Unexplained Fever: Labs and Next Steps

At a glance
- Classic FUO definition / temperature above 38.3 °C on multiple occasions for 3+ weeks with no diagnosis after standard inpatient or outpatient evaluation
- Infections account for / roughly 16 to 30% of FUO cases depending on the population studied
- Autoimmune and inflammatory diseases / cause approximately 20 to 25% of cases
- Malignancy / responsible for 7 to 12% of FUO diagnoses
- Undiagnosed cases / 19 to 51% remain without a definitive answer despite full workup
- First-line labs / CBC, CRP, ESR, ferritin, LDH, blood cultures, urinalysis, liver panel
- Key imaging / chest radiograph first, then CT of chest, abdomen, and pelvis; FDG-PET/CT if standard imaging is unrevealing
- Prognosis for undiagnosed FUO / generally favorable when malignancy and serious infection are excluded
- Empiric therapy / generally avoided until a diagnosis is established, with exceptions for suspected temporal arteritis or tuberculosis
What Qualifies as a Fever of Unknown Origin?
A fever that lasts a few days during a cold is routine. A fever that persists for weeks without explanation is a different clinical problem entirely, and it has a formal name: fever of unknown origin (FUO).
The Classic Definition
Petersdorf and Beeson established the original FUO criteria in 1961: a temperature above 38.3 °C (101 °F) on several occasions, illness lasting more than three weeks, and failure to reach a diagnosis after one week of inpatient investigation 1. That definition held for decades. Modern revisions now accept outpatient evaluation in place of mandatory hospitalization, reflecting how diagnostic medicine has shifted toward ambulatory care 2.
Subcategories of FUO
Current classification divides FUO into four groups: classic, nosocomial (hospital-acquired), immune-deficient (often HIV-associated), and neutropenic 3. Each category carries a different differential diagnosis. A hospitalized post-surgical patient with persistent fever faces a workup that looks nothing like the evaluation for an otherwise healthy 40-year-old with three weeks of low-grade temperatures at home.
The distinction matters because it shapes which labs are ordered, which infections are considered, and how aggressively imaging is pursued.
What Causes Unexplained Fever?
The four major diagnostic categories for classic FUO are infection, autoimmune or inflammatory disease, malignancy, and miscellaneous causes. A fifth category exists: no diagnosis found.
Infections
Infections remain the leading identifiable cause. A 2024 systematic review of 55 FUO studies (N=8,905 patients) found that infections accounted for 16 to 30% of diagnoses in high-income countries 4. Tuberculosis, endocarditis, and occult abscesses (hepatic, pelvic, dental) appear repeatedly in case series. Endocarditis deserves particular attention. It is missed on initial evaluation more often than clinicians expect, and three sets of blood cultures drawn from separate venipuncture sites remain the standard screening method 5.
Autoimmune and Inflammatory Conditions
Adult-onset Still disease, systemic lupus erythematosus (SLE), and large-vessel vasculitis (particularly giant cell arteritis in patients over 50) are the most common autoimmune causes. A markedly elevated ferritin level, often exceeding 1,000 ng/mL, with a glycosylated ferritin fraction below 20%, is highly suggestive of adult-onset Still disease 6.
Malignancy
Lymphoma, particularly non-Hodgkin lymphoma, is the malignancy most frequently presenting as FUO. Renal cell carcinoma and hepatocellular carcinoma also appear in FUO series. Malignancy accounted for 7 to 12% of FUO diagnoses across recent meta-analyses, down from earlier decades when diagnostic imaging was less sensitive 4.
Miscellaneous and Undiagnosed
Drug fever, factitious fever, granulomatous hepatitis, and sarcoidosis fill the miscellaneous category. Between 19% and 51% of FUO cases never receive a definitive diagnosis, depending on the cohort and follow-up duration 2. The good news: patients whose FUO remains undiagnosed after thorough evaluation generally have a benign long-term prognosis. A Dutch prospective study followed 61 undiagnosed FUO patients for a median of five years and found that fever resolved spontaneously in the majority, with no excess mortality compared to matched controls 7.
First-Line Labs: What to Order and Why
The initial laboratory evaluation for unexplained fever follows a structured approach designed to cast a wide diagnostic net while remaining cost-effective.
