When is a fever not ‘just a fever’?

Fever is one of the most common acute presenting complaints in Paediatrics, with 20-40% of parents reporting at least one episode per year.

Assessment and diagnosis can be difficult when covering a broad spectrum, from self-resolving mild upper respiratory tract infections to severe bacterial infections leading to sepsis. Research around this area often focuses on differentiating bacteria from viral infections, ensuring appropriate antibiotic prescribing, and separating those who can recover at home with paracetamol from those who require admission for more intensive observation and management.

However, the list of differentials for fever in a child goes far beyond the initial ‘bacterial vs viral’ debate in those whose fever does not disappear after a few days of antibiotics or supportive care.

Managing these unusual fevers can prove challenging when faced with a never-ending diagnosis list and few clues to help, whether it’s an atypical or non-pathogen.

Sometimes, it can be helpful to take a step back and remind ourselves what a fever is—what physiology are we observing in a child whose temperature is above the acceptable limit?

Why do we mount a fever response?

Fever is one of the most outwardly observable aspects of the body’s immune response and the one we are most used to associating with an illness.

The cellular pathway causing fever is complex, relying on many molecules. Some of these will sound familiar, such as cyclooxygenase 2 (COX2), the target of a well-known antipyretic: ibuprofen.  

Without getting too far down the cellular spiral, this means that whilst our brains jump to infection as the cause when they see fever, that’s not always the case. We’re seeing the stimulation of specific molecules, which might be down to a nasty pathogen but might be something else entirely.

You see an eight-week-old in the Emergency Department presenting with a fever, and this is confirmed in triage.

They have mild tachycardia associated with the fever but are otherwise well, well hydrated, and have no identifiable symptoms.

They had their first set of immunisations 48 hours ago.

What do you do?

The deciding factors here are the infant’s age and the fever—in an older child, a day of fever in an otherwise healthy child would likely not cause much excitement. It is debatable whether we should indiscriminately treat children less than three months of age with fever as suspected sepsis.

But without any focal symptoms to point to a viral URTI or clear environmental cause, the next steps seem pretty clear. The blood-brain barrier is too immature at this age to block potential pathogens from entering, so the worry is not just about a serious bacterial infection (SBI) disseminating through the blood but also into the brain.

While SBIs are usually the overwhelming concern for such cautious management in infants, an additional consideration here, especially in infants under eight weeks, should be the Herpes Simplex Virus (HSV). Disseminated HSV, passed on from maternal genital lesions during delivery or from anyone close in the postnatal period (typically cold sores), can have disastrous outcomes. It can also present in less severe forms, such as skin, eyes, and mouth (SEM) disease or encephalitis.

Post-vaccination fevers

A fever post-immunisation is an expected side effect, and parents are often advised to give paracetamol to combat this.

This can then pose a difficult conundrum if a child under three months presents with a fever after vaccination – do we allow it as a side effect, but could we then be missing an underlying infection?

There is, unfortunately, no simple answer.

A few factors come into play: how well the child appears, how long since the vaccinations and why the parents/carers brought them in. Clearly, if the child is symptomatic or appears unwell, it cannot be attributed to the post-immunisation state, and they should be treated. Conversely, a well-looking child cannot be absolved of any possible infection.

Vaccination in the previous 24 hours may well be responsible for the fever. However, as the child passes this window, this should no longer be a factor in decision-making.

The reasons for attendance can also help risk stratify these children. A parent/carer worried about how they’ve been raises more concern than one who has brought them in as they have been told to attend with any fever but feels the child is well in themselves. Ultimately, this is a clinical judgement each time, based on the individual circumstances.

Remember that your usual inflammatory markers (like CRP) will go up after immunisation as the body is mounting an immune response, so they cannot be used as a reliable indicator.

You decide 48 hours is too long, and you can’t blame the immunisations…

The next step is a partial septic screen, with blood cultures and empiric antibiotics as the minimum. The need to perform a lumbar puncture should be decided clinically.  

You send them off, along with some routine bloods and an extended NPA swab and admit to the ward for IV antibiotics guided by your local antimicrobial policy.


The NPA swab is positive for rhinovirus, and the blood culture is negative at 48 hours – both good evidence to stop the antibiotics and chalk this up to a viral URTI…right?

Did rhinovirus cause this fever?

It could have, but is it convincing enough? Maybe if the fever has settled and the child is well.

Really, though, it’s a ‘we don’t know’. This is not the end of the story. We can’t always figure out where a fever has come from, and it’s always worth considering the possibility of dual pathology – especially if things start to look a little unusual.

Rhinovirus doesn’t have a strong track record of proven causation. A multi-site study across Europe showed that testing positive for rhinovirus meant you were more likely to have a bacterial infection than a viral one.

You get called right before they’re discharged. They’ve spiked a fever again!

You examine them, looking for anything of note– they still have no coryzal symptoms, but a small umbilical granuloma now looks a little red and weepy.

Bingo! You’ve found a source!


With otherwise normal observations, you prescribe a five-day course of flucloxacillin and send them home, with safety netting to return if the fever doesn’t improve after 48 hours of these antibiotics.

