they are the sample that cannot be recovered later. And the timing of the puncture relative to the first dose is written down, because it changes how every result that follows is read.
What the fluid says
The fluid is read as a pattern rather than as a number, and the pattern is what separates the groups. What follows is the shape of each rather than a set of thresholds, because the thresholds overlap and the clinical state is read alongside them.
| Group | What the fluid shows | What it does not settle |
|---|---|---|
| Bacterial meningitis | A high number of cells, predominantly neutrophils; protein raised; glucose low relative to the blood; organisms visible on the Gram stain in a proportion of cases | Not the organism's name, and not its sensitivities. The culture and the panel do that, and both can be sterilised by a dose already given |
| Viral meningitis | A moderate number of cells, predominantly lymphocytes; protein normal or mildly raised; glucose normal | Not which virus, and not a discharge from the question. The pattern is compatible with several, and the specific ones are requested by name |
| Tuberculous meningitis | Lymphocytes, a markedly raised protein, and a glucose that is low — a combination that is characteristic and that evolves over days rather than appearing complete | Not confirmed by one sample. Nucleic acid amplification helps, culture takes weeks, and a negative first sample does not close the question |
| Cryptococcal and fungal meningitis | A modest number of cells, often in someone whose immune response is blunted, with the organism seen on specific staining or its antigen detected | Not excluded by a normal cell count. In an immunosuppressed person the inflammatory response may be small and the organism plentiful |
| Encephalitis | Often a modest lymphocytosis, and sometimes entirely normal — because the disease is in the substance rather than in the fluid | A normal fluid does not exclude encephalitis. This is the single most consequential line on this page |
| After antibiotics have been given | Cultures sterile, cell type shifted, and the pattern blurred — but the chemistry largely intact and the genetic material still detectable | Not a negative result. It is an unanswered question, and the panel is the test that answers it |
Where a first sample has named nothing and the question is still open, the puncture is repeated rather than the question closed. That is the discipline: a second sample taken a day later, in a condition that is evolving, regularly names what the first one did not.
What imaging establishes, and what it does not
3.0T magnetic resonance imaging is the study that answers the question of level, and it is worth being precise about what it settles and what it leaves open. What it settles is whether the brain substance is involved and in what pattern — and in the encephalitides the pattern often points at the cause before the fluid does. The herpes form has a predilection for the temporal lobes and the structures of memory, and where the imaging shows that, the antiviral decision is made on the imaging rather than waiting for a result that takes longer to arrive.
It also settles what else is happening: whether there is a collection, an infarct, a swelling that is shifting structures, or a hydrocephalus that has appeared as a consequence. Those are the findings that change what is done in the next hour rather than the next week.
What it does not do is name the organism, except where the pattern is distinctive enough to narrow the question — which is sometimes, not usually. It does not replace the fluid. It does not tell you whether the pressure is behaving abnormally at the moment the needle would go in, which is why computed tomography rather than magnetic resonance imaging is the study that precedes a deferred puncture: it is fast, it answers the safety question, and it is available at the hour.
Where the consciousness is fluctuating and the imaging has not explained it, long-term video electroencephalography is the study that does. Non-convulsive seizure is a commoner explanation than families are told, and it is treatable rather than observable.
Blood, and the tests beyond the fluid
Blood cultures are drawn before the first dose and are the sample that cannot be recovered later. Alongside them, blood tests establish the inflammatory picture, the kidney and liver function that determine what can be dosed, and — where the immune state is the question rather than the headache — the immune state itself.
Beyond the fluid and the blood, three groups of tests are worth knowing about, because each is requested for a reason rather than as part of a panel. Specific organism testing: nucleic acid amplification for tuberculosis, cryptococcal antigen where the immune state suggests it, and serology where the organism and the exposure make it the right question. Autoimmune antibody panels, requested where the course is subacute, where it relapses, or where the fluid has named nothing — because in those three settings the question has turned rather than merely persisted. And audiology, which is not a diagnostic test at all but which belongs in every plan for a bacterial meningitis, and which is requested before discharge rather than after a family notices something at home.
What should not happen is a workup that continues to add tests without a question attached to each. Adding tests that have no bearing on the distinction is how a week is spent and how a discharge summary ends up saying meningitis, unspecified.
What a finished assessment leaves behind in writing
Five things, and a family is entitled to ask for each of them.
Which level was involved — the membranes, the substance, or both — stated as a judgement rather than as a restatement of a discharge code.
Which organisms were excluded, and on what grounds. This is the item most often missing, and it is the one that matters most to a family who will otherwise spend years not knowing. A negative culture taken after a dose of antibiotics is a different statement from a negative polymerase chain reaction, and both are different from a test that was never sent.
What remains plausible, and what would settle it. Where the question is genuinely open, the answer is a named next test with a reason attached rather than a general instruction to see how things go.
What the follow-up contains, with intervals: audiology, cognitive review, and a review appointment at which the question of a relapse is asked rather than assumed away. Each with a date rather than a general instruction to keep an eye on it.
What would bring the person back. The sentence that answers the question families are most often not given, written down rather than left to be inferred. What treatment consists of is set out on the treatment page.