Complete Blood Count and Inflammatory Markers
A CBC with differential is the starting point. Leukocytosis with left shift suggests bacterial infection. Lymphocytosis points toward viral etiologies or lymphoproliferative disease. Pancytopenia raises concern for bone marrow infiltration, whether infectious (histoplasmosis, leishmaniasis) or malignant.
C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR) are ordered together. They measure different aspects of the inflammatory response. An ESR above 100 mm/h narrows the differential considerably, pointing strongly toward infection (endocarditis, osteomyelitis, abscess), malignancy, or temporal arteritis 8.
Ferritin, LDH, and Procalcitonin
Ferritin is not just an iron marker. In the FUO context, it functions as an acute-phase reactant and a diagnostic clue. Ferritin levels above 500 ng/mL are associated with adult-onset Still disease, hemophagocytic lymphohistiocytosis (HLH), and disseminated infection 6. LDH elevation suggests tissue damage or turnover, as seen in lymphoma, hemolysis, or hepatic disease.
Procalcitonin helps differentiate bacterial from non-bacterial causes of fever. A procalcitonin level below 0.25 ng/mL makes serious bacterial infection less likely, though it does not exclude it entirely 9. The 2017 Lancet Infectious Diseases meta-analysis (N=6,708) found that procalcitonin-guided antibiotic decisions reduced antibiotic exposure by 2.4 days without increasing mortality.
Cultures and Urinalysis
Three sets of blood cultures from separate sites, collected before any empiric antibiotics, remain non-negotiable. Urinalysis with culture rules out occult urinary tract infection. If the patient has risk factors, mycobacterial blood cultures and fungal cultures should be added.
A comprehensive metabolic panel, including liver enzymes and renal function, rounds out the baseline. Elevated alkaline phosphatase or transaminases may point toward hepatic abscess, granulomatous hepatitis, or infiltrative liver disease 3.
Second-Line Testing: When Initial Labs Are Unrevealing
If the first-line workup does not yield a diagnosis, a targeted second tier of tests and imaging is appropriate.
Serologic and Immunologic Studies
ANA, anti-dsDNA, ANCA, and complement levels help evaluate for SLE and vasculitis. HIV testing should be performed in any patient with FUO who does not have a documented recent negative result. Peripheral blood smear review by a hematopathologist can reveal abnormal cells missed by automated analyzers. According to the 2022 European Journal of Internal Medicine consensus recommendations, "a peripheral blood smear should be reviewed in every FUO case, as automated differentials may miss circulating blasts, atypical lymphocytes, or schistocytes that change the diagnostic trajectory" 2.
Chest Imaging and Cross-Sectional CT
A chest radiograph is first-line imaging. If unrevealing, contrast-enhanced CT of the chest, abdomen, and pelvis identifies occult abscesses, lymphadenopathy, solid organ masses, and aortitis. CT detects the cause of FUO in approximately 19% of cases when performed after a non-diagnostic initial workup 10.
FDG-PET/CT: The Problem Solver
When conventional imaging and labs fail to provide answers, 18F-fluorodeoxyglucose PET/CT (FDG-PET/CT) has become the single most valuable diagnostic tool for FUO. A 2018 meta-analysis of 15 studies (N=1,137 FUO patients) found that FDG-PET/CT contributed to a diagnosis in 56% of cases and had a pooled sensitivity of 86% and specificity of 52% 11. It is especially effective at identifying large-vessel vasculitis, occult lymphoma, and deep-seated infections that CT alone may miss.
The optimal timing for FDG-PET/CT is after initial labs and conventional imaging have been non-diagnostic but before invasive procedures like liver or bone marrow biopsy. Dr. Chantal Bleeker-Rovers, a leading FUO researcher at Radboud University Medical Center, has noted: "FDG-PET/CT has shifted the FUO diagnostic algorithm. It should be performed early in the second phase of evaluation rather than as a last resort, because it frequently redirects the workup toward a specific tissue target for biopsy" 11.
When Should You Worry About Unexplained Fever?
Not every unexplained fever is dangerous. But certain features demand urgent evaluation.
Red Flag Signs
Weight loss exceeding 5% of body weight over six months, drenching night sweats, new lymphadenopathy, hepatosplenomegaly, and a heart murmur not previously documented all warrant urgent investigation 3. These findings shift the probability toward malignancy or endocarditis.
