Differentiating Infectious from Non-Infectious Fever in Critical Care: A Clinical Approach

Dr Neeraj Manikath , Claude.ai

Abstract

Fever in critically ill patients represents a diagnostic and therapeutic challenge with significant implications for patient outcomes. While infection remains the primary consideration, numerous non-infectious etiologies can produce fever in the intensive care unit (ICU). The indiscriminate use of empiric antimicrobials contributes to antimicrobial resistance, adverse drug effects, and increased healthcare costs. This review provides a systematic approach to differentiating infectious from non-infectious causes of fever in critically ill patients, with emphasis on diagnostic strategies, clinical reasoning, and evidence-based management.

Introduction

Fever, defined as a core body temperature ≥38.3°C (101°F), occurs in 26-70% of ICU patients depending on the population studied.¹ The traditional reflex to treat all fever with antimicrobials has been challenged by mounting evidence that fever may be protective in certain contexts and that non-infectious causes account for up to 50% of febrile episodes in critically ill patients.²,³

The stakes are high: delayed antimicrobial therapy in sepsis increases mortality by approximately 7.6% per hour, yet unnecessary antimicrobial exposure drives resistance and increases complications.⁴ This review synthesizes current evidence to guide clinicians through this diagnostic dilemma.

Pathophysiology of Fever

Infectious Fever (True Fever)

Infectious fever results from cytokine-mediated upward adjustment of the hypothalamic thermoregulatory set point. Pathogen-associated molecular patterns (PAMPs) trigger release of endogenous pyrogens—particularly interleukin-1 (IL-1), interleukin-6 (IL-6), and tumor necrosis factor-alpha (TNF-α)—which stimulate prostaglandin E2 (PGE2) synthesis in the hypothalamus, raising the temperature set point.⁵

Pearl: The presence of rigors strongly suggests infectious fever, as the shivering response is triggered by the hypothalamus attempting to reach the elevated set point.

Non-Infectious Fever

Non-infectious fever can result from:

Sterile inflammation: Surgery, trauma, pancreatitis (release of damage-associated molecular patterns, or DAMPs)
Drug reactions: Direct hypothalamic effects or hypersensitivity reactions
Central fever: Direct hypothalamic injury
Hyperthermia: Failure of heat dissipation mechanisms (distinct from true fever)

Oyster: Hyperthermia (heat stroke, malignant hyperthermia, neuroleptic malignant syndrome, serotonin syndrome) represents a fundamentally different process where the thermoregulatory set point remains normal but heat dissipation fails. These conditions DO NOT respond to antipyretics and require different management.

Epidemiology of Fever in the ICU

Studies demonstrate variable proportions of infectious versus non-infectious fever:

Medical ICU: 60-75% infectious⁶
Surgical ICU: 40-50% infectious (higher proportion of non-infectious causes)⁷
Neurological ICU: 30-50% infectious⁸
Post-cardiac surgery: Up to 80% non-infectious in first 48 hours⁹

Hack: The "5 W's" mnemonic (Wind-pneumonia, Water-UTI, Wound-surgical site infection, Walking-DVT, Wonder drugs-drug fever) remains useful but was developed for surgical wards. In the ICU, add "Wires" (catheter-related bloodstream infection) and "White cells" (acalculous cholecystitis, sinusitis).

Clinical Approach to Fever Differentiation

History and Timing

Time of onset relative to ICU admission or intervention:

<48 hours post-admission: Consider community-acquired infections, aspiration, pre-existing infection
48-96 hours: Transition period; hospital-acquired infections emerge
>96 hours: Primarily hospital-acquired infections, non-infectious causes more likely

**Post-operative fever timeline:**¹⁰

Day 0-2: Surgical trauma, atelectasis, transfusion reactions (rarely infection)
Day 3-5: Pneumonia, urinary tract infection
Day 5-7: Surgical site infection, Clostridioides difficile infection
>Day 7: Deep abscess, anastomotic leak, prosthetic infection

Pearl: Fever in the immediate post-operative period (0-48 hours) is usually non-infectious. The classic teaching that "atelectasis causes fever" has been challenged—most post-operative fever in this window results from surgical inflammatory response.¹¹

Physical Examination

Signs favoring infection:

Hemodynamic instability with fever
Rigors or severe chills
New or changing infiltrate on lung examination
Purulent secretions (sputum, wound drainage, urine)
Localized signs: erythema, warmth, fluctuance
New heart murmur (endocarditis)

Signs favoring non-infectious causes:

Stable hemodynamics despite high fever
Absence of rigors
Diffuse erythroderma (drug reaction, toxic shock)
Surgical emphysema tracking beyond wound margins
Calf swelling/erythema (DVT—though DVT rarely causes fever >38.5°C)

Oyster: Central fever (neurogenic fever) following brain injury presents with:

Temperature >38.5°C often >39.5°C
Absence of diurnal variation
Elevated temperature unresponsive to antipyretics
Tachycardia often absent despite high fever (uncoupling of temperature-heart rate relationship)
Normal or low inflammatory markers¹²

Laboratory Evaluation

Initial laboratory assessment:

Tests supporting infection:

White blood cell count with differential (leukocytosis, bandemia, leukopenia)
C-reactive protein (CRP) elevation (>100 mg/L highly suggestive)
Procalcitonin elevation (>0.5 ng/mL)¹³
Lactate elevation (tissue hypoperfusion)
New or worsening organ dysfunction

Hack: Procalcitonin algorithms for antibiotic stewardship:

<0.25 ng/mL: Infection unlikely, consider discontinuation
0.25-0.5 ng/mL: Infection possible, clinical judgment needed
0.5-2.0 ng/mL: Infection likely
2.0 ng/mL: Severe bacterial infection or sepsis highly likely¹⁴