Satisfied with another case solved, you finish your shift remembering the importance of a complete top-to-toe exam to look for possible sources of infection.



Two days later, who should turn up with a fever but the same patient?!

Hang on….they’re STILL febrile?

When unsure what’s happening, it’s time to return to basics. Is it the wrong pathogen, the wrong treatment or something else entirely?

A full septic screen means more blood cultures and a lumbar puncture. Given the persistent fever and age, meningitis has to lurk in your top differentials.

Check the inflammatory markers, particularly at the acute and chronic inflammation organ markers, to look for focal inflammation or damage, and a blood film to visualise those cells.

Check the urine and a chest X-ray for the most common sources of infection. Then, consider another imaging modality, such as an abdomen ultrasound, to look for any free fluid or organomegaly. Do they need an echocardiogram to look for vegetation suggesting endocarditis?

Your most useful test will be detailed temperature monitoring with accurate record keeping. An infant with an ongoing fever for two weeks is not the typical presentation, so the more detail you have, the better.

Does it happen at the same time every day? Are there any precipitating factors? Does it swing up and down regardless of antipyretics? Is it EVERY day, or is it intermittent? How high is it going, and how does the fever affect the infant?

Identifying fever patterns can do two things. It can point towards specific pathologies and verify the presence of fever. It can also ensure that it is clearly reported with exact measurements that can be used to justify expensive or difficult-to-access future investigations.

What if the infection isn’t the problem?

This could be an infection that hasn’t responded to antibiotics – if the pathogen is not bacterial or if the bacteria is resistant to the antibiotic being used. It could also be an average infection, and the reason it hasn’t responded the way you expect is the HOST.

An immune deficiency may mean the host doesn’t clear a regular infection as easily as someone with a competent immune system, confusing the diagnosis.

Consider a simultaneous immune system workup, especially if the fever continues and the cultures remain negative. An HIV test, immunoglobulins, lymphocyte subsets, and vaccine responses all help determine where the immune system could go wrong.

Response to vaccinations can be difficult to interpret, especially in the young during the typical DTP priming schedule. Whenever you send these tests, ensure you get a detailed history of their vaccinations and when they’ve had them to help you interpret the results correctly.

So, we haven’t found an infection, and the immune system seems fine – what now?

If the host doesn’t seem to be the problem, we’re back to considering some more unusual causes of fever.

1. Atypical pathogens that are difficult to locate. These include mycobacteria, fungal, and parasitic infections that might not appear on our usual range of tests or cultures.
2. Inflammatory conditions like Kawasakis, primary HLH and juvenile idiopathic arthritis. These often have non-specific initial symptoms that can make them challenging to diagnose.
3. Malignancies like leukaemia and lymphoma
4.  Fever syndromes: PFAPA (Periodic Fever, Aphthous stomatitis, Pharyngitis, Adenitis) and Familial Mediterranean Fever syndrome. These are rare and will be low on the list, but they are an important consideration.  

That’s so many differentials – where do we start?

As in the Emergency Department, we prioritise ‘What will kill you faster?’

Just like following the A to E approach to find the problems that will kill you first, one way to approach a large list of differentials is to consider what is time-sensitive and what will be the most harmful to our patient.

You’ve already done some investigations that will help guide how you rank your differentials, as will the history you’ve gathered and any risk factors you’ve identified.

Travel and exposure history might point towards certain pathogens, even if the risk seems small.

If you haven’t taken a detailed history, now is the time.

We want to know about every relative with a cough, every encounter with new food, water sources, and animals—the weekend trip to the petting zoo, the cousin with pet rats, the wild swimming adventure—we want to know about it all.

Use your detective skills to narrow your differentials, then test based on probability and availability.

Let’s focus on infective causes for the fever

If broad-spectrum antibiotics have not cleared an infection (and it’s definitely an infection), it is either a resistant bacterium or a different type of pathogen.

Viruses

Viruses commonly cause fever, although the usual culprits are often self-limiting. Respiratory or gastrointestinal virus infections have typical symptoms and often a short course. You might be isolated in a nasal pharyngeal swab or a stool sample.

Some childhood viruses, such as chicken pox or measles, have typical symptoms or characteristic rashes that identify them. Awareness of the common presentations of these viruses helps to narrow the differential.

This comes with a caveat. Not every common illness will present as expected. If some of these have been ruled out earlier because they did not match the typical presentation, this might be the time to revisit them.

Fungi

We all know there are more categories of pathogens, but if you don’t encounter them often, it can be tricky to remember to include them in your differentials.

Fungi can cause local or disseminated infection. Mycetoma is a chronic skin infection caused by fungi such as Madurella mycetomatis. It presents insidiously, leading to a delayed diagnosis. Fungal testing can be challenging. Markers include beta-D-glucan and Aspergillus galactomannan. They are more helpful in confirming or monitoring cases rather than as a screening tool.

Parasites

Parasitic infections cause the largest burden of febrile illness in many areas of the world.