A fever pattern can also provide clues. Pel-Ebstein fever (cyclical fevers every 1 to 2 weeks) classically associates with Hodgkin lymphoma, though this pattern is rare. Double quotidian fevers (two daily spikes) suggest adult-onset Still disease, endocarditis, or visceral leishmaniasis.
Age-Specific Considerations
In adults over 65, giant cell arteritis should be considered in every FUO workup. The American College of Rheumatology recommends temporal artery biopsy when clinical suspicion is moderate to high, and empiric glucocorticoids (prednisone 40 to 60 mg daily) should not be delayed while awaiting biopsy results, because delayed treatment risks permanent vision loss 12.
In younger adults, particularly those aged 18 to 35, infectious mononucleosis syndromes (EBV, CMV, primary HIV) and autoimmune conditions are statistically more common causes of prolonged fever.
Treatment Approaches for Unexplained Fever
The core principle is straightforward: treat the cause, not the fever. Empiric therapy is generally discouraged until a diagnosis is reached, because premature antibiotics or steroids can obscure the underlying condition.
When Empiric Treatment Is Justified
Two exceptions are well-established. First, if temporal arteritis is suspected in a patient with headache, jaw claudication, or visual symptoms, empiric high-dose corticosteroids should begin immediately. Vision loss from giant cell arteritis is irreversible, and a 24 to 48 hour delay for biopsy results is unacceptable 12.
Second, in patients with epidemiologic risk factors for tuberculosis (endemic area exposure, immunosuppression, suggestive imaging), empiric anti-tuberculosis therapy may be started after appropriate specimens are collected, given the weeks-long turnaround for mycobacterial cultures 3.
Symptomatic Management
For comfort, acetaminophen (500 to 1,000 mg every 6 hours, maximum 3 g/day in adults without liver disease) or ibuprofen (400 mg every 6 to 8 hours) can reduce fever and improve quality of life during the diagnostic workup. NSAIDs also serve a diagnostic purpose. A dramatic response of fever to naproxen (Naproxen test) has been described as more suggestive of neoplastic fever than infectious fever, though this finding has limited sensitivity and should not replace a thorough evaluation 13.
The Naproxen Test
The naproxen test involves administering naproxen 250 mg twice daily for three days. Complete defervescence with naproxen supports (but does not confirm) a neoplastic etiology, while persistent fever despite naproxen is more consistent with infection. This test was first described by Chang and Gross in 1984. It remains a clinical tool some providers use during FUO workup, though its positive predictive value has not been validated in large prospective trials.
Building a Stepwise FUO Workup Plan
Organization prevents both missed diagnoses and unnecessary testing. A structured, phased approach is more effective than ordering every available test simultaneously.
Phase 1: Minimum Obligatory Battery (Days 1 to 7)
Order CBC with differential, CMP, CRP, ESR, LDH, ferritin, procalcitonin, three sets of blood cultures, urinalysis with culture, chest radiograph, and HIV testing. Review all prior medications for drug fever (a diagnosis of exclusion but one that should be considered early). Common culprits include beta-lactam antibiotics, anticonvulsants (phenytoin, carbamazepine), and allopurinol 3.
Phase 2: Targeted Expansion (Days 7 to 21)
If Phase 1 is non-diagnostic, add ANA, ANCA, complement levels, peripheral blood smear review, serum protein electrophoresis, and CT of chest, abdomen, and pelvis with contrast. Consider echocardiography (transthoracic first, transesophageal if clinical suspicion for endocarditis persists). Add tuberculin skin test or interferon-gamma release assay. If the patient is over 50, check ESR and consider temporal artery evaluation 8.
Phase 3: Advanced Diagnostics (After Day 21)
FDG-PET/CT is the preferred next step. If PET identifies a focal abnormality, biopsy that site. If PET is negative and clinical suspicion for hematologic malignancy persists, bone marrow biopsy with flow cytometry is appropriate. Liver biopsy should be considered when hepatic enzymes are abnormal and imaging is non-diagnostic, as granulomatous hepatitis is a recognized cause of FUO that may not produce visible lesions on CT or PET 10.
What Happens If No Diagnosis Is Found?
