ICU biomarkers: procalcitonin, lactate, troponin, BNP, and CRP · ICU Topics

Cinematic ICU scene of a monitor displaying trends of lactate, procalcitonin, troponin, BNP and CRP, an arterial blood gas and a venous sample on the trolley, clinical-blue lighting, medical educational, no faces, no text — FigureThe ICU biomarkers — the lactate for the perfusion and the clearance, the procalcitonin for the bacterial infection and the antibiotic stewardship, the troponin for the myocardial injury, the BNP for the heart failure. The trend is more useful than the single value; the biomarker complements, never replaces, the clinical assessment.

Biomarker release pathways in critical illness: tissue hypoxia and lactate, bacterial cytokines and procalcitonin, cardiomyocyte injury and troponin — educational diagram — FigureBiomarkers are kinetic signals — interpret trends and pretest probability, never a single number in isolation.

ICU biomarker-guided actions: lactate clearance in sepsis, PCT-guided antibiotic stop rules, troponin for type 1 versus type 2 MI, BNP for volume and heart failure — management pathway — FigureUse kinetics: clear lactate, stop antibiotics with falling PCT and clinical recovery, avoid reflex angiography for isolated type 2 troponin rise.

Key ICU biomarkers comparison

Biomarker	What it measures	When elevated	Clinical use	Limitations
Procalcitonin	Bacterial infection (calcitonin precursor)	Bacterial sepsis (>2 ng/mL), bacterial pneumonia; NOT viral/non-infective	Antibiotic stewardship (stop when <0.5 or falls ≥80%)	False positive (trauma, surgery, cardiogenic shock), false negative (first 6-12 h, localised)
Lactate	Tissue hypoperfusion / anaerobic metabolism	Shock, sepsis, mesenteric ischaemia, seizures, burns	Guide resuscitation (clearance >10%/h), prognosis	Also raised: β2-agonists, malignancy, mitochondrial toxins, liver failure
Troponin / hs-cTn	Myocardial injury (cardiomyocyte necrosis)	Type 1 MI, type 2 MI, MINS, sepsis, PE, myocarditis, renal failure	Diagnose MI, grade cardiac injury, prognostic	Type 2 in ICU (not atherosclerotic) — do NOT reflexively PCI; renal failure elevates
BNP / NT-proBNP	Cardiac stretch / wall tension (volume overload)	Heart failure, volume overload, PE, sepsis	Diagnose heart failure, guide diuresis, prognostic	Raised: renal failure, PE, sepsis, age, AF; LOWER in obesity (haemodilution/clearance)
CRP	Acute-phase inflammation	Infection, inflammation, tissue injury, malignancy, autoimmune	Trend (rising = worsening), monitor infection response	Nonspecific, slow (24-48 h lag), many causes
D-dimer	Cross-linked fibrin degradation (coagulation + fibrinolysis)	VTE (DVT/PE), DIC, arterial thrombosis, inflammation, malignancy, pregnancy, surgery	Rule OUT VTE/DIC (high NPV); ISTH DIC scoring	Non-specific (low specificity); rises with age — use age-adjusted cut-off (age × 10 µg/L in >50 yr)
suPAR	Soluble urokinase plasminogen activator receptor (immune activation)	Sepsis, septic shock, multi-organ failure, inflammation	Prognostic — predicts mortality independent of severity scores	Nonspecific; influenced by renal function, malignancy; not yet routine bedside
Copeptin (CT-proAVP)	C-terminal fragment of pro-arginine-vasopressin (AVP surrogate)	Shock, sepsis, stress, hypotension, hyponatraemia	Surrogate for hard-to-measure AVP; prognostic in shock; guides vasopressin consideration	Released with stress; non-specific; influenced by osmolality, fluid status

[1]

Practical use of biomarkers in septic shock

On presentation — lactate (baseline — guide resuscitation), procalcitonin (baseline — guide antibiotics), CRP (baseline — trend later)
Resuscitation guided by lactate — target lactate clearance >10%/h (or normalisation within 2-6h). If not clearing → ongoing hypoperfusion (assess volume, cardiac output, source control)
Procalcitonin trend — if PCT falls >80% from baseline or <0.5 → consider stopping antibiotics (PROORATA algorithm). Repeat daily
Troponin — check if cardiac symptoms, arrhythmia, or unexplained hypotension. Elevated → assess for type 1 (ECG, ACS) vs type 2 MI (treat underlying cause)
BNP/NT-proBNP — check if heart failure suspected (dyspnoea, oedema). Low → heart failure unlikely. High → heart failure or other cause (renal, PE, sepsis)
Serial CRP — trend every 24-48h. Falling → responding. Rising or persistently high → treatment failure, complication, ongoing infection

[1]

Clinical pearls

High-yield ICU biomarker points for CICM/FFICM exam

Procalcitonin guides ANTIBIOTIC DURATION, not initiation. PCT is a bacterial infection marker — rises in bacterial sepsis, stays low in viral/non-infective. PRORATA trial: PCT-guided algorithm (stop antibiotics when PCT <0.5 or falls 80%) reduced antibiotic days by 2-3 days WITHOUT increasing adverse outcomes. BUT: still start empiric antibiotics for suspected sepsis (don't wait for PCT).[1] }
Lactate clearance >10%/h is the target in septic shock. Jansen (Am J Resp Crit Care 2010): lactate clearance-guided resuscitation reduced mortality in lactate ≥3.0. Measure every 2h. If clearance <10%: reassess (more fluids? inotropes? source control?). Lactate normalisation (≤2) is the goal. CLEARED lactate = adequate resuscitation.[2] }
Troponin in ICU is often TYPE 2 MI (supply-demand), not type 1 (atherosclerotic). Type 2: tachycardia, hypotension, hypoxia, sepsis → supply-demand mismatch → myocardial injury → troponin rise. Treatment: address underlying cause (oxygenate, blood pressure, rate control). Type 1: plaque rupture → thrombus → need antiplatelet, anticoagulant, PCI. CLINICAL: don't automatically PCI every elevated troponin in ICU — assess type (ECG changes? chest pain? ACS features?).[4] }
BNP/NT-proBNP: heart failure diagnosis and prognosis. Low (<300 NT-proBNP) → heart failure UNLIKELY (negative predictive value 90%). High → heart failure likely (but false positives: renal failure, PE, sepsis, age, AF). Prognosis: higher BNP = worse outcome in ICU. Also guides diuresis (reducing BNP = decongestion).[3] }
CRP is NONSPECIFIC — use trends, not single values. CRP rises with ANY inflammation (infection, surgery, trauma, malignancy, autoimmune). Rises 24-48h after insult (slow). Trend (every 24-48h): falling = responding to treatment; rising/persistent = treatment failure. Procalcitonin is MORE SPECIFIC for bacterial infection than CRP.[5] }
Lactate sources — not all elevated lactate is hypoperfusion. TYPE A (hypoperfusion): shock, sepsis, mesenteric ischaemia, CO poisoning. TYPE B (non-hypoperfusion): beta-agonists (salbutamol), malignancy (lymphoma), mitochondrial toxins (metformin, cyanide), thiamine deficiency, seizures (muscle), liver failure (impaired clearance). If lactate high but well perfused → consider type B.[6] }
Procalcitonin kinetics — rising or persistently high suggests ongoing infection. PCT normally falls 50% every 24h with effective antibiotics. If NOT falling: (1) Inadequate antibiotic (resistant organism, wrong drug). (2) Inadequate source control (undrained abscess). (3) New infection. (4) Non-bacterial cause (persistent PCT without bacterial infection — rare). Rising PCT = reassess.[1] }
Lactate >2 is sepsis criterion (qSOFA + lactate). Sepsis-3 definition: suspected infection + SOFA ≥2 (includes lactate >2 as organ dysfunction). Lactate >2 with infection = sepsis (even if normotensive). Lactate >2 + vasopressor requirement = septic shock. Measure lactate in ALL suspected sepsis.[2] }
NT-proBNP vs BNP — different half-lives and thresholds. BNP: half-life 20 min (rapid changes). NT-proBNP: half-life 120 min (more stable, higher values). Both reflect heart failure. NT-proBNP preferred in ICU (more stable, wider dynamic range). Thresholds vary by age, renal function. LOW NT-proBNP excludes heart failure well; HIGH is less specific.[3] }
High-sensitivity troponin (hs-cTn) — detects smaller injury. hs-cTn assays detect myocardial injury at MUCH lower levels (10x more sensitive than conventional). More ICU patients have 'elevated' troponin (any minor injury detected). Interpret carefully — many ICU patients have mildly elevated hs-cTn (not all are MI). Trend matters (rising = ongoing injury).[4] }
Procalcitonin false positives and negatives. FALSE POSITIVE (high PCT without bacterial infection): major surgery, severe trauma, cardiogenic shock, heat stroke, prolonged cardiac arrest, medullary thyroid cancer (produces calcitonin). FALSE NEGATIVE (low PCT with bacterial infection): early infection (PCT rises 6-12h), localized infection (low bacterial load), after antibiotics started. Interpret in clinical context.[1] }
Lactate/pyruvate ratio distinguishes ischaemic from non-ischaemic. Ischaemic (Type A): lactate/pyruvate ratio HIGH (>10-15) — anaerobic metabolism (pyruvate shunted to lactate). Non-ischaemic (Type B): lactate/pyruvate ratio NORMAL — mitochondrial dysfunction (impaired pyruvate oxidation). Not routinely measured, but useful in research/complex cases.[6] }
CRP vs procalcitonin for infection diagnosis. PCT is MORE SPECIFIC for BACTERIAL infection (low in viral). CRP is elevated in ANY inflammation (less specific). Both rise in bacterial infection. PCT rises FASTER (6-12h vs 24-48h for CRP). PCT falls faster with effective treatment. For antibiotic stewardship: PCT preferred (more dynamic, specific).[5] }
Combining biomarkers improves prediction. No single biomarker is perfect. COMBINING: lactate (perfusion) + PCT (bacterial infection) + CRP (inflammation) + clinical assessment → better diagnosis and prognosis than any alone. Example: septic shock with high lactate + high PCT + rising CRP → severe bacterial sepsis, aggressive treatment needed. Low lactate + low PCT + falling CRP → resolving, consider de-escalation.[1] }

Red flags

Prognosis

PRORATA trial (Bouadma 2010, Lancet) — procalcitonin-guided antibiotic therapy

RCT: 621 ICU patients with suspected bacterial infection. Procalcitonin-guided (stop antibiotics when PCT <0.5 or fell ≥80%) vs standard (physician decision).

Antibiotic days (days 1-28): PCT 9.3 vs standard 14.9 (p<0.0001) — REDUCED by 5.6 days
Mortality (day 28): PCT 20.3% vs standard 20.4% (NO difference — safe)
Relapse of infection: similar
CONCLUSION: Procalcitonin-guided antibiotic therapy REDUCES antibiotic exposure (5.6 days) without increasing adverse outcomes. Now standard in many ICU antibiotic stewardship programmes. [1]

Jansen (lactate clearance, 2010): lactate-guided resuscitation reduced mortality in septic shock with lactate ≥3.0 (20% vs 44%, p=0.006). BNP prognosis: elevated BNP in ICU → 2-3x higher mortality.

