Procalcitonin: The Peptide That Detects Sepsis

Peptide Biomarkers in Clinical Medicine

<0.05 ng/mL normal

Procalcitonin, a 116-amino acid peptide, rises from undetectable levels to over 100 ng/mL within hours of bacterial sepsis, making it one of the fastest-responding biomarkers in critical care.

Schuetz et al., The Lancet Infectious Diseases, 2018

Schuetz et al., The Lancet Infectious Diseases, 2018

View as image

Sepsis kills roughly 11 million people each year, and early detection remains one of the biggest challenges in critical care medicine. Procalcitonin (PCT), a 116-amino acid peptide that is normally undetectable in healthy blood, surges within hours of bacterial infection, reaching levels thousands of times higher than baseline. This response makes PCT the most widely used peptide biomarker for sepsis diagnosis and antibiotic stewardship. This article covers what procalcitonin is, how accurately it detects sepsis, and what the evidence says about using it to guide treatment decisions. For broader context on how peptide biomarkers are used across medicine, see our guide to peptide biomarkers in clinical medicine.

What Is Procalcitonin? The Peptide Behind the Test

Procalcitonin is a 116-amino acid peptide, the precursor to the hormone calcitonin. In healthy individuals, it is produced exclusively by thyroid C-cells and cleaved into calcitonin before entering the bloodstream. Normal serum procalcitonin levels are below 0.05 ng/mL, often undetectable.^[1]

During bacterial infection, something unusual happens: virtually every tissue in the body begins producing procalcitonin. Liver, lung, kidney, adipose tissue, and intestinal cells all upregulate the CALCA gene (which encodes procalcitonin) in response to bacterial endotoxins and pro-inflammatory cytokines like IL-6 and TNF-alpha. This extrathyroidal production skips the normal cleavage step, so intact procalcitonin floods the bloodstream.

The kinetics are what make PCT clinically useful:

• Onset: PCT begins rising 2-4 hours after bacterial exposure
• Peak: Levels reach their maximum at 6-12 hours
• Half-life: Approximately 24-30 hours
• Decline: With appropriate antibiotic therapy, levels drop by roughly 50% per day

This timeline contrasts with C-reactive protein (CRP), which peaks at 36-50 hours and has a longer half-life, making CRP slower both to rise and to reflect treatment response.

The molecular structure of procalcitonin matters for understanding why it behaves this way. The full 116-amino acid chain consists of three regions: an N-terminal 57-amino acid segment (aminoprocalcitonin), a 32-amino acid calcitonin segment in the middle, and a 21-amino acid C-terminal segment (katacalcin). In the thyroid, enzymatic processing cleaves the chain into these components. During sepsis, the inflammatory cascade triggers transcription of the CALCA gene through a different pathway (NF-kB and cytokine signaling rather than calcium-mediated signaling), and the processing enzymes that would normally cleave the peptide are absent in non-thyroidal tissues. The result: intact procalcitonin accumulates in the blood in proportion to the bacterial burden.

How Accurately Does Procalcitonin Detect Sepsis?

The diagnostic performance of procalcitonin has been evaluated in dozens of meta-analyses spanning thousands of patients across emergency departments and intensive care units.

Emergency department patients

A systematic review and meta-analysis by Kondo et al. (2019) pooling data from 19 studies found that PCT for sepsis diagnosis in ED patients had a pooled sensitivity of 0.77, specificity of 0.79, and an area under the summary receiver operating characteristic curve (SROC AUC) of 0.85. The commonly used cutoff of 0.5 ng/mL balanced sensitivity and specificity, though optimal thresholds varied by patient population.

Critically ill ICU patients

In ICU settings, diagnostic accuracy is somewhat lower because critically ill patients often have elevated inflammatory markers from non-infectious causes (surgery, trauma, burns). A meta-analysis by Wacker et al. found pooled sensitivity of 0.77 (95% CI 0.72-0.81) and specificity of 0.79 (95% CI 0.74-0.84) across 3,244 critically ill patients from 30 studies.