Caveats for procalcitonin:

Elevated in: severe trauma, surgery (first 48 hours), cardiogenic shock, severe pancreatitis, post-cardiac arrest
May be normal in: localized infections, viral infections, fungal infections, early sepsis
Unreliable in: renal failure (unless CRRT employed), immunocompromised patients

Tests supporting non-infectious causes:

Eosinophilia (drug fever, though absence doesn't exclude it)
Markedly elevated transaminases (drug-induced hepatitis)
Elevated CK (rhabdomyolysis, neuroleptic malignant syndrome)
Normal procalcitonin with elevated CRP (sterile inflammation)

Pearl: The combination of normal procalcitonin (<0.25 ng/mL) with elevated CRP suggests non-infectious inflammation. CRP rises in both infectious and non-infectious inflammation; procalcitonin is more specific for bacterial infection.¹⁵

Microbiological Investigations

Culture strategy:

Before initiating antimicrobials, obtain:

Blood cultures × 2 sets from separate sites (peripheral and central if applicable)
Respiratory cultures: Endotracheal aspirate or bronchoalveolar lavage if intubated
Urine culture: If urinary symptoms or abnormal urinalysis
Wound cultures: From any suspicious surgical sites
Line tip cultures: Only if catheter removed for suspected infection

Hack: The "blood culture stewardship" approach:

Obtain cultures for: new fever >38.3°C, hemodynamic instability, suspected endocarditis, neutropenia
Avoid reflexive cultures for: chronic low-grade temperature elevation, known non-infectious causes, isolated single temperature spike without clinical change¹⁶

Oyster: Positive blood cultures from central lines may represent contamination or catheter colonization rather than true bloodstream infection. Use quantitative cultures when available: colony counts ≥10-fold higher from central versus peripheral sites suggest catheter-related bloodstream infection.¹⁷

Advanced Diagnostic Imaging

Computed tomography (CT):

CT chest: Pneumonia, empyema, pulmonary embolism
CT abdomen/pelvis: Intra-abdominal abscess, acalculous cholecystitis, colitis, mesenteric ischemia
CT sinuses: Nosocomial sinusitis (especially in nasally intubated patients)

Ultrasound:

Bedside cardiac ultrasound: Vegetations (low sensitivity), pericardial effusion
Abdominal ultrasound: Cholecystitis, hepatic/splenic abscess
Lower extremity Doppler: Deep vein thrombosis

Nuclear medicine:

18F-FDG PET/CT: Occult infection or malignancy when source unclear¹⁸
Indium-111 white blood cell scan: Suspected but unlocalized infection

Pearl: In patients with fever of unknown origin in the ICU, whole-body CT or PET/CT imaging has diagnostic yield of 40-60% when standard investigations are unrevealing.¹⁹

Specific Non-Infectious Causes

Drug-Induced Fever

Incidence: 3-7% of all adverse drug reactions²⁰

Common causative agents:

Antimicrobials: β-lactams (especially penicillins), sulfonamides, nitrofurantoin
Cardiovascular: Procainamide, quinidine, methyldopa
Anticonvulsants: Phenytoin, carbamazepine
Others: Allopurinol, H2-blockers, PPIs, heparin

Clinical features:

Onset typically 7-10 days after drug initiation (can be months)
Fever often high (>39°C) with relative bradycardia
Eosinophilia (only present in 20-25%)²¹
Rash (18-20%)
Resolution typically within 48-72 hours of drug cessation

Hack: If drug fever suspected, systematically discontinue non-essential medications starting with those most recently initiated. Consider the "rechallenge" test only for essential medications, as fever recurs within hours if the drug is causative.

Venous Thromboembolism

DVT alone rarely causes fever >38.5°C. However:

Pulmonary embolism may cause fever through cytokine release from pulmonary infarction
Septic thrombophlebitis (suppurative thrombophlebitis) causes high fever with positive blood cultures²²

Pearl: Consider catheter-associated septic thrombophlebitis in patients with persistent bacteremia despite appropriate antimicrobials and catheter removal. Diagnosis requires imaging (CT/ultrasound showing thrombus) and often requires anticoagulation plus antimicrobials for 4-6 weeks.

Transfusion Reactions

Febrile non-hemolytic transfusion reaction (FNHTR):

Most common transfusion reaction (1-3% of transfusions)
Temperature rise ≥1°C during or within 4 hours of transfusion
Managed with antipyretics; leukoreduction of future transfusions²³

Transfusion-related acute lung injury (TRALI):

Occurs within 6 hours of transfusion
Fever with acute respiratory distress and bilateral infiltrates
Requires supportive care; often resolves within 48-96 hours

Septic transfusion reaction:

Rare but serious (bacterial contamination, especially platelets stored at room temperature)
High fever, rigors, severe hemodynamic instability
Requires immediate cessation, cultures of patient and product, broad-spectrum antimicrobials

Pancreatitis

Acute pancreatitis causes sterile inflammation with fever in 70-85% of cases. Fever alone does NOT indicate infected pancreatic necrosis.²⁴

Distinguishing sterile from infected necrosis:

Clinical deterioration or persistent organ failure beyond first week
Procalcitonin may help but lacks specificity
CT-guided fine needle aspiration with Gram stain and culture (gold standard)
Antimicrobial prophylaxis NOT recommended for sterile necrosis²⁵

Hack: In acute pancreatitis, defer antimicrobials unless there is: (1) clinical deterioration with high suspicion of infection, (2) positive cultures, or (3) extrapancreatic infection source. Early antimicrobial use does not prevent infected necrosis and may promote resistant organisms and fungal superinfection.