[1]

Lactate — deep dive (perfusion, clearance, prognostics)

Lactate Type A (hypoperfusion / ischaemic) vs Type B (non-hypoperfusion / impaired clearance)

Feature	Type A — hypoperfusion / tissue ischaemia	Type B — non-hypoperfusion (impaired metabolism or clearance)
Mechanism	Anaerobic glycolysis from oxygen debt; pyruvate shunted to lactate	Aerobic glycolysis (β2-agonists), impaired oxidation (mitochondrial toxins), reduced hepatic clearance
Lactate/pyruvate ratio	High (>10-15)	Normal
Examples	Septic / haemorrhagic / cardiogenic shock, mesenteric ischaemia, severe burns, limb ischaemia, CO poisoning, seizures (delayed)	Salbutamol / ritodrine / adrenaline infusions, metformin (MALA), cyanide, thiamine deficiency (beriberi), malignancy (lymphoma, leukaemia — Warburg), liver failure, ethanol, propofol infusion, linezolid
Clinical sign	Cool, mottled, prolonged cap refill, oliguria, hypotension	Warm, well-perfused, normal ScvO2, may be haemodynamically stable
Management	Resuscitate (fluids, blood, inotropes), source control, restore oxygen delivery	Remove trigger (stop β2-agonist / metformin), haemodialysis (metformin, linezolid), thiamine (alcoholic), treat malignancy
Prognosis	Mortality ∝ lactate level and clearance half-life	Driven by underlying cause

[1]

Lactate clearance protocol in septic shock (Jansen 2010 algorithm)

Draw baseline lactate at recognition of septic shock / suspected sepsis with lactate ≥3.0 mmol/L
Resuscitate with crystalloid 30 mL/kg bolus (Surviving Sepsis Campaign) + early broad-spectrum antibiotics within 1 h + source control
Recheck lactate at 2 h — calculate clearance: (baseline − current) / baseline × 100
If clearance ≥10%/h → on target; continue therapy, recheck at 4-6 h
If clearance <10%/h → reassess: echocardiography (cardiac output / fluid responsiveness), MAP target ≥65 mmHg, consider vasopressor (noradrenaline), inotrope (dobutamine) if low CO, source control (drain abscess, debride tissue), consider ongoing occult bleeding
If lactate rising despite resuscitation → red flag: mesenteric ischaemia, mitochondrial dysfunction (severe sepsis), β2-agonist, metformin toxicity, limb/bowel ischaemia. Escalate: surgical review, CT angiography
Target — lactate normalisation (≤2 mmol/L) within 2-6 h; persistent hyperlactataemia at 24 h = poor prognosis

[1]

Troponin in ICU — type 1 MI, type 2 MI, and MINS

ICU biomarker families: perfusion (lactate), infection stewardship (procalcitonin, CRP), cardiac injury (troponin, BNP), coagulation (D-dimer) — educational classification panel — FigureMatch the biomarker to the question — lactate for perfusion, PCT for stewardship, troponin for myocardial injury, BNP for wall stress.

Type 1 MI vs Type 2 MI vs MINS — distinguishing myocardial injury in ICU

Feature	Type 1 MI	Type 2 MI	MINS (myocardial injury after non-cardiac surgery)
Mechanism	Atherosclerotic plaque rupture/erosion → in-situ thrombus	Supply-demand mismatch (no plaque rupture)	Perioperative supply-demand mismatch ± non-ischaemic injury
Troponin	Rise AND fall, with ≥1 value >99th centile	Rise AND fall, with ≥1 value >99th centile	Peak troponin ≥99th centile within 30 d post-op (no ischaemic symptom required)
Symptoms	Ischaemic chest pain/symptoms	Often absent or atypical in ICU	Usually SILENT (85% without ischaemic features) — detected by routine troponin screening
ECG	ST-elevation, ST-depression, T-wave inversion, new LBBB	Often non-specific, sinus tachycardia, or changes of underlying cause	Non-specific; new ST changes uncommon
Imaging	Regional wall motion abnormality, plaque on angiography	May be normal or fixed defect	Variable
Management	DAPT, anticoagulation, urgent PCI (STEMI) / early invasive strategy (NSTE-ACS)	Treat UNDERLYING cause (oxygen, BP, rate, sepsis, anaemia) — routine PCI not indicated	Risk stratify, address modifiable factors, start aspirin + statin, cardiology follow-up; troponin trend alone
Prognosis	Driven by infarct size and treatment	Mortality ~2-3× non-injured ICU patients	30-day mortality ~8-10%; 1-year mortality ~15-20%

[1]

MINS — myocardial injury after non-cardiac surgery (AHA 2021 Scientific Statement)

MINS is defined as peak troponin ≥99th centile upper reference limit (URL) within 30 days after non-cardiac surgery, in the absence of a non-ischaemic cause (sepsis, PE, cardioversion). Ischaemic symptoms (chest pain, dyspnoea) are NOT required — ~85% of MINS cases are clinically silent and are detected by routine postoperative troponin screening. MINS occurs in ~8% of patients undergoing non-cardiac surgery (rising to ~20-30% in vascular and high-risk surgery). It is independently associated with 30-day mortality ~9%, and increased risk of cardiac arrest, cardiogenic shock, and stroke. It is NOT automatically ACS — the pathophysiology is largely supply-demand (type 2-like) and routine PCI does not improve outcomes. Management per AHA 2021: intensive surveillance, initiation of aspirin and high-intensity statin, glycaemic control, telemetric monitoring, cardiology follow-up; reserve invasive angiography for high-risk troponin trajectories (e.g., rise-and-fall with ischaemic features).[9][10][11]

Troponin elevation in ICU — how to distinguish type 1 MI from type 2 MI / MINS

Confirm rise/fall pattern with serial hs-cTn (0, 1, 3 h). Isolated single rise ≠ acute MI — consider chronic injury (renal failure, structural heart disease)
Assess ischaemic symptoms — chest pain, dyspnoea, diaphoresis, radiation (absent in many type 2 / MINS)
ECG — look for ST-elevation (STE), ST-depression, T-wave inversion, new LBBB, posterior MI. Dynamic ST changes favour type 1. Diffuse / non-specific changes favour type 2
Bedside echo — new regional wall motion abnormality (especially in a coronary territory) favours type 1; global hypokinesis / EF reduction favours type 2
Identify supply-demand drivers of type 2 — tachyarrhythmia (AF with RVR), hypoxia, hypotension, severe anaemia, sepsis, severe hypertension
Confirm non-cardiac triggers of troponin rise — PE, myocarditis, sepsis, renal failure, cardioversion, seizures (rule these out before calling it MI)
Decision — STEMI / NSTE-ACS (type 1) → DAPT + anticoag + cardiology + early invasive strategy. Type 2 → treat underlying cause, monitor, antiplatelet/ statin if 10-yr ASCVD risk warrants. MINS → aspirin + statin + surveillance

[1]

BNP / NT-proBNP — heart failure diagnosis, volume status, prognosis

BNP vs NT-proBNP — what's the difference?

Property	BNP	NT-proBNP
Origin	Cleaved from pro-BNP by corin → active BNP + inert NT-proBNP	Inert N-terminal fragment (released in 1:1 stoichiometry)
Half-life	~20 min	~120 min
Stability	Less stable (requires cold processing)	More stable (room temperature OK for hours)
Removed by	Neutral endopeptidases + clearance receptors	Renal clearance predominantly
Affected by neprilysin inhibitors (sacubitril)	↑ measured BNP (false high)	No significant effect → preferred in patients on ARNI
Bedside assay	Available	Most labs batch; not always point-of-care
Heart-failure rule-out (dyspnoea)	BNP <100 pg/mL	NT-proBNP <300 pg/mL (rule-out across all ages)

[1]

D-dimer — fibrin turnover in VTE, DIC, arterial thrombosis

D-dimer elevation — causes and significance

Cause	Typical magnitude	Use
Acute VTE (DVT, PE)	Moderately-high	Rule OUT (negative D-dimer + low pre-test probability → VTE excluded, no imaging needed). High NPV (≥97%)
DIC (overt)	Very high + falling fibrinogen + rising PT + low platelets	ISTH overt DIC score — D-dimer contributes 0-3 points
Arterial thrombosis (MI, stroke, mesenteric)	Mild-moderate	Not diagnostic — supports prothrombotic state
Aortic dissection / aneurysm	Mild-moderate	Aortic Dissection Detection Risk Score uses D-dimer >500 µg/L
Pregnancy (3rd trimester)	Physiologically elevated	Use pregnancy-adapted YEARS algorithm / PE rule-out criteria
Malignancy, inflammation, sepsis, post-op, trauma	Variable, often high	Non-specific; cannot rule out VTE reliably
Ageing	Rises ~10 µg/L per year	Use age-adjusted cut-off: age × 10 µg/L FEU in >50 yr

[1]

D-dimer interpretation in suspected PE in ICU

Stratify pre-test probability — Wells score (low/moderate/high) or PERC if low-risk and young. Imaging is the rule in unstable patients
Low/intermediate pre-test probability → D-dimer useful. Normal D-dimer (with sensitive assay, incl. age-adjusted cut-off in >50 yr) excludes PE — stop workup. The 2017 van Es IPD meta-analysis confirmed Wells + age-adjusted D-dimer gives a 3-month VTE failure rate <1.5% in PE-rule-out patients, including older patients.[18]
High pre-test probability — do NOT rely on D-dimer (specificity too low). Go straight to CT pulmonary angiography
Haemodynamically unstable — bedside echo (RV strain), in-lab pulmonary angiography if unstable for CT. D-dimer is academic
Pregnancy — use YEARS algorithm: D-dimer + clinical signs of DVT + haemoptysis + most-symptomatic leg is PE → CT; if not, use D-dimer cut-off (rule-out at <1000 if no PE signs; <500 if signs present)
Critically ill patient — D-dimer almost always elevated (sepsis, surgery, trauma, lines) → low rule-out utility; rely on imaging. The value in ICU is in trajectory (rising in DIC, falling in resolving thrombosis) and ISTH DIC scoring

suPAR — soluble urokinase-type plasminogen activator receptor

suPAR — prognostic inflammation marker in sepsis and critical illness

suPAR (soluble urokinase-type plasminogen activator receptor) is shed from activated neutrophils, monocytes and endothelium during inflammation — its level reflects immune activation intensity, not a specific disease. The 2012 Backes systematic review (Intensive Care Med) found suPAR is an independent predictor of mortality in sepsis, septic shock and systemic inflammation, often outperforming CRP and procalcitonin for prognosis (but not for diagnosis of bacterial infection). The 2019 Gussen study (J Intensive Care) showed neutrophils are the main source of circulating suPAR, linking neutrophil activation (NETosis, migratory capacity) to outcome — high suPAR indicates exhausted/dysfunctional neutrophils. Practical use: a rising suPAR in septic shock signals a deteriorating immune response and worse prognosis, supporting escalation (reassess source control, consider immunomodulation). Not yet routine bedside in most ICUs — mainly research / academic interest.[13][14]

Copeptin — vasopressin surrogate

Copeptin (CT-proAVP) — measuring the unmeasurable vasopressin

Copeptin is the C-terminal fragment of the arginine-vasopressin (AVP) prohormone, secreted stoichiometrically with AVP from the posterior pituitary. AVP itself is unstable, platelet-bound, and hard to measure; copeptin is stable, easily assayed, and mirrors AVP release — making it the practical surrogate for the vasopressin axis. Copeptin rises with stress (shock, sepsis, hypotension, hyponatraemia, hypoxia). In septic shock, vasopressin deficiency (relative) is a recognised phenomenon, and exogenous vasopressin (0.03 U/min) is a catecholamine-sparing agent — copeptin levels correlate with endogenous AVP reserve and may identify patients most likely to respond. The 2013 Lee study in paediatric sepsis (Intensive Care Med) showed copeptin distinguished septic shock from uncomplicated sepsis and correlated with severity. Prognostically, high copeptin in shock predicts mortality independent of severity scores. Hyponatraemia work-up: copeptin distinguishes AVP-driven (hypervolaemic hyponatraemia — SIADH, heart failure) from non-AVP-driven causes. Not yet routine bedside in most ICUs — used for rule-out of MI in chest-pain pathways (cTn + copeptin) and emerging in shock prognostication.[15][16]

Procalcitonin — antibiotic stewardship algorithm

PRORATA-based procalcitonin algorithm for antibiotic de-escalation

Draw baseline PCT at antibiotic initiation (do NOT delay empiric antibiotics for suspected sepsis)
Day 1-3 (escalation phase) — continue antibiotics. Recheck PCT daily
Stop antibiotics if PCT <0.5 ng/mL OR has fallen ≥80% from peak (and clinical improvement: defervescence, falling WBC, haemodynamic stability, source controlled)
Continue if PCT >0.5 ng/mL AND <80% fall — reassess for: inadequate source control, resistant organism, wrong drug, new infection, non-bacterial cause (PCT can stay high in cardiogenic shock, severe trauma, medullary thyroid cancer)
Special populations — PCT less reliable in neonates, localised infection (endocarditis, abscess), after major surgery/trauma (false positive), early infection <6 h (false negative), and immunosuppressed patients
NEVER withhold or delay antibiotics in septic shock based on a single PCT value — clinical judgement overrides

[1]

Procalcitonin vs CRP — which is the better infection biomarker?

