What is the comprehensive algorithm for evaluating, investigating, and treating fever in adults, including history, physical exam, red‑flag criteria, work‑up for low‑risk, high‑risk, and prolonged fever, and appropriate empiric therapy?

Fever Work-Up Algorithm, Investigations, and Treatment in Adults

Initial Assessment and Red Flags

Begin with accurate temperature measurement using rectal or oral thermometry (≥38°C/100.4°F); avoid unreliable methods such as axillary, tympanic, or temporal-arterial measurements. 1, 2

Critical Red Flags Requiring Immediate Empiric Antibiotics

Empiric antibiotics must be initiated immediately in the following scenarios, before completing the full diagnostic work-up 1, 3:

• Hemodynamic instability or septic shock (hypotension requiring vasopressors, altered mental status, respiratory distress) 1, 3
• Neutropenia (absolute neutrophil count <500 cells/µL or <1000 cells/µL with expected decline to <500) 3
• Suspected bacterial meningitis (altered mental status, meningeal signs, petechial rash) 1
• Suspected cholangitis (fever, jaundice, right upper quadrant pain) 1
• Suspected tick-borne rickettsial disease (appropriate exposure history with fever and rash) 1

High-Risk Clinical Features

Age ≥50 years combined with leukocyte count ≥15×10⁹/L predicts 36% incidence of serious illness; these patients require careful evaluation and consideration for hospitalization 4

---

Focused History: High-Yield Elements

Document the following systematically 1, 2:

• All indwelling devices with insertion dates (central venous catheters, urinary catheters, endotracheal tubes, surgical drains, IV lines) 1, 2
• Recent procedures, surgeries, or hospitalizations within the past 60 days 2
• Complete medication review for drug-induced fever (especially antibiotics, chemotherapy, anesthetics, neuroleptics, SSRIs) 1, 2
• Immunocompromising conditions (diabetes, malignancy, transplant, HIV, chronic steroids) 2
• Travel history within the past year to tropical or subtropical regions 5, 3
• Timing of fever onset relative to any surgical procedure (immediate = atelectasis; days 3-5 = pneumonia, UTI, catheter infection) 2

---

Targeted Physical Examination

Systematically examine 1, 2:

• Respiratory system: auscultate for crackles, diminished breath sounds, rapid breathing 6
• Skin and mucous membranes: look for petechiae, rash, cyanosis, poor peripheral circulation 3, 6
• All catheter insertion sites: erythema, purulence, tenderness 1, 2
• Abdomen: tenderness, hepatosplenomegaly, surgical scars 3, 2
• Neurological status: altered consciousness, meningeal irritation, seizures 3, 6
• "Silent sources": otitis media, decubitus ulcers (sacrum, back, head), perianal/perineal abscesses, retained foreign bodies (tampons) 5, 2

---

Laboratory Investigations

Blood Cultures: The Cornerstone

Obtain at least two blood culture sets (≈60 mL total) from separate anatomical sites BEFORE initiating antibiotics. 1, 3

• For patients with central venous catheters: draw simultaneous central and peripheral cultures to calculate differential time-to-positivity (≥2 hours suggests catheter-related bloodstream infection) 1
• Sample at least two lumens of any central line to increase diagnostic yield 1
• Skin antisepsis: use 2% chlorhexidine in 70% isopropyl alcohol with mandatory 30-second drying period 1

Complete Blood Count with Differential

• Band neutrophil count >1,500 cells/µL yields likelihood ratio of 14.5 for bacterial infection 1
• Neutrophil proportion >90% yields likelihood ratio of 7.5 for bacterial infection 1
• Age ≥50 years plus leukocyte count ≥15×10⁹/L identifies high-risk patients 4

Biomarkers: Procalcitonin and C-Reactive Protein

Use procalcitonin or CRP only when pre-test probability of bacterial infection is low-to-intermediate; these markers help rule out infection and guide antibiotic discontinuation, not initial diagnosis. 1, 2

• Do not rely on negative biomarkers when clinical suspicion is high—they do not exclude infection 1
• Procalcitonin cutoffs: 0.5 ng/mL for bacterial infection; 2-10 ng/mL for severe sepsis; >10 ng/mL for septic shock 5
• Endotoxin activity assay has 98.6% negative predictive value for Gram-negative infection 5

Respiratory and Urinary Testing

• For suspected pneumonia or new respiratory symptoms: order viral nucleic acid amplification panels plus bacterial cultures 1
• SARS-CoV-2 PCR when community transmission is ongoing 1
• For suspected UTI with pyuria: replace the indwelling catheter and culture from the new catheter, not the old one 1

Additional Laboratory Tests

• Liver function tests, urinalysis, urine culture 3
• Consider malaria testing for all patients who visited tropical countries within 1 year 3

---

Imaging Strategy

First-Line Imaging

Obtain a chest radiograph on all febrile patients—pneumonia is the leading infectious cause of fever 2