Adrenal Insufficiency

Relative adrenal insufficiency in critical illness may present with fever along with:

Refractory hypotension
Hyponatremia
Hyperkalemia
Eosinophilia
Hypoglycemia

Diagnosis: Random cortisol <10 μg/dL or inadequate response to ACTH stimulation test (though interpretation controversial in sepsis)²⁶

Central (Neurogenic) Fever

Occurs in 4-37% of patients with acute brain injury (traumatic brain injury, subarachnoid hemorrhage, intracerebral hemorrhage, ischemic stroke).²⁷

Diagnostic criteria (diagnosis of exclusion):

Timing: Fever within 72 hours of brain injury
High fever: Usually >38.5°C, often >39.5°C
Persistence: Sustained elevation without diurnal variation
Antipyretic resistance: Minimal response to acetaminophen or NSAIDs
Absence of infection: Negative cultures, normal procalcitonin
Neuroanatomic correlation: Hypothalamic or brainstem injury on imaging

Management:

External cooling (surface cooling, intravascular cooling catheters)
Avoid excessive shivering (counterproductive; use sedation/paralysis if necessary)
Avoid aggressive pharmacologic temperature reduction (limited efficacy)

Pearl: Central fever is a diagnosis of exclusion. Thoroughly exclude infection before attributing fever to brain injury alone, as co-infection is common in neurocritical care patients.

Malignancy

Malignancy-related fever characteristics:

Often cyclical (Pel-Ebstein pattern in lymphoma)
Night sweats, weight loss
Elevated LDH
Resolution with NSAIDs (particularly naproxen)²⁸

Tumors commonly causing fever:

Hematologic malignancies (lymphoma, leukemia)
Renal cell carcinoma
Hepatocellular carcinoma
Atrial myxoma

Other Non-Infectious Causes

Alcohol withdrawal: Fever may occur in severe withdrawal/delirium tremens

Thyroid storm: High fever (often >40°C) with tachycardia, agitation, diaphoresis

Adrenal crisis: As above

Gout/pseudogout: Crystal-induced inflammation may cause fever with joint inflammation

Hematoma resorption: Large hematomas (retroperitoneal, intramuscular) cause fever during resorption

Antimicrobial Stewardship Principles

When to Initiate Empiric Antimicrobials

Strong indications (treat immediately):

Sepsis or septic shock (qSOFA ≥2, hypotension, lactate ≥2 mmol/L)²⁹
Neutropenic fever
Suspected meningitis/encephalitis
Necrotizing soft tissue infection
Suspected catheter-related bloodstream infection with hemodynamic instability

Weak indications (observe, investigate, consider clinical context):

Isolated fever without hemodynamic compromise or organ dysfunction
Post-operative day 0-2
Known non-infectious cause likely
Central fever in neurocritical care patient with normal inflammatory markers

Hack: The "1-hour bundle" for sepsis management emphasizes rapid antimicrobial administration, but this should not override clinical judgment. In stable patients without clear infection source, brief observation (4-6 hours) while completing diagnostic workup is reasonable and may prevent unnecessary antimicrobial exposure.³⁰

De-escalation and Duration

Procalcitonin-guided antimicrobial discontinuation:

Discontinue if procalcitonin decreases by >80% from peak or to <0.5 ng/mL
Meta-analyses show reduced antimicrobial duration without increased mortality¹⁴

Culture-guided de-escalation:

Narrow spectrum based on culture and susceptibility results (typically by 48-72 hours)
Discontinue antimicrobials if cultures negative and alternative diagnosis established

Duration principles:

Uncomplicated infections: Shorter courses (5-7 days) often adequate³¹
Complicated infections: 7-14 days typically sufficient
Avoid prolonged courses "just in case"

Management of Non-Infectious Fever

Antipyretic Therapy

Evidence for fever treatment:

No mortality benefit demonstrated for routine antipyresis in critically ill patients³²
HEAT trial (2015): Paracetamol vs. placebo in ICU patients showed no difference in ICU-free days³³
Potential benefits: Reduced metabolic demand, improved comfort
Potential harms: Masking infection, hepatotoxicity (acetaminophen), bleeding/renal injury (NSAIDs)

Recommendations:

Consider antipyretics for: patient comfort, excessive metabolic demand (severe cardiac/pulmonary disease), neurologic injury with intracranial hypertension
Avoid routine antipyresis in hemodynamically stable patients without above indications
Acetaminophen: 650-1000 mg q6h (maximum 4 g/day; reduce in hepatic impairment)
NSAIDs: Generally avoid in critically ill patients due to bleeding/renal risks

Physical Cooling

Indications:

Hyperthermia syndromes (temperature >40°C)
Central fever refractory to antipyretics
Severe intracranial hypertension

Methods:

External cooling: Cooling blankets, ice packs, cooling fans
Intravascular cooling: Catheter-based systems (more effective, more invasive)
Evaporative cooling: Tepid water misting with fans

Hack: When using external cooling, monitor for shivering (counterproductive, increases metabolic rate). Manage shivering with: buspirone (15-30 mg enterally), magnesium sulfate (4-6 g IV), dexmedetomidine (0.2-0.7 μg/kg/h), or neuromuscular blockade if intubated.³⁴

Diagnostic Algorithm: A Practical Approach

Step 1: Initial Assessment

Hemodynamically stable? Organ dysfunction present?
Timing of fever relative to admission/interventions?
Risk factors for infection vs. non-infectious causes?