Feature	Procalcitonin	CRP
Biologically specific for bacterial infection	Yes (calcitonin precursor from extra-thyroidal tissues)	No — generic acute-phase reactant
Onset after insult	6-12 h	24-48 h (slower)
Falls with effective therapy	~50% per day	Slow (days)
Useful for antibiotic stewardship	Yes — PRORATA, SAPS, SABS validated algorithms	No — too slow and non-specific
Differentiates bacterial vs viral	Reasonably (low in pure viral)	No
False positives	Major surgery, severe trauma, cardiogenic shock, prolonged cardiac arrest, medulla of thyroid (calcitonin-producing tumour)	Any inflammation, malignancy, autoimmune, tissue injury
Cost / availability	More expensive	Cheap, universal

[1]

Combining biomarkers — multimarker panels improve prediction

How to combine biomarkers in common ICU scenarios

Scenario	Leading biomarkers	What each tells you	Integrated interpretation
Septic shock	Lactate + PCT + CRP + suPAR	Lactate → perfusion; PCT → bacterial likelihood; CRP → inflammation; suPAR → prognosis	High lactate + high PCT + rising suPAR = severe, escalate; clearance of lactate + falling PCT = improving
Acute dyspnoea (HF vs COPD)	BNP/NT-proBNP	Low → not HF; high → likely HF (mind renal/AF/obesity)	Combine with echo + clinical to confirm
Post-op troponin rise	hs-cTn trend	MINS vs type 1 MI	Aspirin + statin; invasive strategy only if ischaemic features
Suspected PE	D-dimer (low pre-test)	Normal D-dimer rules out PE in low-risk	If elevated → CTPA
Bleeding + abnormal coagulation	D-dimer + fibrinogen + PT + platelets	Overt DIC if ISTH ≥5	Treat trigger + factor replacement
Septic shock on rising vasopressors	Copeptin + lactate + ScvO2	Copeptin → AVP reserve; lactate → perfusion; ScvO2 → oxygen utilisation	Low copeptin (vasopressin depleted) + refractory shock → add vasopressin 0.03 U/min

[1]

Clinical pearls — second set (deep dive)

High-yield ICU biomarker pearls — set 2 (CICM / FFICM / EDIC)

Lactate clearance is not the same as lactate normalisation. Clearance ≥10%/h is a rate; normalisation (≤2 mmol/L) is a goal. Both matter — Jansen used clearance (target ≥10%/h), ANDROMEDA-SHOCK questioned whether chasing lactate with more fluids is harmful. Best practice: resuscitate to perfusion endpoints (CRT, mottling, MAP, urine output) AND track lactate — do not chase a number with fluid alone.[2][17]
A normal lactate does not exclude shock. Early septic shock (compensated), cardiogenic shock with high output, distributive shock — lactate may be normal initially. Recheck at 2-4 h. Persistent normal lactate with hypotension → consider adrenal insufficiency, anaphylaxis, vasoplegia from toxins.
Sepsis-3 definitions anchor on lactate. Lactate >2 mmol/L + suspected infection = sepsis (organ dysfunction). Lactate >2 + vasopressor requirement to keep MAP ≥65 = septic shock. Measure lactate in EVERY suspected sepsis — it triages severity, triggers the 1-h bundle, and predicts mortality (lactate ≥4 doubles mortality).[7]
Beta-hydroxybutyrate vs total lactate in DKA. Don't be fooled — DKA ketones (β-hydroxybutyrate) can give a mild lactate rise via altered redox, but the dominant acid is ketoacidosis. A high lactate (>5) in apparent DKA → look for sepsis, mesenteric ischaemia, or metformin.
Lactate >10 mmol/L has mortality >70% in septic shock — this is a marker of catastrophic hypoperfusion or mitochondrial dysfunction. Combined with vasopressor-refractory shock, consider sources (mesenteric ischaemia — lactate ≥5 + abdominal pain, sudden rise), and discuss goals of care early.[6]
MINS is prognostic, not automatically ACS. ~85% of perioperative troponin elevations have NO ischaemic features. The temptation to anticoagulate / PCI every post-op troponin rise is to be resisted. AHA 2021 supports aspirin + statin + surveillance, reserving invasive management for rising-and-falling troponin with ischaemic symptoms / ECG changes. PCI in MINS without ACS has not been shown to improve outcomes.[9][11]
Type 2 MI needs type-2 treatment. Treating tachyarrhythmia, hypoxia, anaemia, sepsis and shock is the therapy — not cath lab referral. Routine DAPT/anticoag in type 2 MI is not indicated unless co-existing type 1. The 2022 Bularga cohort (Circulation) confirmed that type 2 MI patients DO have a higher burden of coronary disease than previously assumed — so risk-stratify for secondary prevention, but the acute event is supply-demand.[8]
Renal failure elevates troponin and BNP. Both are cleared / metabolised by the kidney; eGFR <30 may double hs-cTn and triple NT-proBNP. In a dialysis patient with elevated troponin, the trend (rising-and-falling pattern) and the clinical picture matter more than the absolute value. Chronic troponin elevation in renal failure without dynamic change = chronic injury (NOT acute MI).[4]
BNP falls with decongestion; that is a GOOD sign. A falling BNP/NT-proBNP during diuresis of acute decompensated HF predicts lower readmission and mortality (haemodynamic decongestion). A persistently high or rising BNP despite diuresis = refractory congestion → consider ultrafiltration, inotrope (HFrEF), or mechanical support.
Obesity lowers BNP/NT-proBNP — use obesity cut-offs. BMI ≥30 can halve BNP even in true HF (haemodilution + increased clearance via adipose natriuretic peptide receptors). Consider a lower threshold (e.g., BNP <50 instead of <100) when ruling out HF in obesity — or rely on echo + JVP / POCUS (B-lines).
Sacubitril (in ARNI) increases measured BNP but not NT-proBNP. Because neprilysin inhibition blocks BNP breakdown, measured BNP rises within hours of ARNI initiation — this is a drug artefact, not worsening HF. Use NT-proBNP (not BNP) to monitor patients on ARNI. Stop ARNI for ≥36 h before switching from ACE-i (enalapril) to avoid angioedema.
CRP lags PCT by 24-48 h. In early sepsis, PCT rises faster (6-12 h) and falls faster (half-life ~24 h vs 48 h for CRP). A rising CRP with a falling PCT suggests resolving bacterial infection (CRP lag). A rising CRP with rising PCT suggests ongoing or worsening bacterial infection. Use both trends — PCT for direction of travel, CRP for confirmation.
suPAR is the only biomarker here that predicts mortality independent of disease severity scores. APACHE II, SOFA, SAPS — suPAR adds prognostic information beyond these. In septic shock, a suPAR >12 ng/mL has been associated with mortality >50%. Currently research-only in most ICUs but a high-yield exam topic.[13][14]
Copeptin + troponin doubles the rule-out power for MI in chest pain. A negative high-sensitivity troponin + low copeptin at presentation can rule out non-ST-elevation MI without serial sampling — FDA-approved pathway in some systems. Conceptually: troponin detects cardiomyocyte injury, copeptin detects the endogenous stress response (which rises within minutes of myocardial ischaemia). In ICU, copeptin adds shock-prognosis data.
D-dimer is an exclusion test, not a diagnostic test. The greatest utility of a normal D-dimer is ruling OUT VTE/DIC in low-pre-test-probability patients. An elevated D-dimer is meaningless in isolation (especially in ICU — sepsis, surgery, trauma, malignancy, pregnancy all raise it). The age-adjusted cut-off (age × 10 µg/L FEU in >50 yr) reduces false positives in older adults without missing PE.[18]
In overt DIC, treat the trigger — the lab will follow. DIC is never a primary diagnosis. Abruption, sepsis, APL, amniotic fluid embolism, trauma, snake bite — find and treat the cause while supporting haemostasis (cryoprecipitate to fibrinogen ≥1.5, platelets ≥50 if bleeding, FFP for prolonged PT). Therapeutic anticoagulation (not prophylactic) is reserved for thrombotic-dominant DIC with extensive organ thrombosis.[12]
Procalcitonin in pure viral infection is low — COVID-19, influenza, RSV. A high PCT in viral respiratory illness suggests bacterial co-infection (pneumococcal / staphylococcal superinfection) — the Surviving Sepsis COVID guidelines recommended PCT to identify bacterial superinfection and limit antibiotic exposure.
Beware the 'washout' lactate. After revascularisation of an ischaemic limb or bowel, lactate can briefly rise as accumulated lactate washes into circulation. Distinguish from worsening: trend over 1-2 h, assess perfusion clinically, and look for the source (surgery / clot lysis).
Troponin and renal failure — chronic vs acute. A stable, mildly elevated hs-cTn in a stable dialysis patient is chronic injury (NOT MI). A rising-and-falling pattern with clinical deterioration = acute myocardial injury. Use absolute change (delta) — a delta >20% (or >50% if baseline elevated) supports acute injury in renal failure.[4]
Don't interpret biomarkers in isolation. A high lactate + normal PCT + high CRP → consider non-bacterial causes (mesenteric ischaemia, malignancy, toxin, post-ictal). A normal lactate + high PCT → early bacterial infection before hypoperfusion. Always integrate biomarkers with history, examination, imaging, and clinical trajectory.

Red flags — extended set

Biomarker red flags demanding escalation

Lactate >4 mmol/L + vasopressor requirement → septic shock (Sepsis-3); start 1-h bundle.[7] }
Lactate rising despite adequate resuscitation → mesenteric ischaemia, mitochondrial dysfunction, ongoing bleeding, metformin toxicity → surgical/imaging review.[6] }
Lactate >10 mmol/L → catastrophic perfusion failure; mortality >70% — discuss goals of care.[6] }
Procalcitonin >2 ng/mL → bacterial sepsis likely; start antibiotics.[1] }
Procalcitonin not falling ≥80% by day 5 → reassess source control, resistant organism, wrong drug, new infection.[1] }
Troponin rising with chest pain + dynamic ST changes → type 1 MI → DAPT, anticoagulation, urgent PCI (STEMI) / early invasive (NSTE-ACS).[4] }
Troponin rising post-op without ischaemic features → MINS — aspirin + statin + surveillance; NOT routine PCI.[9] }
BNP/NT-proBNP rising despite aggressive diuresis → refractory congestion — escalate (inotrope, ultrafiltration, mechanical support).[3] }
D-dimer + low fibrinogen + low platelets + prolonged PT → overt DIC (ISTH ≥5) — treat trigger, factor replacement, anticoagulate if thrombotic.[12] }
D-dimer normal in low-risk suspected PE → VTE excluded, NO imaging needed.[18] }
suPAR rising in septic shock → immune exhaustion / worse prognosis — escalate, reassess source control, consider immunomodulation.[14] }
Copeptin very high in refractory shock → AVP axis maximally activated — add vasopressin 0.03 U/min (catecholamine-sparing).[15] }
CRP rising after an initial fall → new infection, complication, treatment failure, drug fever, or non-infective flare.[5] }

Common biomarker interpretation traps

Renal failure falsely elevates troponin AND NT-proBNP — use the trend (rise-and-fall), not the absolute value.
Obesity lowers BNP/NT-proBNP — use lower rule-out thresholds or rely on echo / POCUS.
Sacubitril (ARNI) raises measured BNP (drug artefact) — use NT-proBNP for monitoring.
Major surgery / severe trauma / prolonged cardiac arrest raise procalcitonin (false positive for infection) — don't over-treat with antibiotics.
Pregnancy raises D-dimer physiologically — use pregnancy-adapted (YEARS) algorithm.
Beta-agonists / metformin / linezolid raise lactate without hypoperfusion (Type B) — treat cause, don't flood with fluids.
Early infection (<6 h) may have a normal procalcitonin — repeat at 12-24 h; do not withhold antibiotics if sepsis suspected.
Single troponin value is not diagnostic — MI requires a rise-and-fall pattern with ≥1 value >99th centile URL.

[1]

Trial cards — the evidence base

ANDROMEDA-SHOCK (Hernández 2019, JAMA) — peripheral perfusion vs lactate-targeted resuscitation

Multicentre RCT (Latin America, n=424) in septic shock: capillary refill time (CRT)-guided vs lactate-guided resuscitation.

28-day mortality: CRT 34.9% vs lactate 43.4% (RR 0.75, 95% CI 0.55-1.02; OR 0.61 in adjusted analysis; p=0.06 unadjusted but adjusted was significant)
Fluids given at 8 h: significantly LESS in CRT group (~1.8 L vs 2.3 L) — fewer harms of fluid overload
SOFA, vasopressor days, RRT use: favoured CRT strategy
CONCLUSION: A peripheral-perfusion (CRT, mottling, central-toe temperature) strategy may improve outcomes vs a lactate-chasing strategy in septic shock, in part by avoiding excessive fluid resuscitation. Practical: combine both — lactate for prognosis, perfusion for when to stop resuscitating.[17]

Jansen lactate clearance trial (2010, AJRCCM) — early lactate-guided therapy

Multicentre open-label RCT (n=348) of ventilated ICU patients with lactate ≥3.0 mmol/L: lactate-clearance-guided (target ≥20% per 2 h) vs no lactate guidance.