Malaria is endemic in 84 countries, causing an estimated 247 million cases in 2021. Depending on the species, it can remain latent for years in the liver or blood if incompletely treated. This means you must take a long travel history to cover potential exposure.

Gastrointestinal parasites like giardia and cryptosporidiosis can cause fever, usually alongside diarrhoea, nausea, and vomiting. This is more likely if the patient has drunk contaminated food or water.

It’s important to ask the right questions. A cursory ‘Did you eat anything unusual?’ might not trigger the correct answers. Ask about where they ate, how the food was prepared, whether there was undercooked meat and where the water came from.

Mycobacteria

Mycobacterium are a bacterial species worth considering in their own right – not just Mycobacterium tuberculosis. TB can present with prolonged fevers, and in 2022, there were an estimated 1.3 million paediatric cases.

The prevalence of non-tuberculous mycobacterial disease is increasing in North America, Europe and Australia. It can present with a spectrum of diseases, with cervical lymphadenitis being most common in children, coupled with low-grade fever and general malaise. With such non-specific symptoms and a 67% sensitivity rate on culture (the diagnosis gold standard), it can be a difficult pathogen to pin down.

When do we start to look for zebras rather than horses?

Of course, even the rarest conditions will occasionally pop up. Understanding them means you will have more luck pinpointing these without over-investigating every fever you encounter.

PFAPA is named for its presenting symptoms: Periodic Fever, Aphthous stomatitis, Pharyngitis, and Adenitis. The periodic fever is often the hallmark that leads to diagnosis once other causes have been excluded. Children are typically younger than five, with a fever lasting up to a weekend and occurring every two months or so. Crucially, these children will be well between episodes without impacting growth or development. Any positive family history or a correlation of fevers with throat pain at regular intervals should prompt you to consider it. Excluding all other diagnoses is the only way to confirm it.

Whilst PFAPA is rare, it is the most common paediatric periodic fever syndrome. Familial Mediterranean Fever (FMF), CAPS, TRAPS and Hyper-IgD are all even rarer genetic causes of period fever to consider.

The family history, whether of a formal diagnosis or just of having similar symptoms, is often the most helpful tool, alongside a well-kept fever diary with timing, duration and associated symptoms. Each condition has a specific phenotype that might be worth reviewing when presented with a tricky case of persistent fever.

That’s a lot of thinking for one fever – and we still don’t have an answer for our case!

Often, a fever is just the body’s innate immune response to a common pathogen. It’ll either fight it off alone or with the help of some antibiotics.

In the Emergency Department, our job is to work out who needs that help and who doesn’t, and that doesn’t always require much thinking! Some cases might throw up a few extra questions when they don’t quite fit your expected pattern.

Remember, every persistent and recurrent fever started as just a regular old fever. Sometimes, some clues might help you develop a wider differential early on.

The eight-week-old you saw in the ED with the fever that kept coming back? Sometimes, we never find the answer, but that doesn’t make the questions less important!

Selected references

Bouza E, Almirante B, García Rodríguez J, Garnacho-Montero J, Salavert M, Muñoz P, et al. Biomarkers of fungal infection: Expert opinion on the current situations. Revista Española de Quimioterapia. 2020 Jan 23;33(01):1–10.

Burstein B, Lirette MP, Beck C, Chauvin-Kimoff L, Chan K. Management of well-appearing febrile young infants aged ≤90 days. Paediatr Child Health. 2024 Feb 7;29(1):50–7.

Geneva: World Health Organization. .World Malaria Report 2022.  Licence:  CC BY-NC-SA 3.0 IGO.

Espin Diaz PC, Singh K, Kher P, Avanthika C, Jhaveri S, Saad Y, et al. Periodic Fever in Children: Etiology and Diagnostic Challenges. Cureus. 2022 Jul 25;

Meoli A, Deolmi M, Iannarella R, Esposito S. Non-Tuberculous Mycobacterial Diseases in Children. Pathogens. 2020 Jul 9;9(7):553.

NICE. Fever in under 5s: assessment and initial management [Internet]. [cited 2024 Mar 19]. Available from: https://www.nice.org.uk/guidance/ng143/chapter/Context

Shah P, Voice M, Calvo-Bado L, Rivero-Calle I, Morris S, Nijman R, et al. Relationship between molecular pathogen detection and clinical disease in febrile children across Europe: a multicentre, prospective observational study. The Lancet Regional Health – Europe. 2023 Sep;32:100682.

Wang A, Manthiram K, Dedeoglu F, Licameli GR. Periodic fever, aphthous stomatitis, pharyngitis, and adenitis (PFAPA) syndrome: A review. World J Otorhinolaryngol Head Neck Surg. 2021 Jul;7(3):166–73.

WHO. Global Tuberculosis Report 2023 [Internet]. [cited 2024 Mar 19]. Available from: https://www.who.int/publications/i/item/9789240083851

Wolff M, Bachur R. Serious Bacterial Infection in Recently Immunized Young Febrile Infants. Academic Emergency Medicine. 2009 Dec;16(12):1284–9.