Step 2: Basic Investigations

CBC with differential, CRP, procalcitonin, lactate
Basic metabolic panel, liver function tests
Urinalysis
Chest radiograph

Step 3: Risk Stratification

HIGH RISK (treat empirically):

Septic shock
Neutropenia
Immunosuppression
Procalcitonin >2 ng/mL with hemodynamic instability

INTERMEDIATE RISK (culture before treating, short observation acceptable):

Fever with localizing signs but hemodynamically stable
Procalcitonin 0.5-2.0 ng/mL

LOW RISK (observe, investigate, hold antimicrobials):

Isolated fever, hemodynamically stable
Procalcitonin <0.5 ng/mL
Clear alternative diagnosis (drug fever, central fever, post-operative day 0-2)

Step 4: Source Control

Remove potentially infected catheters
Drain collections
Surgical intervention if indicated (e.g., acalculous cholecystitis, necrotizing fasciitis)

Step 5: Antimicrobial Selection (if indicated)

Empiric regimen based on likely source, local resistance patterns
Cover MRSA if risk factors present (prior MRSA, high local prevalence)
Consider antifungals if prolonged ICU stay, broad-spectrum antimicrobials, TPN, immunosuppression

Step 6: Reassessment

Daily reassessment of antimicrobial necessity
Culture results available? De-escalate.
Clinical improvement? Consider duration/discontinuation.
No improvement? Consider non-infectious causes, resistant organisms, inadequate source control.

Special Populations

Immunocompromised Patients

Lower threshold for empiric antimicrobials
Consider atypical pathogens (fungi, viruses, parasites, opportunistic bacteria)
Non-infectious causes common: GVHD, drug reactions, underlying malignancy
Procalcitonin less reliable

Post-Cardiac Surgery

Fever common in first 48-72 hours (inflammatory response to cardiopulmonary bypass)
Infection uncommon in first 48 hours
After 72 hours: Consider sternal wound infection, pneumonia, endocarditis, post-pericardiotomy syndrome³⁵

Burns

Hypermetabolic state with elevated baseline temperature
Frequent dressing changes cause temperature spikes
High risk for infection (impaired barrier, immunosuppression)
Balance between infection risk and inflammatory response challenging³⁶

Clinical Pearls and Oysters Summary

Pearls:

Rigors strongly suggest true infectious fever
Normal procalcitonin + elevated CRP suggests non-infectious inflammation
Post-operative fever in days 0-2 is usually non-infectious
Central fever: high (>39.5°C), persistent, antipyretic-resistant, uncoupled heart rate
Drug fever typically resolves within 48-72 hours of medication discontinuation
DVT rarely causes fever >38.5°C
Procalcitonin algorithms reduce antimicrobial exposure without increasing mortality
Brief observation (4-6 hours) in stable patients allows diagnostic workup and may prevent unnecessary antimicrobials
Fever itself is not harmful in most ICU patients; routine antipyresis shows no mortality benefit
Daily antimicrobial reassessment is critical for stewardship

Oysters:

Hyperthermia (NMS, serotonin syndrome, malignant hyperthermia) ≠ fever; does NOT respond to antipyretics
Central fever is diagnosis of exclusion; thoroughly rule out infection first
Positive central line cultures may represent colonization, not infection
Fever in pancreatitis does NOT automatically mean infected necrosis
Atelectasis causing post-operative fever is a myth; surgical inflammatory response is responsible
Eosinophilia only present in 20-25% of drug fever cases; absence doesn't exclude it
Early antimicrobials in sterile pancreatitis do NOT prevent infected necrosis and may harm
"One-hour sepsis bundle" should not override clinical judgment in stable, unclear cases
Procalcitonin unreliable in: renal failure without CRRT, immunocompromised, early sepsis, localized/viral/fungal infections
Septic thrombophlebitis requires weeks of antimicrobials plus anticoagulation, not just catheter removal

Conclusion

Differentiating infectious from non-infectious fever in critically ill patients requires systematic clinical reasoning, judicious use of biomarkers, and thoughtful antimicrobial stewardship. While infection remains the primary concern, recognizing non-infectious causes prevents unnecessary antimicrobial exposure, reduces resistance, and improves patient outcomes. The clinician must balance the urgency of treating sepsis against the harms of reflexive antimicrobial use, employing risk stratification, biomarkers like procalcitonin, and clinical judgment to navigate this complex decision-making process.

The art of critical care medicine lies not in treating every fever, but in discerning which fevers require treatment and which represent the body's appropriate inflammatory response to non-infectious stressors.