In-hospital mortality: lactate-guided 23% vs no guidance 33% (statistically significant in the subgroup with lactate ≥3.0 entry criterion — 20% vs 44% in severe shock subgroup)
Less organ failure in lactate-guided arm
CONCLUSION: Early lactate-guided therapy reduced mortality and organ failure in patients with hyperlactataemia. Lactate clearance ≥10-20%/h became the standard resuscitation target, integrated into the Surviving Sepsis Campaign 2021.[2][7]

MINS — the VISION cohort data (Devereaux, Curr Opin Cardiol 2014; AHA 2021 Scientific Statement)

Pooled cohort data from the international VISION (Vascular Events In Noncardiac Surgery Patients Cohort Evaluation) study (>30,000 patients across multiple countries):

Incidence of MINS: ~8% of all non-cardiac surgery; 20-30% in vascular surgery; higher with prior CAD, HF, stroke, emergency surgery
30-day mortality post-MINS: ~9% (3-4× higher than patients without MINS)
Independent association with cardiac arrest, cardiogenic shock, acute HF, stroke — even when asymptomatic
~85% of MINS cases have no ischaemic symptoms — detected by routine postoperative troponin screening (AHA recommends hs-cTn at 24-48 h post-op in high-risk patients)
AHA 2021 management: aspirin + statin, glycaemic control, intensive surveillance, cardiology follow-up; PCI reserved for high-risk trajectories (ischaemic features / dynamic ECG)
PREVENT-MINS (Szczeklik 2025, Circulation): ivabradine did NOT reduce MINS — supports the supply-demand (rate-independent) pathophysiology in most cases.[9][10][11]

Backes 2012 systematic review (Intensive Care Med) — suPAR as biomarker in systemic inflammation

Systematic review of suPAR in sepsis, septic shock, and systemic inflammation across 14 studies (>3,500 patients):

suPAR is an independent predictor of mortality in sepsis and septic shock, often superior to CRP, procalcitonin, IL-6 for prognosis
Higher suPAR → higher 28-day and 90-day mortality; cut-off ~12 ng/mL carries mortality >50%
Less useful for DIAGNOSIS of infection (PCT remains better for distinguishing bacterial from non-bacterial)
Confirmed by Gussen 2019 (J Intensive Care): neutrophils are the dominant source of circulating suPAR; suPAR marks exhausted neutrophil function — high suPAR = dysfunctional, migratory-impaired neutrophils → worse outcome
Status: not yet routine bedside in most ICUs, but a high-yield exam concept (immunoparalysis, neutrophil exhaustion).[13][14]

van Es 2017 IPD meta-analysis (J Thromb Haemost) — Wells rule + age-adjusted D-dimer for PE exclusion

Individual patient data meta-analysis (>7,000 patients with suspected PE) of the original and simplified Wells rules combined with age-adjusted D-dimer cut-off (age × 10 µg/L FEU in >50 yr):

3-month VTE failure rate after PE exclusion: <1.5% (well below the 3% safety threshold)
Age-adjusted cut-off safely increased the proportion of older patients in whom PE could be ruled out by D-dimer alone (without imaging) — from ~6% to ~30% in >50 yr
CONCLUSION: A negative D-dimer (using the age-adjusted threshold) combined with a low/intermediate Wells score safely excludes PE. In ICU, D-dimer is rarely rule-out useful (most critically ill have elevated D-dimer) — rely on imaging, except for low-risk young patients.[18]

Lee 2022 BMJ systematic review — NT-proBNP rule-in/rule-out pathway for acute heart failure

Meta-analysis (>40,000 patients with acute dyspnoea) validating NT-proBNP thresholds for acute heart failure:

Rule-out: NT-proBNP <300 ng/L → NPV ~98% across all ages (heart failure unlikely)
Rule-in (age-stratified): <50 yr >450 ng/L; 50-75 yr >900 ng/L; >75 yr >1800 ng/L → specificity ~90%
Confounders: renal failure raises (use higher thresholds or alternative diagnosis), obesity lowers (use 50% lower thresholds), AF raises independently
CONCLUSION: NT-proBNP is a powerful rule-out for acute heart failure in the dyspnoeic ICU patient; rule-in is age-stratified and confounder-aware. Trending NT-proBNP during decongestion predicts outcome.[3]

Special populations

Biomarker interpretation pitfalls by population

Population	Pitfall	Workaround
Renal failure (eGFR <30, dialysis)	Troponin and NT-proBNP chronically elevated	Use delta troponin (>20% rise) for acute injury; rely on echo, JVP, fluid balance
Obesity (BMI ≥30)	BNP/NT-proBNP falsely low	Use obesity cut-offs (50% lower); use POCUS B-lines / JVP / echo
Pregnancy	D-dimer physiologically elevated from 1st trimester; alkaline phosphatase ↑	Use YEARS algorithm; do NOT use plain D-dimer threshold
Cirrhosis / liver failure	Lactate clearance impaired; coagulopathy (PT, fibrinogen) altered; CRP often low (decreased hepatic synthesis)	Lactate may be mildly high without hypoperfusion; do not over-resuscitate
Cardiac surgery / post-op	Troponin, BNP, CRP all elevated at baseline	Use trends; post-CABG troponin rise is expected (surgical injury) — MINS definition (≥99th centile) still applies but threshold often higher
Malignancy	D-dimer chronically high; CRP high; suPAR high	D-dimer cannot rule out VTE — image; suPAR adds prognosis
Neutropenia / post-chemotherapy	PCT, CRP may be blunted; suPAR high	Do not rely on fever or PCT — start empiric antibiotics in any febrile neutropenic; serial CRP/PCT
On sacubitril/valsartan (ARNI)	BNP falsely high	Use NT-proBNP

[1]

Exam synthesis — CICM / FFICM / EDIC one-minute answer

The exam-friendly synthesis of ICU biomarkers

ICU biomarkers complement, not replace, clinical assessment.

Procalcitonin (PCT): bacterial infection marker → antibiotic stewardship (PRORATA: stop at PCT <0.5 or ≥80% fall). NEVER delay empiric antibiotics in sepsis.
Lactate: hypoperfusion/anaerobiosis → resuscitation target (clearance ≥10%/h, Jansen 2010); ≥2 mmol/L = Sepsis-3 organ dysfunction; ≥4 + vasopressors = septic shock. Combine with peripheral perfusion (ANDROMEDA-SHOCK 2019 — CRT-targeted may outperform lactate-chasing).
Troponin (hs-cTn): myocardial injury → distinguish type 1 MI (plaque rupture → PCI/DAPT) from type 2 MI (supply-demand → treat cause) from MINS (post-op, often silent, aspirin + statin; AHA 2021).
BNP/NT-proBNP: cardiac stretch → heart failure rule-out (<300 ng/L NT-proBNP ≈ rules out) and prognosis; renal failure raises, obesity lowers, ARNI raises BNP (use NT-proBNP).
CRP: non-specific inflammation → use trends, lags PCT by 24-48 h.
D-dimer: fibrin turnover → rules OUT VTE/DIC (high NPV with age-adjusted cut-off) but rules nothing in; ISTH overt DIC score uses magnitude of D-dimer rise.
suPAR: immune activation / neutrophil exhaustion → prognostic for mortality in sepsis, independent of severity scores; not yet routine bedside.
Copeptin (CT-proAVP): vasopressin surrogate → shock prognosis; high copeptin + refractory shock → consider exogenous vasopressin; also used for early MI rule-out with troponin.

[1]

One-glance biomarker-cause summary table

Marker up	Bacterial infection	Hypoperfusion	Cardiac	Inflammation	Thrombosis	Stress / prognosis
Procalcitonin	✓✓✓	—	—	✓ (surgery, trauma)	—	✓
Lactate	✓ (sepsis)	✓✓✓	✓ (shock)	—	—	✓✓
Troponin	—	✓ (shock)	✓✓✓	✓ (myocarditis)	✓ (type 1)	✓
BNP/NT-proBNP	—	—	✓✓✓	✓ (sepsis)	—	✓
CRP	✓	—	—	✓✓✓	—	—
D-dimer	—	—	—	✓ (DIC)	✓✓✓ (VTE/DIC)	—
suPAR	✓✓	—	—	✓✓	—	✓✓✓
Copeptin	—	✓ (shock)	✓ (stress)	—	—	✓✓✓

[1]

Exam practice

SAQ — Procalcitonin-guided antibiotic stewardship in severe community-acquired pneumonia

10 minutes · 10 marks

A 68-year-old man is admitted to ICU with a 3-day history of fever, productive cough and pleuritic chest pain. He is confused, HR 112, BP 92/55 (MAP 67), RR 30, SpO2 91% on 6 L nasal prongs. Chest X-ray shows right middle and lower lobe consolidation. Lactate 2.6 mmol/L, WCC 18.5, CRP 185 mg/L, procalcitonin 8.4 ng/mL. Blood cultures have been taken and empirical piperacillin-tazobactam started.

[1]

SAQ — Postoperative troponin elevation: distinguishing MINS, type 1 MI and type 2 MI

10 minutes · 10 marks

A 72-year-old man with a background of hypertension and type 2 diabetes undergoes an emergency laparotomy for perforated diverticulitis. Twelve hours post-op in ICU he develops new atrial fibrillation with rapid ventricular response (HR 140), SpO2 92% on 2 L, BP 88/50 (MAP 63). He has no chest pain. High-sensitivity troponin T (hs-cTnT) is 14 ng/L pre-op, 280 ng/L at 12 hours post-op, and 360 ng/L at 15 hours (99th centile URL 40 ng/L). ECG shows AF with lateral T-wave inversion. Lactate 2.4 mmol/L. Haemoglobin 96 g/L.

[1]

References

[1]Bouadma L, Luyt CE, Tubach F, et al. Use of procalcitonin to reduce patients' exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial Lancet, 2010.PMID 20097417
[2]Jansen TC, van Bommel J, Schoonderbeek FJ, et al. Early lactate-guided therapy in intensive care unit patients: a multicenter, open-label, randomized controlled trial Am J Respir Crit Care Med, 2010.PMID 20463176
[3]Lee KK, Doudesis D, Anwar M, et al. Development and validation of a decision support tool for the diagnosis of acute heart failure: systematic review, meta-analysis, and modelling study BMJ, 2022.PMID 35697365
[4]Chapman AR, Adamson PD, Shah ASV, et al. High-Sensitivity Cardiac Troponin and the Universal Definition of Myocardial Infarction Circulation, 2020.PMID 31587565
[5]Póvoa P. C-reactive protein: a valuable marker of sepsis Intensive Care Med, 2002.PMID 11904651
[6]Vincent JL, Bakker J. Blood lactate levels in sepsis: in 8 questions Curr Opin Crit Care, 2021.PMID 33852499
[7]Evans L, Rhodes A, Alhazzani W, et al. Executive Summary: Surviving Sepsis Campaign: International Guidelines for the Management of Sepsis and Septic Shock 2021 Crit Care Med, 2021.PMID 34643578
[8]Bularga A, Hung J, Daghem M, et al. Coronary Artery and Cardiac Disease in Patients With Type 2 Myocardial Infarction: A Prospective Cohort Study Circulation, 2022.PMID 35341327
[9]Ruetzler K, Smilowitz NR, Berger JS, et al. Diagnosis and Management of Patients With Myocardial Injury After Noncardiac Surgery: A Scientific Statement From the American Heart Association Circulation, 2021.PMID 34601955
[10]Khan J, Alonso-Coello P, Devereaux PJ. Myocardial injury after noncardiac surgery Curr Opin Cardiol, 2014.PMID 25029449
[11]Ekeloef S, Alamili M, Devereaux PJ, et al. Troponin elevations after non-cardiac, non-vascular surgery are predictive of major adverse cardiac events and mortality: a systematic review and meta-analysis Br J Anaesth, 2016.PMID 27799170
[12]Levi M, Toh CH, Thachil J, et al. Guidelines for the diagnosis and management of disseminated intravascular coagulation. British Committee for Standards in Haematology Br J Haematol, 2009.PMID 19222477
[13]Backes Y, van der Sluijs KF, Mackie DP, et al. Usefulness of suPAR as a biological marker in patients with systemic inflammation or infection: a systematic review Intensive Care Med, 2012.PMID 22706919
[14]Gussen H, Hohlstein P, Bartneck M, et al. Neutrophils are a main source of circulating suPAR predicting outcome in critical illness J Intensive Care, 2019.PMID 31061709
[15]Lee JH, Chan YH, Lai OF, et al. Vasopressin and copeptin levels in children with sepsis and septic shock Intensive Care Med, 2013.PMID 23344831
[16]Dabla PK, Dabla V, Arora S. Co-peptin: Role as a novel biomarker in clinical practice Clin Chim Acta, 2011.PMID 20920496
[17]Hernández G, Ospina-Tascón GA, Damiani LP, et al. Effect of a Resuscitation Strategy Targeting Peripheral Perfusion Status vs Serum Lactate Levels on 28-Day Mortality Among Patients With Septic Shock: The ANDROMEDA-SHOCK Randomized Clinical Trial JAMA, 2019.PMID 30772908
[18]van Es N, Kraaijpoel N, Klok FA, et al. The original and simplified Wells rules and age-adjusted D-dimer testing to rule out pulmonary embolism: an individual patient data meta-analysis J Thromb Haemost, 2017.PMID 28106338