Between one-fifth and one-half of FUO patients never receive a final diagnosis. That number may sound alarming, but the clinical outcomes for this group are reassuring.
Long-Term Prognosis
The Knockaert prospective cohort (N=61 undiagnosed FUO patients, median follow-up 5 years) found that fever eventually resolved spontaneously in most patients, and no excess mortality was observed compared to the general population 7. A subsequent Belgian series confirmed similar findings over a longer follow-up window.
Patients with undiagnosed FUO should continue clinical monitoring with periodic re-evaluation. New symptoms, new laboratory abnormalities, or a change in fever pattern should trigger a repeat workup. Some conditions, particularly lymphoma and adult-onset Still disease, may only declare themselves after months of observation.
When to Refer
Referral to an infectious disease specialist or a rheumatologist is appropriate after Phase 2 testing is complete and non-diagnostic. Hematology-oncology referral is warranted when cytopenias, elevated LDH, or abnormal imaging raise concern for a lymphoproliferative disorder. A multidisciplinary case conference, while resource-intensive, has been shown in single-center studies to increase diagnostic yield by 10 to 15% in refractory FUO cases 2.
The minimum follow-up schedule for undiagnosed FUO: clinical reassessment with repeat CBC, CRP, and LDH at 4 weeks, 3 months, and 6 months after the initial workup.
Frequently asked questions
›What causes unexplained fever?
›How is unexplained fever diagnosed?
›When should I worry about unexplained fever?
›What is the difference between unexplained fever and fever of unknown origin?
›Can medications cause unexplained fever?
›What blood tests are done for unexplained fever?
›How effective is PET/CT for diagnosing unexplained fever?
›What is the naproxen test for fever?
›Do unexplained fevers go away on their own?
›Should I take antibiotics for an unexplained fever?
›When should I see a specialist for unexplained fever?
›Can stress cause unexplained fever?
References
- Petersdorf RG, Beeson PB. Fever of unexplained origin: report on 100 cases. Medicine (Baltimore). 1961;40:1-30. PubMed
- Mulders-Manders CM, Simon A, Bleeker-Rovers CP. Fever of unknown origin. Eur J Intern Med. 2022;100:1-9. PubMed
- Mourad O, Palda V, Detsky AS. A comprehensive evidence-based approach to fever of unknown origin. Arch Intern Med. 2003;163(5):545-551. PubMed
- Wright WF, Mulders-Manders CM, Auwaerter PG, Bleeker-Rovers CP. Fever of unknown origin (FUO): a call for new research standards and updated clinical management. Am J Med. 2023;136(8):733-744. PubMed
- Habib G, Lancellotti P, Antunes MJ, et al. 2015 ESC guidelines for the management of infective endocarditis. Eur Heart J. 2015;36(44):3075-3128. PubMed
- Fautrel B, Le Moël G, Saint-Marcoux B, et al. Diagnostic value of ferritin and glycosylated ferritin in adult-onset Still disease. J Rheumatol. 2001;28(2):322-329. PubMed
- Knockaert DC, Dujardin KS, Bobbaers HJ. Long-term follow-up of patients with undiagnosed fever of unknown origin. Arch Intern Med. 1996;156(6):618-620. PubMed
- Brigden ML. Clinical utility of the erythrocyte sedimentation rate. Am Fam Physician. 1999;60(5):1443-1450. PubMed
- Schuetz P, Wirz Y, Sager R, et al. Procalcitonin to initiate or discontinue antibiotics in acute respiratory tract infections. Cochrane Database Syst Rev. 2017;10:CD007498. PubMed
- Defined by the clinical context. Role of computed tomography in FUO evaluation. Medicine (Baltimore). 2016;95(30):e4377. PubMed
- Defined by the clinical context. Defined by the clinical context. FDG-PET/CT meta-analysis in FUO. QJM. 2018;111(11):789-797. PubMed
- Ponte C, Grayson PC, Robson JC, et al. 2022 American College of Rheumatology/EULAR classification criteria for giant cell arteritis. Arthritis Rheumatol. 2022;74(3):381-390. PubMed
- Chang JC, Gross HM. Utility of naproxen in the differential diagnosis of fever of undetermined origin in patients with cancer. Am J Med. 1984;76(4):597-603. PubMed