What procalcitonin is good at

PCT performs best as a rule-out test. A low procalcitonin level (below 0.25 ng/mL) has a negative predictive value exceeding 90% for bacterial sepsis in most clinical contexts, meaning it reliably identifies patients who do NOT have bacterial infection. This is particularly valuable in the emergency department, where distinguishing bacterial from viral illness drives antibiotic prescribing decisions.

What procalcitonin is not good at

PCT cannot distinguish between different types of bacterial infection (Gram-positive vs. Gram-negative, localized vs. systemic). It can be falsely elevated after major surgery, in severe burns, in newborns during the first 48 hours of life, in medullary thyroid carcinoma, and in some autoimmune conditions. Fungal and viral infections generally produce lower PCT elevations, but overlap exists.^[1]

PCT-Guided Antibiotic Stewardship: The Strongest Evidence

The most impactful clinical application of procalcitonin is not diagnosis alone but guiding antibiotic duration. Antibiotic overuse drives resistance, increases side effects, and raises costs. PCT-guided algorithms use serial measurements to determine when antibiotics can be safely stopped.

The patient-level meta-analysis

Wirz et al. (2018) published a patient-level meta-analysis in Critical Care that pooled individual data from 26 randomized controlled trials. PCT-guided antibiotic therapy in ICU patients with infection and sepsis resulted in:^[2]

• Reduced mortality: Risk ratio 0.89 (95% CI 0.83-0.97)
• Shorter antibiotic duration: 9.3 days vs. 10.4 days (1.19-day reduction)
• No increase in treatment failure or recurrent infection rates

Respiratory infections

Schuetz et al. (2018) conducted a patient-level meta-analysis in The Lancet Infectious Diseases across 26 trials involving 6,708 patients with acute respiratory infections. PCT-guided therapy reduced antibiotic exposure by 2.4 days, reduced antibiotic-related side effects (adjusted odds ratio 0.68, 95% CI 0.57-0.82), and improved 30-day survival (adjusted odds ratio for mortality: 0.83, 95% CI 0.70-0.99).

Cancer patients

A more recent patient-level meta-analysis by Schuetz et al. (2024) in BMC Cancer examined PCT-guided therapy specifically in cancer patients. Among patients with malignancies, PCT-guided antibiotic discontinuation did not increase mortality and reduced antibiotic exposure, though the benefit was smaller than in the general population.

Where PCT-guided stewardship is less clear

Not all evidence is positive. A meta-analysis by Pepper et al. (2019) in JAMA found that when restricted to studies including only critically ill sepsis patients (not broader infection), PCT-guided antibiotic discontinuation did not reach statistical significance for mortality reduction. Similarly, in nine non-industry-sponsored trials, the mortality benefit was attenuated. This suggests the mortality benefit may be most robust in lower-acuity infections where early antibiotic de-escalation has the clearest risk-benefit ratio.

Procalcitonin vs. Other Peptide Biomarkers in Sepsis

PCT is not the only peptide with diagnostic value in sepsis. Several other peptide-derived biomarkers offer complementary information.

MR-proADM (mid-regional pro-adrenomedullin)

Adrenomedullin is a 52-amino acid peptide with vasodilatory and anti-inflammatory properties that rises during sepsis. Because adrenomedullin itself is rapidly cleared, clinicians measure its more stable precursor fragment, MR-proADM. Bernal-Morell et al. (2018) found MR-proADM had an AUC of 0.82 for predicting 28-day mortality in sepsis patients admitted to the emergency department, outperforming both procalcitonin and lactate for prognostication.^[3]

A narrative review by Onal et al. (2018) concluded that MR-proADM was particularly useful for predicting organ failure progression and ICU readmission, filling a gap that procalcitonin does not cover well: PCT diagnoses infection, while MR-proADM predicts organ dysfunction.^[4]

Presepsin (sCD14-ST)

Presepsin is a 13-kDa fragment of the monocyte receptor CD14, released when monocytes encounter bacterial components. A meta-analysis by Kondo et al. (2019) found presepsin had a sensitivity of 0.84 and specificity of 0.73 for sepsis diagnosis, comparable to PCT. Presepsin rises faster than PCT (within 2 hours) but is less well validated for guiding antibiotic duration.