References

Laupland KB, Shahpori R, Kirkpatrick AW, et al. Occurrence and outcome of fever in critically ill adults. Crit Care Med. 2008;36(5):1531-1535.
O'Grady NP, Barie PS, Bartlett JG, et al. Guidelines for evaluation of new fever in critically ill adult patients: 2008 update from the American College of Critical Care Medicine and the Infectious Diseases Society of America. Crit Care Med. 2008;36(4):1330-1349.
Circiumaru B, Baldock G, Cohen J. A prospective study of fever in the intensive care unit. Intensive Care Med. 1999;25(7):668-673.
Kumar A, Roberts D, Wood KE, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med. 2006;34(6):1589-1596.
Dinarello CA, Porat R. Fever and hyperthermia. In: Jameson JL, Fauci AS, Kasper DL, et al., eds. Harrison's Principles of Internal Medicine. 20th ed. McGraw Hill; 2018.
Marik PE. Fever in the ICU. Chest. 2000;117(3):855-869.
Barie PS, Hydo LJ, Eachempati SR. Causes and consequences of fever complicating critical surgical illness. Surg Infect (Larchmt). 2004;5(2):145-159.
Hocker SE, Tian L, Li G, et al. Indicators of central fever in the neurologic intensive care unit. JAMA Neurol. 2013;70(12):1499-1504.
Tamayo E, Álvarez FJ, Martínez-Rafa B, et al. Fever in the cardiac postoperative period: a retrospective observational study. J Cardiothorac Vasc Anesth. 2012;26(5):791-796.
Garibaldi RA, Brodine S, Matsumiya S, Coleman M. Evidence for the non-infectious etiology of early postoperative fever. Infect Control. 1985;6(7):273-277.
Mavros MN, Velmahos GC, Falagas ME. Atelectasis as a cause of postoperative fever: where is the clinical evidence? Chest. 2011;140(2):418-424.
Thompson HJ, Tkacs NC, Saatman KE, et al. Hyperthermia following traumatic brain injury: a critical evaluation. Neurobiol Dis. 2003;12(3):163-173.
Wacker C, Prkno A, Brunkhorst FM, Schlattmann P. Procalcitonin as a diagnostic marker for sepsis: a systematic review and meta-analysis. Lancet Infect Dis. 2013;13(5):426-435.
Schuetz P, Wirz Y, Sager R, et al. Effect of procalcitonin-guided antibiotic treatment on mortality in acute respiratory infections: a patient level meta-analysis. Lancet Infect Dis. 2018;18(1):95-107.
Póvoa P, Coelho L, Almeida E, et al. C-reactive protein as a marker of infection in critically ill patients. Clin Microbiol Infect. 2005;11(2):101-108.
Dempsey C, Skoglund E, Muldrew KL, Garey KW. Economic health care costs of blood culture contamination: a systematic review. Am J Infect Control. 2019;47(8):963-967.
Blot F, Nitenberg G, Chachaty E, et al. Diagnosis of catheter-related bacteraemia: a prospective comparison of the time to positivity of hub-blood versus peripheral-blood cultures. Lancet. 1999;354(9184):1071-1077.
Bleeker-Rovers CP, Vos FJ, Mudde AH, et al. A prospective multi-centre study of the value of FDG-PET as part of a structured diagnostic protocol in patients with fever of unknown origin. Eur J Nucl Med Mol Imaging. 2007;34(5):694-703.
Haidar R, Musallam KM, Taher AT. Fever of unknown origin in the intensive care unit: an observational study of 87 episodes. Am J Med Sci. 2014;347(4):282-286.
Mackowiak PA, LeMaistre CF. Drug fever: a critical appraisal of conventional concepts. Ann Intern Med. 1987;106(5):728-733.
Patel RA, Gallagher JC. Drug fever. Pharmacotherapy. 2010;30(1):57-69.
Mermel LA, Allon M, Bouza E, et al. Clinical practice guidelines for the diagnosis and management of intravascular catheter-related infection: 2009 update by the Infectious Diseases Society of America. Clin Infect Dis. 2009;49(1):1-45.
Delaney M, Wendel S, Bercovitz RS, et al. Transfusion reactions: prevention, diagnosis, and treatment. Lancet. 2016;388(10061):2825-2836.
Banks PA, Bollen TL, Dervenis C, et al. Classification of acute pancreatitis—2012: revision of the Atlanta classification and definitions by international consensus. Gut. 2013;62(1):102-111.
Tenner S, Baillie J, DeWitt J, Vege SS. American College of Gastroenterology guideline: management of acute pancreatitis. Am J Gastroenterol. 2013;108(9):1400-1415.
Marik PE, Pastores SM, Annane D, et al. Recommendations for the diagnosis and management of corticosteroid insufficiency in critically ill adult patients: consensus statements from an international task force. Crit Care Med. 2008;36(6):1937-1949.
Kilpatrick MM, Lowry DW, Firlik AD, et al. Hyperthermia in the neurosurgical intensive care unit. Neurosurgery. 2000;47(4):850-856.
Knockaert DC, Vanneste LJ, Bobbaers HJ. Fever of unknown origin in elderly patients. J Am Geriatr Soc. 1993;41(11):1187-1192.
Singer M, Deutschman CS, Seymour CW, et al. The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3). JAMA. 2016;315(8):801-810.
Kalil AC, Gilbert DN, Winslow DL, et al. Infectious Diseases Society of America (IDSA) POSITION STATEMENT: Why IDSA Did Not Endorse the Surviving Sepsis Campaign Guidelines. Clin Infect Dis. 2018;66(10):1631-1635.
Chastre J, Wolff M, Fagon JY, et al. Comparison of 8 vs 15 days of antibiotic therapy for ventilator-associated pneumonia in adults: a randomized trial. JAMA. 2003;290(19):2588-2598.
Niven DJ, Stelfox HT, Léger C, et al. Assessment of the safety and feasibility of administering antipyretic therapy in critically ill adults: a systematic review. J Crit Care. 2013;28(5):886-894.
Young P, Saxena M, Bellomo R, et al. Acetaminophen for fever in critically ill patients with suspected infection. N Engl J Med. 2015;373(23):2215-2224.
Badjatia N, Strongilis E, Gordon E, et al. Metabolic impact of shivering during therapeutic temperature modulation: the Bedside Shivering Assessment Scale. Stroke. 2008;39(12):3242-3247.
Lahtinen P, Koskenvuo J, Friberg O, et al. Post-pericardiotomy syndrome and post-operative mediastinitis: similar inflammatory processes? Interact Cardiovasc Thorac Surg. 2011;13(1):3-7.
Greenhalgh DG, Saffle JR, Holmes JH 4th, et al. American Burn Association consensus conference to define sepsis

and infection in burns. J Burn Care Res. 2007;28(6):776-790.