[1]

Key ICU biomarkers comparison

Biomarker	What it measures	When elevated	Clinical use	Limitations
Procalcitonin	Bacterial infection (calcitonin precursor)	Bacterial sepsis (>2 ng/mL), bacterial pneumonia; NOT viral/non-infective	Antibiotic stewardship (stop when <0.5 or falls ≥80%)	False positive (trauma, surgery, cardiogenic shock), false negative (first 6-12 h, localised)
Lactate	Tissue hypoperfusion / anaerobic metabolism	Shock, sepsis, mesenteric ischaemia, seizures, burns	Guide resuscitation (clearance >10%/h), prognosis	Also raised: β2-agonists, malignancy, mitochondrial toxins, liver failure
Troponin / hs-cTn	Myocardial injury (cardiomyocyte necrosis)	Type 1 MI, type 2 MI, MINS, sepsis, PE, myocarditis, renal failure	Diagnose MI, grade cardiac injury, prognostic	Type 2 in ICU (not atherosclerotic) — do NOT reflexively PCI; renal failure elevates
BNP / NT-proBNP	Cardiac stretch / wall tension (volume overload)	Heart failure, volume overload, PE, sepsis	Diagnose heart failure, guide diuresis, prognostic	Raised: renal failure, PE, sepsis, age, AF; LOWER in obesity (haemodilution/clearance)
CRP	Acute-phase inflammation	Infection, inflammation, tissue injury, malignancy, autoimmune	Trend (rising = worsening), monitor infection response	Nonspecific, slow (24-48 h lag), many causes
D-dimer	Cross-linked fibrin degradation (coagulation + fibrinolysis)	VTE (DVT/PE), DIC, arterial thrombosis, inflammation, malignancy, pregnancy, surgery	Rule OUT VTE/DIC (high NPV); ISTH DIC scoring	Non-specific (low specificity); rises with age — use age-adjusted cut-off (age × 10 µg/L in >50 yr)
suPAR	Soluble urokinase plasminogen activator receptor (immune activation)	Sepsis, septic shock, multi-organ failure, inflammation	Prognostic — predicts mortality independent of severity scores	Nonspecific; influenced by renal function, malignancy; not yet routine bedside
Copeptin (CT-proAVP)	C-terminal fragment of pro-arginine-vasopressin (AVP surrogate)	Shock, sepsis, stress, hypotension, hyponatraemia	Surrogate for hard-to-measure AVP; prognostic in shock; guides vasopressin consideration	Released with stress; non-specific; influenced by osmolality, fluid status

[1]

Practical use of biomarkers in septic shock

On presentation — lactate (baseline — guide resuscitation), procalcitonin (baseline — guide antibiotics), CRP (baseline — trend later)
Resuscitation guided by lactate — target lactate clearance >10%/h (or normalisation within 2-6h). If not clearing → ongoing hypoperfusion (assess volume, cardiac output, source control)
Procalcitonin trend — if PCT falls >80% from baseline or <0.5 → consider stopping antibiotics (PROORATA algorithm). Repeat daily
Troponin — check if cardiac symptoms, arrhythmia, or unexplained hypotension. Elevated → assess for type 1 (ECG, ACS) vs type 2 MI (treat underlying cause)
BNP/NT-proBNP — check if heart failure suspected (dyspnoea, oedema). Low → heart failure unlikely. High → heart failure or other cause (renal, PE, sepsis)
Serial CRP — trend every 24-48h. Falling → responding. Rising or persistently high → treatment failure, complication, ongoing infection

[1]

Clinical pearls

High-yield ICU biomarker points for CICM/FFICM exam

Procalcitonin guides ANTIBIOTIC DURATION, not initiation. PCT is a bacterial infection marker — rises in bacterial sepsis, stays low in viral/non-infective. PRORATA trial: PCT-guided algorithm (stop antibiotics when PCT <0.5 or falls 80%) reduced antibiotic days by 2-3 days WITHOUT increasing adverse outcomes. BUT: still start empiric antibiotics for suspected sepsis (don't wait for PCT).[1] }
Lactate clearance >10%/h is the target in septic shock. Jansen (Am J Resp Crit Care 2010): lactate clearance-guided resuscitation reduced mortality in lactate ≥3.0. Measure every 2h. If clearance <10%: reassess (more fluids? inotropes? source control?). Lactate normalisation (≤2) is the goal. CLEARED lactate = adequate resuscitation.[2] }
Troponin in ICU is often TYPE 2 MI (supply-demand), not type 1 (atherosclerotic). Type 2: tachycardia, hypotension, hypoxia, sepsis → supply-demand mismatch → myocardial injury → troponin rise. Treatment: address underlying cause (oxygenate, blood pressure, rate control). Type 1: plaque rupture → thrombus → need antiplatelet, anticoagulant, PCI. CLINICAL: don't automatically PCI every elevated troponin in ICU — assess type (ECG changes? chest pain? ACS features?).[4] }
BNP/NT-proBNP: heart failure diagnosis and prognosis. Low (<300 NT-proBNP) → heart failure UNLIKELY (negative predictive value 90%). High → heart failure likely (but false positives: renal failure, PE, sepsis, age, AF). Prognosis: higher BNP = worse outcome in ICU. Also guides diuresis (reducing BNP = decongestion).[3] }
CRP is NONSPECIFIC — use trends, not single values. CRP rises with ANY inflammation (infection, surgery, trauma, malignancy, autoimmune). Rises 24-48h after insult (slow). Trend (every 24-48h): falling = responding to treatment; rising/persistent = treatment failure. Procalcitonin is MORE SPECIFIC for bacterial infection than CRP.[5] }
Lactate sources — not all elevated lactate is hypoperfusion. TYPE A (hypoperfusion): shock, sepsis, mesenteric ischaemia, CO poisoning. TYPE B (non-hypoperfusion): beta-agonists (salbutamol), malignancy (lymphoma), mitochondrial toxins (metformin, cyanide), thiamine deficiency, seizures (muscle), liver failure (impaired clearance). If lactate high but well perfused → consider type B.[6] }
Procalcitonin kinetics — rising or persistently high suggests ongoing infection. PCT normally falls 50% every 24h with effective antibiotics. If NOT falling: (1) Inadequate antibiotic (resistant organism, wrong drug). (2) Inadequate source control (undrained abscess). (3) New infection. (4) Non-bacterial cause (persistent PCT without bacterial infection — rare). Rising PCT = reassess.[1] }
Lactate >2 is sepsis criterion (qSOFA + lactate). Sepsis-3 definition: suspected infection + SOFA ≥2 (includes lactate >2 as organ dysfunction). Lactate >2 with infection = sepsis (even if normotensive). Lactate >2 + vasopressor requirement = septic shock. Measure lactate in ALL suspected sepsis.[2] }
NT-proBNP vs BNP — different half-lives and thresholds. BNP: half-life 20 min (rapid changes). NT-proBNP: half-life 120 min (more stable, higher values). Both reflect heart failure. NT-proBNP preferred in ICU (more stable, wider dynamic range). Thresholds vary by age, renal function. LOW NT-proBNP excludes heart failure well; HIGH is less specific.[3] }
High-sensitivity troponin (hs-cTn) — detects smaller injury. hs-cTn assays detect myocardial injury at MUCH lower levels (10x more sensitive than conventional). More ICU patients have 'elevated' troponin (any minor injury detected). Interpret carefully — many ICU patients have mildly elevated hs-cTn (not all are MI). Trend matters (rising = ongoing injury).[4] }
Procalcitonin false positives and negatives. FALSE POSITIVE (high PCT without bacterial infection): major surgery, severe trauma, cardiogenic shock, heat stroke, prolonged cardiac arrest, medullary thyroid cancer (produces calcitonin). FALSE NEGATIVE (low PCT with bacterial infection): early infection (PCT rises 6-12h), localized infection (low bacterial load), after antibiotics started. Interpret in clinical context.[1] }
Lactate/pyruvate ratio distinguishes ischaemic from non-ischaemic. Ischaemic (Type A): lactate/pyruvate ratio HIGH (>10-15) — anaerobic metabolism (pyruvate shunted to lactate). Non-ischaemic (Type B): lactate/pyruvate ratio NORMAL — mitochondrial dysfunction (impaired pyruvate oxidation). Not routinely measured, but useful in research/complex cases.[6] }
CRP vs procalcitonin for infection diagnosis. PCT is MORE SPECIFIC for BACTERIAL infection (low in viral). CRP is elevated in ANY inflammation (less specific). Both rise in bacterial infection. PCT rises FASTER (6-12h vs 24-48h for CRP). PCT falls faster with effective treatment. For antibiotic stewardship: PCT preferred (more dynamic, specific).[5] }
Combining biomarkers improves prediction. No single biomarker is perfect. COMBINING: lactate (perfusion) + PCT (bacterial infection) + CRP (inflammation) + clinical assessment → better diagnosis and prognosis than any alone. Example: septic shock with high lactate + high PCT + rising CRP → severe bacterial sepsis, aggressive treatment needed. Low lactate + low PCT + falling CRP → resolving, consider de-escalation.[1] }

Red flags

Prognosis

PRORATA trial (Bouadma 2010, Lancet) — procalcitonin-guided antibiotic therapy

RCT: 621 ICU patients with suspected bacterial infection. Procalcitonin-guided (stop antibiotics when PCT <0.5 or fell ≥80%) vs standard (physician decision).

Antibiotic days (days 1-28): PCT 9.3 vs standard 14.9 (p<0.0001) — REDUCED by 5.6 days
Mortality (day 28): PCT 20.3% vs standard 20.4% (NO difference — safe)
Relapse of infection: similar
CONCLUSION: Procalcitonin-guided antibiotic therapy REDUCES antibiotic exposure (5.6 days) without increasing adverse outcomes. Now standard in many ICU antibiotic stewardship programmes. [1]

[1]

Lactate — deep dive (perfusion, clearance, prognostics)

Lactate Type A (hypoperfusion / ischaemic) vs Type B (non-hypoperfusion / impaired clearance)

Feature	Type A — hypoperfusion / tissue ischaemia	Type B — non-hypoperfusion (impaired metabolism or clearance)
Mechanism	Anaerobic glycolysis from oxygen debt; pyruvate shunted to lactate	Aerobic glycolysis (β2-agonists), impaired oxidation (mitochondrial toxins), reduced hepatic clearance
Lactate/pyruvate ratio	High (>10-15)	Normal
Examples	Septic / haemorrhagic / cardiogenic shock, mesenteric ischaemia, severe burns, limb ischaemia, CO poisoning, seizures (delayed)	Salbutamol / ritodrine / adrenaline infusions, metformin (MALA), cyanide, thiamine deficiency (beriberi), malignancy (lymphoma, leukaemia — Warburg), liver failure, ethanol, propofol infusion, linezolid
Clinical sign	Cool, mottled, prolonged cap refill, oliguria, hypotension	Warm, well-perfused, normal ScvO2, may be haemodynamically stable
Management	Resuscitate (fluids, blood, inotropes), source control, restore oxygen delivery	Remove trigger (stop β2-agonist / metformin), haemodialysis (metformin, linezolid), thiamine (alcoholic), treat malignancy
Prognosis	Mortality ∝ lactate level and clearance half-life	Driven by underlying cause

[1]

Lactate clearance protocol in septic shock (Jansen 2010 algorithm)