Targeted Imaging Based on Clinical Suspicion

• CT chest with IV contrast identifies pulmonary source in ≈72% of surgical-ICU patients when chest X-ray is nondiagnostic 1
• Abdominal ultrasound for abdominal pain, abnormal liver tests, or recent abdominal surgery 1, 2
• CT abdomen/pelvis with IV contrast has 81.8% positive predictive value for septic foci (most commonly abdomen/pelvis/genitourinary tract) 1
• CT of the operative region for fever occurring several days after thoracic, abdominal, or pelvic surgery without another explanation 1

Imaging to Avoid

• Avoid routine abdominal imaging in patients lacking abdominal signs, symptoms, or liver-function abnormalities 1
• Avoid routine sinus CT in prolonged febrile neutropenia without localizing symptoms—findings are often nondiscriminatory 1

---

Advanced Imaging for Prolonged Fever of Unknown Origin

When initial work-up is inconclusive, ¹⁸F-FDG PET/CT provides the highest diagnostic yield with sensitivity of 84-86% and overall yield of ≈56%. 1

• Perform PET/CT within 3 days of initiating oral glucocorticoids if steroids are required 1
• Early PET/CT use is cost-effective compared with prolonged empiric testing strategies 1
• A negative PET/CT predicts favorable prognosis and supports watchful waiting in selected patients 1

---

Non-Infectious Causes: Always Consider

Medication-Related Fever

• Drug fever typically appears around 21 days after drug initiation; resolves within 1-7 days after discontinuation 1, 2
• Malignant hyperthermia can develop up to 24 hours after exposure to succinylcholine or halogenated anesthetics 5, 1
• Neuroleptic malignant syndrome (e.g., haloperidol) presents with muscular rigidity and elevated creatine phosphokinase 5, 1
• Serotonin syndrome (e.g., SSRIs, linezolid) manifests with autonomic instability and neuromuscular hyperactivity 1
• Withdrawal syndromes (alcohol, opioids, benzodiazepines, barbiturates) produce fever, tachycardia, diaphoresis 1

Other Non-Infectious Etiologies

• Vascular: venous thrombosis, pulmonary embolism, myocardial infarction 1, 2
• Inflammatory: gout, pancreatitis, Dressler syndrome, transplant rejection 1, 2
• Endocrine: thyroid storm, adrenal insufficiency 1, 2
• Malignancy: tumor fever, cytokine release syndrome 2
• Transfusion-related: CMV reactivation (presents 1 month post-transfusion with high fever, pancytopenia, atypical lymphocytosis) 5

---

Empiric Antibiotic Therapy: When and What

Stable Patients Without Red Flags

In clinically stable adults, empirical antibiotics should be avoided because they can obscure the underlying diagnosis and cause harm; therapeutic decisions must be driven by targeted clinical assessment. 1, 2

Unstable Patients or Red-Flag Scenarios

Initiate empiric antibiotics within 1 hour after diagnosis of sepsis, immediately after obtaining blood cultures. 3

• Target likely pathogens based on suspected source, risk for multidrug-resistant organisms, and local antimicrobial susceptibility patterns 3
• For neutropenic fever (ANC <100 cells/µL expected >7 days): empiric β-lactam monotherapy (e.g., piperacillin-tazobactam) plus vancomycin 1, 3
• For suspected catheter-related bloodstream infection: do not routinely remove central venous catheters; removal is reserved for clinically unstable patients or microbiologic evidence of catheter infection 1

---

Special Populations

Neutropenic Patients

• Hospitalization and empiric therapy with vancomycin plus antipseudomonal antibiotics is mandatory 3
• Daily surveillance (physical exam, review of systems, targeted cultures) to detect emerging infections early 1
• Maintain high suspicion for opportunistic infections such as CMV reactivation, which may present with fever and normal white-blood-cell count 5, 1

Returning Travelers

• Detailed travel history including all geographical locations visited 5, 3
• Malaria testing for all patients who visited tropical countries within 1 year of presentation 3
• Consider rapidly fatal tropical diseases (e.g., dengue, typhoid, rickettsial infections) 5

Critically Ill Patients

• Thorough evaluation for "silent" sources of infection 3
• Consider invasive diagnostic procedures (aspiration/biopsy of skin and soft tissue lesions) if non-invasive tests are unrevealing 3
• Use core temperature measurement (pulmonary artery, bladder, esophageal) rather than peripheral methods 1

---

Common Pitfalls to Avoid

• Automatic "fever workup" panels lead to unnecessary testing, blood loss, radiation exposure, and patient transport risks 1, 2
• Assuming infection without clinical evidence—up to 75% of fever-of-unknown-origin cases resolve spontaneously 1, 2
• Starting empiric antibiotics in stable patients—this can mask the underlying diagnosis and may be harmful 1, 2
• Overlooking "silent sources" such as otitis media, decubitus ulcers, perianal abscesses, retained foreign bodies 5, 2
• Using unreliable temperature methods (oral, tympanic, temporal artery) when accurate assessment is critical 1, 2
• Routine removal of central venous catheters without microbiologic evidence or clinical instability 1