Schortgen F, Clabault K, Katsahian S, et al. Fever control using external cooling in septic shock: a randomized controlled trial. Am J Respir Crit Care Med. 2012;185(10):1088-1095.
Drewry AM, Ablordeppey EA, Murray ET, et al. Antipyretic therapy in critically ill septic patients: a systematic review and meta-analysis. Crit Care Med. 2017;45(5):806-813.
Lee BH, Inui D, Suh GY, et al. Association of body temperature and antipyretic treatments with mortality of critically ill patients with and without sepsis: multi-centered prospective observational study. Crit Care. 2012;16(1):R33.
Niven DJ, Gaudet JE, Laupland KB, et al. Accuracy of peripheral thermometers for estimating temperature: a systematic review and meta-analysis. Ann Intern Med. 2015;163(10):768-777.

Supplementary Clinical Scenarios

To consolidate learning, consider these common ICU scenarios:

Case 1: The Post-Operative Surgical Patient

Presentation: 62-year-old male, post-operative day 1 from elective colorectal surgery. Temperature 38.7°C, heart rate 95 bpm, blood pressure 128/76 mmHg, respiratory rate 18/min. No respiratory distress. Surgical site clean and dry. Patient mobilizing with physiotherapy.

Laboratory values: WBC 14,000/μL (82% neutrophils, no bands), CRP 85 mg/L, procalcitonin 0.3 ng/mL.

Approach:

Diagnosis: Post-operative inflammatory response (surgical trauma)
Action: Observe, encourage mobilization, adequate analgesia, respiratory exercises
Avoid: Empiric antimicrobials (procalcitonin <0.5, POD 1, no localizing signs)
Reassess: If fever persists beyond 48-72 hours or patient deteriorates

Rationale: Early post-operative fever (POD 0-2) is typically non-infectious. The mildly elevated procalcitonin is expected from surgical stress. Immediate antimicrobial therapy would be inappropriate and contributes to resistance.

Case 2: The Ventilated Trauma Patient

Presentation: 45-year-old female, day 7 in ICU following polytrauma (femur fracture, rib fractures, pulmonary contusion). Mechanically ventilated. New fever 39.1°C. Increased purulent endotracheal secretions. Hemodynamically stable.

Laboratory values: WBC 16,500/μL (88% neutrophils, 8% bands), procalcitonin 2.8 ng/mL, chest X-ray shows new left lower lobe infiltrate.

Approach:

Diagnosis: Ventilator-associated pneumonia (VAP) - highly probable
Action:
- Obtain endotracheal aspirate/BAL for culture and Gram stain
- Initiate empiric antimicrobials (anti-pseudomonal β-lactam + vancomycin or linezolid based on local MRSA risk)
- Ensure source control (secretion clearance, ventilator bundle compliance)
De-escalation plan: Narrow based on culture results at 48-72 hours; target 7-day duration for uncomplicated VAP

Rationale: Strong clinical and laboratory evidence of bacterial pneumonia. Elevated procalcitonin (>2.0) with new infiltrate and purulent secretions in mechanically ventilated patient mandates empiric therapy. However, plan for de-escalation from initiation.

Case 3: The Neurocritical Care Dilemma

Presentation: 38-year-old male, day 4 post-severe traumatic brain injury with diffuse axonal injury and intraventricular hemorrhage. External ventricular drain in place. Temperature persistently 39.5-40.2°C despite acetaminophen. No rigors. Heart rate 88 bpm (on β-blocker). Minimal response to antipyretics.

Laboratory values: WBC 11,200/μL, procalcitonin 0.15 ng/mL, CRP 45 mg/L. Blood cultures, urine culture, CSF analysis (cell count, Gram stain, culture) all negative. CT brain shows expected findings, no new lesions or collections. Chest X-ray clear.

Approach:

Diagnosis: Central (neurogenic) fever - diagnosis of exclusion after thorough workup
Action:
- Continue infection surveillance (do not assume all fever is central)
- External cooling measures: Cooling blankets, ice packs to groin/axillae/neck
- Consider intravascular cooling catheter if sustained elevation
- Manage shivering: Buspirone 30 mg TID via feeding tube, magnesium 4-6 g IV as needed
- Target temperature <38.5°C to reduce intracranial pressure and metabolic demand
Avoid: Prolonged empiric antimicrobials without evidence of infection

Rationale: Classic presentation of central fever - high, persistent, antipyretic-resistant fever following brain injury with normal procalcitonin and negative cultures. However, ongoing vigilance for infection is essential as co-infection is common.

Case 4: The Drug Fever Conundrum

Presentation: 55-year-old female with urosepsis, day 10 of piperacillin-tazobactam. Initial clinical improvement, but now has recurrent fever 38.9-39.4°C for past 3 days. Hemodynamically stable. No new symptoms.

Laboratory values: WBC 9,800/μL (6% eosinophils), procalcitonin decreased from 4.2 to 0.8 ng/mL, then stable. Repeat blood and urine cultures negative. CT abdomen/pelvis shows resolving pyelonephritis, no abscess.

Approach:

Diagnosis: Drug fever (piperacillin-tazobactam most likely)
Action:
- Discontinue piperacillin-tazobactam (adequate treatment duration for urosepsis)
- Observe temperature curve over next 48-72 hours
- If continued antimicrobials needed for other indication, switch to alternative agent
Expect: Defervescence within 48-72 hours

Rationale: Clinical improvement with decreasing procalcitonin but persistent fever suggests non-infectious cause. Eosinophilia (though mild) and timing (7-10 days after drug initiation) support drug fever. β-lactams are the most common causative agents.