Draw baseline lactate at recognition of septic shock / suspected sepsis with lactate ≥3.0 mmol/L
Resuscitate with crystalloid 30 mL/kg bolus (Surviving Sepsis Campaign) + early broad-spectrum antibiotics within 1 h + source control
Recheck lactate at 2 h — calculate clearance: (baseline − current) / baseline × 100
If clearance ≥10%/h → on target; continue therapy, recheck at 4-6 h
If clearance <10%/h → reassess: echocardiography (cardiac output / fluid responsiveness), MAP target ≥65 mmHg, consider vasopressor (noradrenaline), inotrope (dobutamine) if low CO, source control (drain abscess, debride tissue), consider ongoing occult bleeding
If lactate rising despite resuscitation → red flag: mesenteric ischaemia, mitochondrial dysfunction (severe sepsis), β2-agonist, metformin toxicity, limb/bowel ischaemia. Escalate: surgical review, CT angiography
Target — lactate normalisation (≤2 mmol/L) within 2-6 h; persistent hyperlactataemia at 24 h = poor prognosis

[1]

Troponin in ICU — type 1 MI, type 2 MI, and MINS

Type 1 MI vs Type 2 MI vs MINS — distinguishing myocardial injury in ICU

Feature	Type 1 MI	Type 2 MI	MINS (myocardial injury after non-cardiac surgery)
Mechanism	Atherosclerotic plaque rupture/erosion → in-situ thrombus	Supply-demand mismatch (no plaque rupture)	Perioperative supply-demand mismatch ± non-ischaemic injury
Troponin	Rise AND fall, with ≥1 value >99th centile	Rise AND fall, with ≥1 value >99th centile	Peak troponin ≥99th centile within 30 d post-op (no ischaemic symptom required)
Symptoms	Ischaemic chest pain/symptoms	Often absent or atypical in ICU	Usually SILENT (85% without ischaemic features) — detected by routine troponin screening
ECG	ST-elevation, ST-depression, T-wave inversion, new LBBB	Often non-specific, sinus tachycardia, or changes of underlying cause	Non-specific; new ST changes uncommon
Imaging	Regional wall motion abnormality, plaque on angiography	May be normal or fixed defect	Variable
Management	DAPT, anticoagulation, urgent PCI (STEMI) / early invasive strategy (NSTE-ACS)	Treat UNDERLYING cause (oxygen, BP, rate, sepsis, anaemia) — routine PCI not indicated	Risk stratify, address modifiable factors, start aspirin + statin, cardiology follow-up; troponin trend alone
Prognosis	Driven by infarct size and treatment	Mortality ~2-3× non-injured ICU patients	30-day mortality ~8-10%; 1-year mortality ~15-20%

[1]

MINS — myocardial injury after non-cardiac surgery (AHA 2021 Scientific Statement)

Troponin elevation in ICU — how to distinguish type 1 MI from type 2 MI / MINS

Confirm rise/fall pattern with serial hs-cTn (0, 1, 3 h). Isolated single rise ≠ acute MI — consider chronic injury (renal failure, structural heart disease)
Assess ischaemic symptoms — chest pain, dyspnoea, diaphoresis, radiation (absent in many type 2 / MINS)
ECG — look for ST-elevation (STE), ST-depression, T-wave inversion, new LBBB, posterior MI. Dynamic ST changes favour type 1. Diffuse / non-specific changes favour type 2
Bedside echo — new regional wall motion abnormality (especially in a coronary territory) favours type 1; global hypokinesis / EF reduction favours type 2
Identify supply-demand drivers of type 2 — tachyarrhythmia (AF with RVR), hypoxia, hypotension, severe anaemia, sepsis, severe hypertension
Confirm non-cardiac triggers of troponin rise — PE, myocarditis, sepsis, renal failure, cardioversion, seizures (rule these out before calling it MI)
Decision — STEMI / NSTE-ACS (type 1) → DAPT + anticoag + cardiology + early invasive strategy. Type 2 → treat underlying cause, monitor, antiplatelet/ statin if 10-yr ASCVD risk warrants. MINS → aspirin + statin + surveillance

[1]

BNP / NT-proBNP — heart failure diagnosis, volume status, prognosis

BNP vs NT-proBNP — what's the difference?

Property	BNP	NT-proBNP
Origin	Cleaved from pro-BNP by corin → active BNP + inert NT-proBNP	Inert N-terminal fragment (released in 1:1 stoichiometry)
Half-life	~20 min	~120 min
Stability	Less stable (requires cold processing)	More stable (room temperature OK for hours)
Removed by	Neutral endopeptidases + clearance receptors	Renal clearance predominantly
Affected by neprilysin inhibitors (sacubitril)	↑ measured BNP (false high)	No significant effect → preferred in patients on ARNI
Bedside assay	Available	Most labs batch; not always point-of-care
Heart-failure rule-out (dyspnoea)	BNP <100 pg/mL	NT-proBNP <300 pg/mL (rule-out across all ages)

[1]

D-dimer — fibrin turnover in VTE, DIC, arterial thrombosis

D-dimer elevation — causes and significance

Cause	Typical magnitude	Use
Acute VTE (DVT, PE)	Moderately-high	Rule OUT (negative D-dimer + low pre-test probability → VTE excluded, no imaging needed). High NPV (≥97%)
DIC (overt)	Very high + falling fibrinogen + rising PT + low platelets	ISTH overt DIC score — D-dimer contributes 0-3 points
Arterial thrombosis (MI, stroke, mesenteric)	Mild-moderate	Not diagnostic — supports prothrombotic state
Aortic dissection / aneurysm	Mild-moderate	Aortic Dissection Detection Risk Score uses D-dimer >500 µg/L
Pregnancy (3rd trimester)	Physiologically elevated	Use pregnancy-adapted YEARS algorithm / PE rule-out criteria
Malignancy, inflammation, sepsis, post-op, trauma	Variable, often high	Non-specific; cannot rule out VTE reliably
Ageing	Rises ~10 µg/L per year	Use age-adjusted cut-off: age × 10 µg/L FEU in >50 yr

[1]

D-dimer interpretation in suspected PE in ICU

Stratify pre-test probability — Wells score (low/moderate/high) or PERC if low-risk and young. Imaging is the rule in unstable patients
Low/intermediate pre-test probability → D-dimer useful. Normal D-dimer (with sensitive assay, incl. age-adjusted cut-off in >50 yr) excludes PE — stop workup. The 2017 van Es IPD meta-analysis confirmed Wells + age-adjusted D-dimer gives a 3-month VTE failure rate <1.5% in PE-rule-out patients, including older patients.[18]
High pre-test probability — do NOT rely on D-dimer (specificity too low). Go straight to CT pulmonary angiography
Haemodynamically unstable — bedside echo (RV strain), in-lab pulmonary angiography if unstable for CT. D-dimer is academic
Pregnancy — use YEARS algorithm: D-dimer + clinical signs of DVT + haemoptysis + most-symptomatic leg is PE → CT; if not, use D-dimer cut-off (rule-out at <1000 if no PE signs; <500 if signs present)
Critically ill patient — D-dimer almost always elevated (sepsis, surgery, trauma, lines) → low rule-out utility; rely on imaging. The value in ICU is in trajectory (rising in DIC, falling in resolving thrombosis) and ISTH DIC scoring

suPAR — soluble urokinase-type plasminogen activator receptor

suPAR — prognostic inflammation marker in sepsis and critical illness

Copeptin — vasopressin surrogate

Copeptin (CT-proAVP) — measuring the unmeasurable vasopressin

Procalcitonin — antibiotic stewardship algorithm

PRORATA-based procalcitonin algorithm for antibiotic de-escalation

Draw baseline PCT at antibiotic initiation (do NOT delay empiric antibiotics for suspected sepsis)
Day 1-3 (escalation phase) — continue antibiotics. Recheck PCT daily
Stop antibiotics if PCT <0.5 ng/mL OR has fallen ≥80% from peak (and clinical improvement: defervescence, falling WBC, haemodynamic stability, source controlled)
Continue if PCT >0.5 ng/mL AND <80% fall — reassess for: inadequate source control, resistant organism, wrong drug, new infection, non-bacterial cause (PCT can stay high in cardiogenic shock, severe trauma, medullary thyroid cancer)
Special populations — PCT less reliable in neonates, localised infection (endocarditis, abscess), after major surgery/trauma (false positive), early infection <6 h (false negative), and immunosuppressed patients
NEVER withhold or delay antibiotics in septic shock based on a single PCT value — clinical judgement overrides

[1]

Procalcitonin vs CRP — which is the better infection biomarker?

Feature	Procalcitonin	CRP
Biologically specific for bacterial infection	Yes (calcitonin precursor from extra-thyroidal tissues)	No — generic acute-phase reactant
Onset after insult	6-12 h	24-48 h (slower)
Falls with effective therapy	~50% per day	Slow (days)
Useful for antibiotic stewardship	Yes — PRORATA, SAPS, SABS validated algorithms	No — too slow and non-specific
Differentiates bacterial vs viral	Reasonably (low in pure viral)	No
False positives	Major surgery, severe trauma, cardiogenic shock, prolonged cardiac arrest, medulla of thyroid (calcitonin-producing tumour)	Any inflammation, malignancy, autoimmune, tissue injury
Cost / availability	More expensive	Cheap, universal

[1]

Combining biomarkers — multimarker panels improve prediction

How to combine biomarkers in common ICU scenarios

Scenario	Leading biomarkers	What each tells you	Integrated interpretation
Septic shock	Lactate + PCT + CRP + suPAR	Lactate → perfusion; PCT → bacterial likelihood; CRP → inflammation; suPAR → prognosis	High lactate + high PCT + rising suPAR = severe, escalate; clearance of lactate + falling PCT = improving
Acute dyspnoea (HF vs COPD)	BNP/NT-proBNP	Low → not HF; high → likely HF (mind renal/AF/obesity)	Combine with echo + clinical to confirm
Post-op troponin rise	hs-cTn trend	MINS vs type 1 MI	Aspirin + statin; invasive strategy only if ischaemic features
Suspected PE	D-dimer (low pre-test)	Normal D-dimer rules out PE in low-risk	If elevated → CTPA
Bleeding + abnormal coagulation	D-dimer + fibrinogen + PT + platelets	Overt DIC if ISTH ≥5	Treat trigger + factor replacement
Septic shock on rising vasopressors	Copeptin + lactate + ScvO2	Copeptin → AVP reserve; lactate → perfusion; ScvO2 → oxygen utilisation	Low copeptin (vasopressin depleted) + refractory shock → add vasopressin 0.03 U/min

[1]

Clinical pearls — second set (deep dive)

High-yield ICU biomarker pearls — set 2 (CICM / FFICM / EDIC)