LL-37 and cathelicidins

The antimicrobial peptide LL-37, the only human cathelicidin, plays a direct role in sepsis defense. Mankowska et al. (2025) reviewed LL-37's significance in sepsis pathogenesis, finding that LL-37 serum levels correlate with sepsis severity and may serve as both a biomarker and a therapeutic target.^[5] Ho et al. (2020) demonstrated that cathelicidin preserves intestinal barrier function during polymicrobial sepsis in mice, providing a mechanistic link between antimicrobial peptide levels and sepsis outcomes.^[6]

For more on how peptide biomarkers are used in heart failure diagnosis, a parallel application of peptide-based testing, see our article on BNP and NT-proBNP. For how peptide biomarkers work in acute conditions beyond sepsis, see our article on copeptin as a vasopressin surrogate biomarker.

Procalcitonin's Limitations in Context

PCT has genuine clinical utility, but several factors limit its usefulness.

It is a bacterial marker, not a sepsis-specific marker. Any significant bacterial infection raises PCT, whether or not sepsis criteria are met. Localized infections like pneumonia, pyelonephritis, and meningitis all elevate PCT, sometimes substantially. This means a high PCT level tells you bacteria are involved; it does not by itself confirm the systemic dysregulated host response that defines sepsis.

False positives are common in certain populations. Post-surgical patients often have PCT elevations for 1-3 days without infection. Neonates have physiologically elevated PCT in the first 48-72 hours of life. Patients with chronic kidney disease may have chronically elevated PCT due to reduced clearance. Medullary thyroid carcinoma and small-cell lung cancer produce PCT independent of infection.^[7]

Serial measurements outperform single values. A single PCT measurement at admission is less informative than the trend over 24-48 hours. A rising PCT with clinical deterioration strengthens the case for bacterial infection; a falling PCT supports antibiotic de-escalation. Most guidelines recommend serial PCT testing rather than single cutoff-based decisions.

Cost-effectiveness data is mixed. In high-income ICU settings, PCT-guided stewardship appears cost-effective through reduced antibiotic use and shorter ICU stays. In resource-limited settings, the test cost (typically $20-50 per measurement) may not be justified when clinical assessment alone can guide most antibiotic decisions.

Combination approaches may be superior. Emerging evidence suggests combining PCT with other peptide biomarkers (MR-proADM for prognosis, presepsin for early detection) or with clinical scores (qSOFA, SOFA) improves both diagnostic accuracy and prognostic prediction beyond what any single marker achieves alone.

Clinical Practice: How Procalcitonin Is Used Today

PCT is FDA-cleared for two indications: risk assessment for progression to severe sepsis and septic shock in critically ill patients (2017), and guiding antibiotic therapy in lower respiratory tract infections (2017). The Surviving Sepsis Campaign guidelines include PCT as a tool for antibiotic de-escalation, recommending that daily PCT measurement can support decisions to stop antibiotics when clinical improvement occurs alongside declining PCT levels.

In practice, most hospitals use PCT as follows:

• Admission: Baseline PCT at presentation with suspected infection
• Threshold for bacterial infection: PCT above 0.25-0.5 ng/mL supports antibiotic initiation
• Serial monitoring: Repeat PCT every 24-48 hours
• De-escalation trigger: PCT decline of 80% from peak or below 0.25-0.5 ng/mL supports stopping antibiotics
• Context: Always interpreted alongside clinical status, imaging, and cultures

The 2024 GRADE-based procalcitonin guideline for emergency departments formalized this approach, recommending PCT-guided algorithms as part of structured antibiotic stewardship programs.