Case 5: The Occult Intra-Abdominal Source

Presentation: 67-year-old male, day 14 in ICU following complicated appendectomy for perforated appendicitis. Initially improved on antimicrobials, completed 7-day course. Now new fever 38.8°C on day 3 post-antimicrobial completion. Mild abdominal distension, no peritoneal signs.

Laboratory values: WBC 13,400/μL, procalcitonin 1.2 ng/mL (was 0.3 ng/mL two days ago), CRP 120 mg/L.

Approach:

Diagnosis: Concern for intra-abdominal abscess or other complication
Action:
- CT abdomen/pelvis with IV contrast
- If abscess identified: percutaneous or surgical drainage + antimicrobials
- If no abscess but phlegmon/inflammation: consider antimicrobials alone vs. continued observation
- Ensure no other sources (C. difficile testing, review all lines/wounds)

Rationale: Rising procalcitonin after antimicrobial completion suggests recurrent/persistent infection. Intra-abdominal abscess is common after perforated viscus. Source control is paramount - antimicrobials alone insufficient for undrained collections.

Advanced Diagnostic Pearls

The "Sepsis Mimics" - Conditions That Look Like Sepsis But Aren't Infection

Adrenal Crisis
- Hypotension, fever, altered mental status
- Risk factors: chronic steroids, pituitary disease, sepsis itself (relative adrenal insufficiency)
- Clue: Hyponatremia, hyperkalemia, hypoglycemia, eosinophilia
- Treatment: Hydrocortisone 50-100 mg IV q6-8h
Thyroid Storm
- High fever (>40°C), tachycardia out of proportion to fever, agitation, diaphoresis
- Risk factors: Known hyperthyroidism, recent iodine load, surgery, infection trigger
- Clue: Inappropriately fast heart rate, tremor, goiter, lid lag
- Treatment: Propylthiouracil, β-blockers, hydrocortisone, supportive care
Acute Porphyria
- Fever, abdominal pain, neuropsychiatric symptoms, hyponatremia
- Triggers: Medications (many antimicrobials, anticonvulsants), fasting, stress
- Clue: Dark/red urine, peripheral neuropathy, psychiatric symptoms
- Diagnosis: Urine porphobilinogen
- Treatment: Stop offending agents, IV hemin, glucose loading
Hemophagocytic Lymphohistiocytosis (HLH)
- Persistent fever, cytopenias, hepatosplenomegaly, hyperferritinemia
- Can be triggered by infection but becomes self-perpetuating inflammatory syndrome
- Clue: Ferritin >3000 ng/mL, triglycerides >265 mg/dL, low fibrinogen despite acute illness
- Diagnosis: HLH-2004 criteria, bone marrow hemophagocytosis
- Treatment: Immunosuppression (dexamethasone, etoposide), treat trigger
Systemic Mastocytosis
- Flushing, hypotension, bronchospasm, fever
- Clue: Urticaria pigmentosa, elevated tryptase
- Treatment: Antihistamines, mast cell stabilizers, avoid triggers

Temperature Measurement: Technical Considerations

Site accuracy hierarchy:

Pulmonary artery catheter (gold standard, rarely used solely for temperature)
Bladder catheter thermistor (accurate, requires specialized catheter)
Esophageal probe (accurate, requires intubation)
Rectal (accurate core temperature, 0.4°C higher than oral)
Oral (reasonably accurate if proper technique)
Axillary (least accurate, 0.5-1.0°C lower than core)
Tympanic (variable accuracy, operator-dependent)
Temporal artery (reasonable screening, may underestimate high fevers)

Hack: In the ICU, bladder temperature (via specialized Foley catheter) is highly accurate and convenient for continuous monitoring, especially when targeted temperature management is employed.⁴⁰

Pearl: Peripheral thermometers (oral, axillary, temporal) may underestimate core temperature by 0.5-1.0°C in patients with poor perfusion or vasoconstriction. Consider core temperature measurement in shock states.

The "Fever Workup Checklist" - A Practical Tool

When evaluating ICU fever, systematically assess:

Clinical Assessment

☐ Hemodynamic status (MAP, vasopressor requirements, lactate)
☐ Respiratory status (SpO₂, work of breathing, ventilator parameters)
☐ Mental status changes
☐ Rigors present?
☐ New localizing symptoms/signs

Device Assessment

☐ Central lines: Duration, site inspection, need for removal
☐ Urinary catheter: Duration, urine characteristics
☐ Endotracheal tube: Secretion quality/quantity
☐ Drains: Output characteristics
☐ Wounds: Inspection for erythema, drainage, dehiscence

Medication Review

☐ New medications in past 7-14 days
☐ High-risk drugs: β-lactams, anticonvulsants, allopurinol, H2-blockers
☐ Duration of current antimicrobials

Laboratory Risk Stratification

☐ Procalcitonin (<0.5 low risk; 0.5-2.0 intermediate; >2.0 high risk)
☐ WBC trend (leukocytosis, leukopenia, bandemia, eosinophilia)
☐ CRP (elevated in both infectious and non-infectious)
☐ Lactate (tissue hypoperfusion)
☐ New organ dysfunction (creatinine, bilirubin, platelets, coagulation)

Microbiological Cultures (Before Antimicrobials)

☐ Blood cultures × 2 sets
☐ Respiratory culture (if intubated or productive cough)
☐ Urine culture (if urinary symptoms or abnormal UA)
☐ Wound/drain fluid cultures (if applicable)
☐ Stool for C. difficile (if diarrhea or recent antimicrobials)

Imaging

☐ Chest X-ray (pneumonia, effusion, device position)
☐ Consider CT if source unclear and patient stable
☐ Consider ultrasound for focused assessment (cardiac, abdomen, DVT)

Non-Infectious Differential

☐ Post-operative day 0-2? (surgical inflammatory response)
☐ Central nervous system injury? (central fever)
☐ Drug fever candidate?
☐ VTE risk/symptoms?
☐ Recent transfusion?
☐ Pancreatitis present?