Lactate clearance is not the same as lactate normalisation. Clearance ≥10%/h is a rate; normalisation (≤2 mmol/L) is a goal. Both matter — Jansen used clearance (target ≥10%/h), ANDROMEDA-SHOCK questioned whether chasing lactate with more fluids is harmful. Best practice: resuscitate to perfusion endpoints (CRT, mottling, MAP, urine output) AND track lactate — do not chase a number with fluid alone.[2][17]
A normal lactate does not exclude shock. Early septic shock (compensated), cardiogenic shock with high output, distributive shock — lactate may be normal initially. Recheck at 2-4 h. Persistent normal lactate with hypotension → consider adrenal insufficiency, anaphylaxis, vasoplegia from toxins.
Sepsis-3 definitions anchor on lactate. Lactate >2 mmol/L + suspected infection = sepsis (organ dysfunction). Lactate >2 + vasopressor requirement to keep MAP ≥65 = septic shock. Measure lactate in EVERY suspected sepsis — it triages severity, triggers the 1-h bundle, and predicts mortality (lactate ≥4 doubles mortality).[7]
Beta-hydroxybutyrate vs total lactate in DKA. Don't be fooled — DKA ketones (β-hydroxybutyrate) can give a mild lactate rise via altered redox, but the dominant acid is ketoacidosis. A high lactate (>5) in apparent DKA → look for sepsis, mesenteric ischaemia, or metformin.
Lactate >10 mmol/L has mortality >70% in septic shock — this is a marker of catastrophic hypoperfusion or mitochondrial dysfunction. Combined with vasopressor-refractory shock, consider sources (mesenteric ischaemia — lactate ≥5 + abdominal pain, sudden rise), and discuss goals of care early.[6]
MINS is prognostic, not automatically ACS. ~85% of perioperative troponin elevations have NO ischaemic features. The temptation to anticoagulate / PCI every post-op troponin rise is to be resisted. AHA 2021 supports aspirin + statin + surveillance, reserving invasive management for rising-and-falling troponin with ischaemic symptoms / ECG changes. PCI in MINS without ACS has not been shown to improve outcomes.[9][11]
Type 2 MI needs type-2 treatment. Treating tachyarrhythmia, hypoxia, anaemia, sepsis and shock is the therapy — not cath lab referral. Routine DAPT/anticoag in type 2 MI is not indicated unless co-existing type 1. The 2022 Bularga cohort (Circulation) confirmed that type 2 MI patients DO have a higher burden of coronary disease than previously assumed — so risk-stratify for secondary prevention, but the acute event is supply-demand.[8]
Renal failure elevates troponin and BNP. Both are cleared / metabolised by the kidney; eGFR <30 may double hs-cTn and triple NT-proBNP. In a dialysis patient with elevated troponin, the trend (rising-and-falling pattern) and the clinical picture matter more than the absolute value. Chronic troponin elevation in renal failure without dynamic change = chronic injury (NOT acute MI).[4]
BNP falls with decongestion; that is a GOOD sign. A falling BNP/NT-proBNP during diuresis of acute decompensated HF predicts lower readmission and mortality (haemodynamic decongestion). A persistently high or rising BNP despite diuresis = refractory congestion → consider ultrafiltration, inotrope (HFrEF), or mechanical support.
Obesity lowers BNP/NT-proBNP — use obesity cut-offs. BMI ≥30 can halve BNP even in true HF (haemodilution + increased clearance via adipose natriuretic peptide receptors). Consider a lower threshold (e.g., BNP <50 instead of <100) when ruling out HF in obesity — or rely on echo + JVP / POCUS (B-lines).
Sacubitril (in ARNI) increases measured BNP but not NT-proBNP. Because neprilysin inhibition blocks BNP breakdown, measured BNP rises within hours of ARNI initiation — this is a drug artefact, not worsening HF. Use NT-proBNP (not BNP) to monitor patients on ARNI. Stop ARNI for ≥36 h before switching from ACE-i (enalapril) to avoid angioedema.
CRP lags PCT by 24-48 h. In early sepsis, PCT rises faster (6-12 h) and falls faster (half-life ~24 h vs 48 h for CRP). A rising CRP with a falling PCT suggests resolving bacterial infection (CRP lag). A rising CRP with rising PCT suggests ongoing or worsening bacterial infection. Use both trends — PCT for direction of travel, CRP for confirmation.
suPAR is the only biomarker here that predicts mortality independent of disease severity scores. APACHE II, SOFA, SAPS — suPAR adds prognostic information beyond these. In septic shock, a suPAR >12 ng/mL has been associated with mortality >50%. Currently research-only in most ICUs but a high-yield exam topic.[13][14]
Copeptin + troponin doubles the rule-out power for MI in chest pain. A negative high-sensitivity troponin + low copeptin at presentation can rule out non-ST-elevation MI without serial sampling — FDA-approved pathway in some systems. Conceptually: troponin detects cardiomyocyte injury, copeptin detects the endogenous stress response (which rises within minutes of myocardial ischaemia). In ICU, copeptin adds shock-prognosis data.
D-dimer is an exclusion test, not a diagnostic test. The greatest utility of a normal D-dimer is ruling OUT VTE/DIC in low-pre-test-probability patients. An elevated D-dimer is meaningless in isolation (especially in ICU — sepsis, surgery, trauma, malignancy, pregnancy all raise it). The age-adjusted cut-off (age × 10 µg/L FEU in >50 yr) reduces false positives in older adults without missing PE.[18]
In overt DIC, treat the trigger — the lab will follow. DIC is never a primary diagnosis. Abruption, sepsis, APL, amniotic fluid embolism, trauma, snake bite — find and treat the cause while supporting haemostasis (cryoprecipitate to fibrinogen ≥1.5, platelets ≥50 if bleeding, FFP for prolonged PT). Therapeutic anticoagulation (not prophylactic) is reserved for thrombotic-dominant DIC with extensive organ thrombosis.[12]
Procalcitonin in pure viral infection is low — COVID-19, influenza, RSV. A high PCT in viral respiratory illness suggests bacterial co-infection (pneumococcal / staphylococcal superinfection) — the Surviving Sepsis COVID guidelines recommended PCT to identify bacterial superinfection and limit antibiotic exposure.
Beware the 'washout' lactate. After revascularisation of an ischaemic limb or bowel, lactate can briefly rise as accumulated lactate washes into circulation. Distinguish from worsening: trend over 1-2 h, assess perfusion clinically, and look for the source (surgery / clot lysis).
Troponin and renal failure — chronic vs acute. A stable, mildly elevated hs-cTn in a stable dialysis patient is chronic injury (NOT MI). A rising-and-falling pattern with clinical deterioration = acute myocardial injury. Use absolute change (delta) — a delta >20% (or >50% if baseline elevated) supports acute injury in renal failure.[4]
Don't interpret biomarkers in isolation. A high lactate + normal PCT + high CRP → consider non-bacterial causes (mesenteric ischaemia, malignancy, toxin, post-ictal). A normal lactate + high PCT → early bacterial infection before hypoperfusion. Always integrate biomarkers with history, examination, imaging, and clinical trajectory.

Red flags — extended set

Biomarker red flags demanding escalation

Lactate >4 mmol/L + vasopressor requirement → septic shock (Sepsis-3); start 1-h bundle.[7] }
Lactate rising despite adequate resuscitation → mesenteric ischaemia, mitochondrial dysfunction, ongoing bleeding, metformin toxicity → surgical/imaging review.[6] }
Lactate >10 mmol/L → catastrophic perfusion failure; mortality >70% — discuss goals of care.[6] }
Procalcitonin >2 ng/mL → bacterial sepsis likely; start antibiotics.[1] }
Procalcitonin not falling ≥80% by day 5 → reassess source control, resistant organism, wrong drug, new infection.[1] }
Troponin rising with chest pain + dynamic ST changes → type 1 MI → DAPT, anticoagulation, urgent PCI (STEMI) / early invasive (NSTE-ACS).[4] }
Troponin rising post-op without ischaemic features → MINS — aspirin + statin + surveillance; NOT routine PCI.[9] }
BNP/NT-proBNP rising despite aggressive diuresis → refractory congestion — escalate (inotrope, ultrafiltration, mechanical support).[3] }
D-dimer + low fibrinogen + low platelets + prolonged PT → overt DIC (ISTH ≥5) — treat trigger, factor replacement, anticoagulate if thrombotic.[12] }
D-dimer normal in low-risk suspected PE → VTE excluded, NO imaging needed.[18] }
suPAR rising in septic shock → immune exhaustion / worse prognosis — escalate, reassess source control, consider immunomodulation.[14] }
Copeptin very high in refractory shock → AVP axis maximally activated — add vasopressin 0.03 U/min (catecholamine-sparing).[15] }
CRP rising after an initial fall → new infection, complication, treatment failure, drug fever, or non-infective flare.[5] }

Common biomarker interpretation traps

Renal failure falsely elevates troponin AND NT-proBNP — use the trend (rise-and-fall), not the absolute value.
Obesity lowers BNP/NT-proBNP — use lower rule-out thresholds or rely on echo / POCUS.
Sacubitril (ARNI) raises measured BNP (drug artefact) — use NT-proBNP for monitoring.
Major surgery / severe trauma / prolonged cardiac arrest raise procalcitonin (false positive for infection) — don't over-treat with antibiotics.
Pregnancy raises D-dimer physiologically — use pregnancy-adapted (YEARS) algorithm.
Beta-agonists / metformin / linezolid raise lactate without hypoperfusion (Type B) — treat cause, don't flood with fluids.
Early infection (<6 h) may have a normal procalcitonin — repeat at 12-24 h; do not withhold antibiotics if sepsis suspected.
Single troponin value is not diagnostic — MI requires a rise-and-fall pattern with ≥1 value >99th centile URL.

[1]

Trial cards — the evidence base

ANDROMEDA-SHOCK (Hernández 2019, JAMA) — peripheral perfusion vs lactate-targeted resuscitation

Multicentre RCT (Latin America, n=424) in septic shock: capillary refill time (CRT)-guided vs lactate-guided resuscitation.

28-day mortality: CRT 34.9% vs lactate 43.4% (RR 0.75, 95% CI 0.55-1.02; OR 0.61 in adjusted analysis; p=0.06 unadjusted but adjusted was significant)
Fluids given at 8 h: significantly LESS in CRT group (~1.8 L vs 2.3 L) — fewer harms of fluid overload
SOFA, vasopressor days, RRT use: favoured CRT strategy
CONCLUSION: A peripheral-perfusion (CRT, mottling, central-toe temperature) strategy may improve outcomes vs a lactate-chasing strategy in septic shock, in part by avoiding excessive fluid resuscitation. Practical: combine both — lactate for prognosis, perfusion for when to stop resuscitating.[17]

Jansen lactate clearance trial (2010, AJRCCM) — early lactate-guided therapy

Multicentre open-label RCT (n=348) of ventilated ICU patients with lactate ≥3.0 mmol/L: lactate-clearance-guided (target ≥20% per 2 h) vs no lactate guidance.

In-hospital mortality: lactate-guided 23% vs no guidance 33% (statistically significant in the subgroup with lactate ≥3.0 entry criterion — 20% vs 44% in severe shock subgroup)
Less organ failure in lactate-guided arm
CONCLUSION: Early lactate-guided therapy reduced mortality and organ failure in patients with hyperlactataemia. Lactate clearance ≥10-20%/h became the standard resuscitation target, integrated into the Surviving Sepsis Campaign 2021.[2][7]

MINS — the VISION cohort data (Devereaux, Curr Opin Cardiol 2014; AHA 2021 Scientific Statement)

Pooled cohort data from the international VISION (Vascular Events In Noncardiac Surgery Patients Cohort Evaluation) study (>30,000 patients across multiple countries):

Incidence of MINS: ~8% of all non-cardiac surgery; 20-30% in vascular surgery; higher with prior CAD, HF, stroke, emergency surgery
30-day mortality post-MINS: ~9% (3-4× higher than patients without MINS)
Independent association with cardiac arrest, cardiogenic shock, acute HF, stroke — even when asymptomatic
~85% of MINS cases have no ischaemic symptoms — detected by routine postoperative troponin screening (AHA recommends hs-cTn at 24-48 h post-op in high-risk patients)
AHA 2021 management: aspirin + statin, glycaemic control, intensive surveillance, cardiology follow-up; PCI reserved for high-risk trajectories (ischaemic features / dynamic ECG)
PREVENT-MINS (Szczeklik 2025, Circulation): ivabradine did NOT reduce MINS — supports the supply-demand (rate-independent) pathophysiology in most cases.[9][10][11]

Backes 2012 systematic review (Intensive Care Med) — suPAR as biomarker in systemic inflammation

Systematic review of suPAR in sepsis, septic shock, and systemic inflammation across 14 studies (>3,500 patients):

suPAR is an independent predictor of mortality in sepsis and septic shock, often superior to CRP, procalcitonin, IL-6 for prognosis
Higher suPAR → higher 28-day and 90-day mortality; cut-off ~12 ng/mL carries mortality >50%
Less useful for DIAGNOSIS of infection (PCT remains better for distinguishing bacterial from non-bacterial)
Confirmed by Gussen 2019 (J Intensive Care): neutrophils are the dominant source of circulating suPAR; suPAR marks exhausted neutrophil function — high suPAR = dysfunctional, migratory-impaired neutrophils → worse outcome
Status: not yet routine bedside in most ICUs, but a high-yield exam concept (immunoparalysis, neutrophil exhaustion).[13][14]

van Es 2017 IPD meta-analysis (J Thromb Haemost) — Wells rule + age-adjusted D-dimer for PE exclusion

Individual patient data meta-analysis (>7,000 patients with suspected PE) of the original and simplified Wells rules combined with age-adjusted D-dimer cut-off (age × 10 µg/L FEU in >50 yr):

3-month VTE failure rate after PE exclusion: <1.5% (well below the 3% safety threshold)
Age-adjusted cut-off safely increased the proportion of older patients in whom PE could be ruled out by D-dimer alone (without imaging) — from ~6% to ~30% in >50 yr
CONCLUSION: A negative D-dimer (using the age-adjusted threshold) combined with a low/intermediate Wells score safely excludes PE. In ICU, D-dimer is rarely rule-out useful (most critically ill have elevated D-dimer) — rely on imaging, except for low-risk young patients.[18]

