Red flags
- Neutropenia (known or chemotherapy history) -> febrile neutropenia pathway, do not wait for a focus to emerge
- Haemodynamic instability, altered mental state -> sepsis pathway regardless of source found
- Non-blanching rash + fever -> meningococcaemia until excluded
- Returned traveller with fever -> malaria until excluded (can kill within days if missed)
- Immunosuppression (HIV, transplant, biologics, splenectomy) - broadens differential and lowers threshold for empirical treatment
- New murmur + fever -> endocarditis
Differential by mechanism
- Occult UTI/pyelonephritis, early pneumonia, intra-abdominal collection, endocarditis, dental/sinus source, occult line infection
- Malignancy (lymphoma, renal cell, hepatocellular), drug fever, connective tissue disease/vasculitis, thromboembolism (PE/DVT can present as fever), thyrotoxicosis
- Neutropenic: bacterial translocation from gut/skin flora - treat empirically, do not wait for a focus
- Returned traveller: malaria, typhoid, dengue, rickettsial disease - destination and exposure history changes the entire differential
- Immunosuppressed/HIV: opportunistic infection expands the differential substantially (see HIV-related notes)
- Post-operative: the "5 Ws" - Wind (pneumonia/atelectasis), Water (UTI), Wound, Walking (DVT/PE), Wonder drugs (drug fever) - timed roughly to days post-op
Focused history
- Fever pattern, duration, associated symptoms (even subtle - night sweats, weight loss, localising pain)
- Travel history - destination, dates, prophylaxis, exposures (fresh water, animals, food, sexual contacts)
- Immunosuppression - HIV status, chemotherapy, biologics, transplant, splenectomy, recent steroid use
- Recent procedures, hospitalisation, antibiotic use, indwelling devices
- Animal/occupational/recreational exposures (farms, ticks, unpasteurised products)
- Sexual history, IV drug use
- Full medication review - drug fever is a diagnosis of exclusion but common
Focused examination
- Full systematic exam - skin (rash, petechiae, splinter haemorrhages), lymphadenopathy, cardiac auscultation (new murmur), abdominal exam, joints
- Line sites, surgical wounds, pressure areas
- Fundoscopy (Roth spots), nail beds, palms/soles (endocarditis stigmata)
- Lymphoreticular exam (hepatosplenomegaly) if malignancy/systemic infection suspected
- Repeat examination if initial exam unremarkable - findings can evolve
Investigation strategy
- First-line: FBE with differential, UEC, LFT, CRP, blood cultures (x2 sets), urine MCS, CXR
- Directed by exposure history: malaria thick/thin films (repeat x3 if negative and suspicion remains), blood culture held longer if suspecting fastidious organisms (endocarditis - HACEK), serology per travel/exposure
- Echocardiography if new murmur or unexplained bacteraemia
- CT imaging (chest/abdomen/pelvis) if no focus found after initial work-up and patient remains unwell
- Do not over-image or over-culture a well patient who defervesces spontaneously - most short-lived undifferentiated fevers are self-limiting viral illness
Management
1. Risk-stratify first - neutropenic, septic, or high-risk exposure (malaria) patients get empirical treatment immediately, not sequential work-up
2. Well, immunocompetent patient with no red flags: symptomatic management and safety-netting is often appropriate while first-line investigations return - most resolve spontaneously (viral)
3. Persistent fever without a focus (>1 week) despite initial work-up: consider the diagnosis as pyrexia of unknown origin (PUO) - escalate imaging, consider malignancy/rheumatological work-up, infectious diseases referral
- Suspected malaria: treat empirically if severe/high-risk exposure while confirmation pending, per antimalarial guidelines
- Suspected meningococcaemia: immediate IV ceftriaxone/benzylpenicillin before any confirmatory test
- Neutropenic fever: immediate broad-spectrum antibiotics (see Febrile neutropenia note)
- Otherwise avoid "blind" broad-spectrum antibiotics in a stable patient without a source - masks culture yield and drives resistance without proven benefit
- Safety-netting advice and early review if fever persists or new symptoms develop
- Specialist infectious diseases input for prolonged/unexplained fever
Traps
- Missing malaria in a returned traveller because fever pattern was atypical or rapid antigen test was falsely negative on a single sample
- Starting broad-spectrum antibiotics reflexively in a well patient with undifferentiated fever, obscuring the diagnosis and driving resistance
- Missing neutropenia because a recent FBE was not checked in a chemotherapy patient presenting with fever
- Attributing fever to a urinary source based on asymptomatic bacteriuria alone
- Not repeating examination/investigations when the initial work-up is unrevealing but the patient remains unwell
Talk track
Risk-stratify before investigating in detail - neutropenia, sepsis, or a high-risk travel/exposure history (malaria, meningococcaemia) trigger immediate empirical treatment rather than sequential work-up. In a well, immunocompetent patient with no red flags, most undifferentiated fever is self-limiting viral illness and does not need blind broad-spectrum antibiotics. Persistent fever beyond about a week despite first-line work-up becomes a pyrexia of unknown origin problem needing broader imaging and specialist input rather than repeating the same tests.
8 of 8 sections written · drafted 2026-09-13