Future Directions and Emerging Technologies

Novel Biomarkers

Presepsin (sCD14-ST):

Released during bacterial phagocytosis
Faster kinetics than procalcitonin (peaks at 2-3 hours)
May differentiate bacterial from viral infections
Not yet widely available; more studies needed

Neutrophil CD64:

Surface marker upregulated in bacterial infection
Flow cytometry-based
Shows promise but limited by technical requirements

MicroRNA panels:

Emerging research on gene expression profiles to differentiate infection from inflammation
Years away from clinical application

Artificial Intelligence and Machine Learning

Several AI algorithms are in development to predict infection vs. non-infectious fever using:

Continuous vital sign data
Electronic health record data mining
Pattern recognition in clinical trajectories

Early studies show promise but require validation before clinical implementation.

Rapid Molecular Diagnostics

PCR-based multiplex panels:

Blood culture ID panels (identify pathogens in hours vs. days)
Respiratory pathogen panels (viral vs. bacterial differentiation)
Gastrointestinal panels (C. difficile, other enteric pathogens)

Limitation: Detect nucleic acid (presence) not viable organisms (infection); cannot distinguish colonization from infection

Next-generation sequencing (NGS):

Metagenomic sequencing of plasma/CSF
Can identify rare/unexpected pathogens
Currently expensive and slow (24-48 hours)
Role in fever of unknown origin investigations

Teaching Points for Rounds

When teaching trainees about fever in the ICU, emphasize:

"Not all fever is infection, not all infection causes fever"
- 20-30% of septic patients are normothermic or hypothermic
- Up to 50% of ICU fever is non-infectious
"Cultures before antimicrobials, antimicrobials within the hour for sepsis"
- Balance diagnostic yield with urgency
- In true sepsis, don't delay; in uncertain cases, brief observation acceptable
"Procalcitonin guides, clinical judgment decides"
- Biomarkers are adjuncts, not replacements for clinical assessment
- Know the limitations
"Source control trumps antimicrobials"
- Undrained abscess won't improve with antimicrobials alone
- Remove infected devices
"De-escalation should be the expectation, not the exception"
- Plan for de-escalation from the moment you start empiric therapy
- Narrow spectrum at 48-72 hours when cultures available
- Define duration upfront
"Fever itself is rarely harmful"
- Routine antipyresis shows no benefit
- Selective use for specific indications
"Consider non-infectious causes especially when the clinical picture doesn't fit"
- Stable patient with isolated fever and low procalcitonin
- Fever unresponsive to appropriate antimicrobials
- Atypical timeline (very early post-op, delayed after antimicrobial course)
"Every antimicrobial day counts"
- Each day of unnecessary antimicrobials increases resistance risk
- Shorter courses are often adequate
- Daily reassessment is mandatory

Key Takeaway Messages

Fever in the ICU demands systematic evaluation, not reflexive antimicrobial therapy. Risk stratification using clinical assessment, procalcitonin, and diagnostic workup allows tailored management.
Non-infectious causes account for up to 50% of ICU fever, particularly in surgical, neurocritical care, and early post-operative patients. Recognizing these patterns prevents unnecessary antimicrobial exposure.
Procalcitonin is a valuable tool for differentiating bacterial infection from other causes and guiding antimicrobial duration, but must be interpreted in clinical context with awareness of its limitations.
Source control is paramount. No amount of antimicrobial therapy will cure an undrained abscess or overcome an infected device left in situ.
Antimicrobial stewardship begins at initiation, not at de-escalation. Plan for narrowing spectrum and duration from the moment empiric therapy starts. Daily reassessment prevents antimicrobial "inertia."
Fever itself is not the enemy. Routine antipyresis provides no mortality benefit in most ICU patients. Target temperature management for specific indications: patient comfort, excessive metabolic demand, or intracranial hypertension.
When in doubt, observe (if the patient is stable). A brief period of diagnostic evaluation before initiating antimicrobials may prevent weeks of unnecessary therapy. However, never delay antimicrobials in sepsis or septic shock.
Central fever is a diagnosis of exclusion. Thoroughly rule out infection before attributing fever to neurologic injury alone, as co-infection is common in neurocritical care patients.

Conclusion

The febrile critically ill patient presents one of the most common and challenging scenarios in intensive care medicine. Success requires integration of clinical reasoning, diagnostic acumen, microbiological principles, and antimicrobial stewardship. By systematically differentiating infectious from non-infectious causes, judiciously using biomarkers and advanced diagnostics, and thoughtfully applying antimicrobials only when indicated, intensivists can optimize patient outcomes while combating the growing threat of antimicrobial resistance.

The ultimate goal is not to treat every fever, but to identify and appropriately manage those that require intervention while recognizing that fever itself is often an appropriate physiological response that requires no specific therapy. In the words of William Osler, "It is much more important to know what sort of patient has a disease than what sort of disease a patient has" - a principle that remains profoundly relevant in the modern ICU.

This review article synthesizes current evidence and clinical experience to provide a practical framework for managing fever in critical care. Clinicians should adapt these principles to their local epidemiology, resistance patterns, and institutional protocols while maintaining the core tenets of thoughtful clinical reasoning and antimicrobial stewardship.

Friday, October 17, 2025