Lee 2022 BMJ systematic review — NT-proBNP rule-in/rule-out pathway for acute heart failure

Meta-analysis (>40,000 patients with acute dyspnoea) validating NT-proBNP thresholds for acute heart failure:

Rule-out: NT-proBNP <300 ng/L → NPV ~98% across all ages (heart failure unlikely)
Rule-in (age-stratified): <50 yr >450 ng/L; 50-75 yr >900 ng/L; >75 yr >1800 ng/L → specificity ~90%
Confounders: renal failure raises (use higher thresholds or alternative diagnosis), obesity lowers (use 50% lower thresholds), AF raises independently
CONCLUSION: NT-proBNP is a powerful rule-out for acute heart failure in the dyspnoeic ICU patient; rule-in is age-stratified and confounder-aware. Trending NT-proBNP during decongestion predicts outcome.[3]

Special populations

Biomarker interpretation pitfalls by population

Population	Pitfall	Workaround
Renal failure (eGFR <30, dialysis)	Troponin and NT-proBNP chronically elevated	Use delta troponin (>20% rise) for acute injury; rely on echo, JVP, fluid balance
Obesity (BMI ≥30)	BNP/NT-proBNP falsely low	Use obesity cut-offs (50% lower); use POCUS B-lines / JVP / echo
Pregnancy	D-dimer physiologically elevated from 1st trimester; alkaline phosphatase ↑	Use YEARS algorithm; do NOT use plain D-dimer threshold
Cirrhosis / liver failure	Lactate clearance impaired; coagulopathy (PT, fibrinogen) altered; CRP often low (decreased hepatic synthesis)	Lactate may be mildly high without hypoperfusion; do not over-resuscitate
Cardiac surgery / post-op	Troponin, BNP, CRP all elevated at baseline	Use trends; post-CABG troponin rise is expected (surgical injury) — MINS definition (≥99th centile) still applies but threshold often higher
Malignancy	D-dimer chronically high; CRP high; suPAR high	D-dimer cannot rule out VTE — image; suPAR adds prognosis
Neutropenia / post-chemotherapy	PCT, CRP may be blunted; suPAR high	Do not rely on fever or PCT — start empiric antibiotics in any febrile neutropenic; serial CRP/PCT
On sacubitril/valsartan (ARNI)	BNP falsely high	Use NT-proBNP

[1]

Exam synthesis — CICM / FFICM / EDIC one-minute answer

The exam-friendly synthesis of ICU biomarkers

ICU biomarkers complement, not replace, clinical assessment.

Procalcitonin (PCT): bacterial infection marker → antibiotic stewardship (PRORATA: stop at PCT <0.5 or ≥80% fall). NEVER delay empiric antibiotics in sepsis.
Lactate: hypoperfusion/anaerobiosis → resuscitation target (clearance ≥10%/h, Jansen 2010); ≥2 mmol/L = Sepsis-3 organ dysfunction; ≥4 + vasopressors = septic shock. Combine with peripheral perfusion (ANDROMEDA-SHOCK 2019 — CRT-targeted may outperform lactate-chasing).
Troponin (hs-cTn): myocardial injury → distinguish type 1 MI (plaque rupture → PCI/DAPT) from type 2 MI (supply-demand → treat cause) from MINS (post-op, often silent, aspirin + statin; AHA 2021).
BNP/NT-proBNP: cardiac stretch → heart failure rule-out (<300 ng/L NT-proBNP ≈ rules out) and prognosis; renal failure raises, obesity lowers, ARNI raises BNP (use NT-proBNP).
CRP: non-specific inflammation → use trends, lags PCT by 24-48 h.
D-dimer: fibrin turnover → rules OUT VTE/DIC (high NPV with age-adjusted cut-off) but rules nothing in; ISTH overt DIC score uses magnitude of D-dimer rise.
suPAR: immune activation / neutrophil exhaustion → prognostic for mortality in sepsis, independent of severity scores; not yet routine bedside.
Copeptin (CT-proAVP): vasopressin surrogate → shock prognosis; high copeptin + refractory shock → consider exogenous vasopressin; also used for early MI rule-out with troponin.

[1]

One-glance biomarker-cause summary table

Marker up	Bacterial infection	Hypoperfusion	Cardiac	Inflammation	Thrombosis	Stress / prognosis
Procalcitonin	✓✓✓	—	—	✓ (surgery, trauma)	—	✓
Lactate	✓ (sepsis)	✓✓✓	✓ (shock)	—	—	✓✓
Troponin	—	✓ (shock)	✓✓✓	✓ (myocarditis)	✓ (type 1)	✓
BNP/NT-proBNP	—	—	✓✓✓	✓ (sepsis)	—	✓
CRP	✓	—	—	✓✓✓	—	—
D-dimer	—	—	—	✓ (DIC)	✓✓✓ (VTE/DIC)	—
suPAR	✓✓	—	—	✓✓	—	✓✓✓
Copeptin	—	✓ (shock)	✓ (stress)	—	—	✓✓✓

[1]

Exam practice

SAQ — Procalcitonin-guided antibiotic stewardship in severe community-acquired pneumonia

10 minutes · 10 marks

[1]

SAQ — Postoperative troponin elevation: distinguishing MINS, type 1 MI and type 2 MI

10 minutes · 10 marks

[1]

References

[1]Bouadma L, Luyt CE, Tubach F, et al. Use of procalcitonin to reduce patients' exposure to antibiotics in intensive care units (PRORATA trial): a multicentre randomised controlled trial Lancet, 2010.PMID 20097417
[2]Jansen TC, van Bommel J, Schoonderbeek FJ, et al. Early lactate-guided therapy in intensive care unit patients: a multicenter, open-label, randomized controlled trial Am J Respir Crit Care Med, 2010.PMID 20463176
[3]Lee KK, Doudesis D, Anwar M, et al. Development and validation of a decision support tool for the diagnosis of acute heart failure: systematic review, meta-analysis, and modelling study BMJ, 2022.PMID 35697365
[4]Chapman AR, Adamson PD, Shah ASV, et al. High-Sensitivity Cardiac Troponin and the Universal Definition of Myocardial Infarction Circulation, 2020.PMID 31587565
[5]Póvoa P. C-reactive protein: a valuable marker of sepsis Intensive Care Med, 2002.PMID 11904651
[6]Vincent JL, Bakker J. Blood lactate levels in sepsis: in 8 questions Curr Opin Crit Care, 2021.PMID 33852499
[7]Evans L, Rhodes A, Alhazzani W, et al. Executive Summary: Surviving Sepsis Campaign: International Guidelines for the Management of Sepsis and Septic Shock 2021 Crit Care Med, 2021.PMID 34643578
[8]Bularga A, Hung J, Daghem M, et al. Coronary Artery and Cardiac Disease in Patients With Type 2 Myocardial Infarction: A Prospective Cohort Study Circulation, 2022.PMID 35341327
[9]Ruetzler K, Smilowitz NR, Berger JS, et al. Diagnosis and Management of Patients With Myocardial Injury After Noncardiac Surgery: A Scientific Statement From the American Heart Association Circulation, 2021.PMID 34601955
[10]Khan J, Alonso-Coello P, Devereaux PJ. Myocardial injury after noncardiac surgery Curr Opin Cardiol, 2014.PMID 25029449
[11]Ekeloef S, Alamili M, Devereaux PJ, et al. Troponin elevations after non-cardiac, non-vascular surgery are predictive of major adverse cardiac events and mortality: a systematic review and meta-analysis Br J Anaesth, 2016.PMID 27799170
[12]Levi M, Toh CH, Thachil J, et al. Guidelines for the diagnosis and management of disseminated intravascular coagulation. British Committee for Standards in Haematology Br J Haematol, 2009.PMID 19222477
[13]Backes Y, van der Sluijs KF, Mackie DP, et al. Usefulness of suPAR as a biological marker in patients with systemic inflammation or infection: a systematic review Intensive Care Med, 2012.PMID 22706919
[14]Gussen H, Hohlstein P, Bartneck M, et al. Neutrophils are a main source of circulating suPAR predicting outcome in critical illness J Intensive Care, 2019.PMID 31061709
[15]Lee JH, Chan YH, Lai OF, et al. Vasopressin and copeptin levels in children with sepsis and septic shock Intensive Care Med, 2013.PMID 23344831
[16]Dabla PK, Dabla V, Arora S. Co-peptin: Role as a novel biomarker in clinical practice Clin Chim Acta, 2011.PMID 20920496
[17]Hernández G, Ospina-Tascón GA, Damiani LP, et al. Effect of a Resuscitation Strategy Targeting Peripheral Perfusion Status vs Serum Lactate Levels on 28-Day Mortality Among Patients With Septic Shock: The ANDROMEDA-SHOCK Randomized Clinical Trial JAMA, 2019.PMID 30772908
[18]van Es N, Kraaijpoel N, Klok FA, et al. The original and simplified Wells rules and age-adjusted D-dimer testing to rule out pulmonary embolism: an individual patient data meta-analysis J Thromb Haemost, 2017.PMID 28106338

Your progress

Clinical pearls

Red flags

Prognosis

PRORATA trial (Bouadma 2010, Lancet) — procalcitonin-guided antibiotic therapy

Lactate — deep dive (perfusion, clearance, prognostics)

Troponin in ICU — type 1 MI, type 2 MI, and MINS

BNP / NT-proBNP — heart failure diagnosis, volume status, prognosis

D-dimer — fibrin turnover in VTE, DIC, arterial thrombosis

suPAR — soluble urokinase-type plasminogen activator receptor

Copeptin — vasopressin surrogate

Procalcitonin — antibiotic stewardship algorithm

Combining biomarkers — multimarker panels improve prediction

Clinical pearls — second set (deep dive)

Red flags — extended set

Trial cards — the evidence base

ANDROMEDA-SHOCK (Hernández 2019, JAMA) — peripheral perfusion vs lactate-targeted resuscitation

Jansen lactate clearance trial (2010, AJRCCM) — early lactate-guided therapy

MINS — the VISION cohort data (Devereaux, Curr Opin Cardiol 2014; AHA 2021 Scientific Statement)

Backes 2012 systematic review (Intensive Care Med) — suPAR as biomarker in systemic inflammation

van Es 2017 IPD meta-analysis (J Thromb Haemost) — Wells rule + age-adjusted D-dimer for PE exclusion

Lee 2022 BMJ systematic review — NT-proBNP rule-in/rule-out pathway for acute heart failure

Special populations

Exam synthesis — CICM / FFICM / EDIC one-minute answer

Exam practice

SAQ — Procalcitonin-guided antibiotic stewardship in severe community-acquired pneumonia

SAQ — Postoperative troponin elevation: distinguishing MINS, type 1 MI and type 2 MI

References

Your progress

Clinical pearls

Red flags

Prognosis

PRORATA trial (Bouadma 2010, Lancet) — procalcitonin-guided antibiotic therapy

Lactate — deep dive (perfusion, clearance, prognostics)

Troponin in ICU — type 1 MI, type 2 MI, and MINS

BNP / NT-proBNP — heart failure diagnosis, volume status, prognosis

D-dimer — fibrin turnover in VTE, DIC, arterial thrombosis

suPAR — soluble urokinase-type plasminogen activator receptor

Copeptin — vasopressin surrogate

Procalcitonin — antibiotic stewardship algorithm

Combining biomarkers — multimarker panels improve prediction

Clinical pearls — second set (deep dive)

Red flags — extended set

Trial cards — the evidence base

ANDROMEDA-SHOCK (Hernández 2019, JAMA) — peripheral perfusion vs lactate-targeted resuscitation

Jansen lactate clearance trial (2010, AJRCCM) — early lactate-guided therapy

MINS — the VISION cohort data (Devereaux, Curr Opin Cardiol 2014; AHA 2021 Scientific Statement)

Backes 2012 systematic review (Intensive Care Med) — suPAR as biomarker in systemic inflammation

van Es 2017 IPD meta-analysis (J Thromb Haemost) — Wells rule + age-adjusted D-dimer for PE exclusion

Lee 2022 BMJ systematic review — NT-proBNP rule-in/rule-out pathway for acute heart failure

Special populations

Exam synthesis — CICM / FFICM / EDIC one-minute answer

Exam practice

SAQ — Procalcitonin-guided antibiotic stewardship in severe community-acquired pneumonia

SAQ — Postoperative troponin elevation: distinguishing MINS, type 1 MI and type 2 MI

References