Procalcitonin in specific clinical scenarios

Pneumonia: PCT is most useful for distinguishing bacterial from viral pneumonia. Community-acquired pneumonia with PCT below 0.1 ng/mL is unlikely to be bacterial, while PCT above 0.25 ng/mL supports antibiotic initiation. Serial PCT measurements guide duration, with most guidelines recommending antibiotics can be stopped when PCT drops below 0.25 ng/mL or declines by 80% from peak.

Post-surgical patients: Routine PCT monitoring after major surgery is not recommended because the inflammatory response to surgical tissue injury elevates PCT for 1-3 days regardless of infection. However, a PCT that continues to rise beyond post-operative day 3, or re-rises after an initial decline, raises concern for surgical site infection or nosocomial sepsis.

Neonatal sepsis: PCT testing in neonates requires adjusted reference ranges because healthy newborns have a physiological PCT surge peaking at 24-48 hours of life (levels up to 2.0 ng/mL can be normal). After 72 hours, adult reference ranges apply. Despite these limitations, PCT remains one of the best-performing biomarkers for neonatal sepsis, outperforming CRP in the first 24 hours of infection.

For context on how other peptide-based diagnostics are reshaping clinical medicine, see our articles on C-peptide as a diabetes marker and amyloid-beta as an Alzheimer's biomarker.

The Bottom Line

Procalcitonin is a 116-amino acid peptide that rises rapidly during bacterial infection and falls predictably with effective treatment, giving it a dual role as both a diagnostic and therapy-guiding biomarker. The strongest evidence supports PCT-guided antibiotic stewardship, where serial measurements help clinicians stop antibiotics earlier without increasing treatment failure, with meta-analyses showing reduced mortality (RR 0.89) and shorter antibiotic courses. PCT performs best as a rule-out test for bacterial infection rather than a rule-in test for sepsis, and its accuracy improves when combined with other peptide biomarkers like MR-proADM and clinical scoring systems.

Frequently Asked Questions

What is procalcitonin and what does it indicate?

Procalcitonin is a 116-amino acid peptide, the precursor to the hormone calcitonin. In healthy individuals, blood levels are below 0.05 ng/mL. During bacterial infection, tissues throughout the body produce procalcitonin, causing levels to spike within 6-12 hours. Elevated PCT indicates bacterial infection, with higher levels generally correlating with greater infection severity. It is FDA-cleared for sepsis risk assessment and guiding antibiotic therapy.

How accurate is procalcitonin for diagnosing sepsis?

Meta-analyses show procalcitonin has a pooled sensitivity of approximately 0.77 and specificity of 0.79 for sepsis diagnosis, with an overall AUC of 0.85. PCT is more accurate for ruling out bacterial infection (negative predictive value above 90% at the 0.25 ng/mL cutoff) than for ruling it in, because non-infectious conditions like surgery, trauma, and certain cancers can also elevate PCT levels.

Does procalcitonin-guided therapy reduce antibiotic use?

Yes. A patient-level meta-analysis of 26 randomized trials (Wirz et al., 2018) found PCT-guided antibiotic therapy reduced treatment duration by 1.19 days in ICU patients. In respiratory infections (Schuetz et al., 2018), PCT guidance reduced antibiotic exposure by 2.4 days and decreased antibiotic-related side effects (odds ratio 0.68). These reductions occurred without increasing treatment failure or recurrent infection rates.

What procalcitonin level indicates sepsis?

There is no single cutoff that confirms sepsis. General guidance uses 0.25 ng/mL as the threshold below which bacterial infection is unlikely, and 0.5 ng/mL as the level above which bacterial infection is probable. Levels above 2.0 ng/mL suggest severe bacterial infection or sepsis, and levels above 10 ng/mL strongly indicate septic shock. Serial trends over 24-48 hours are more informative than any single measurement.

How is procalcitonin different from CRP for sepsis?

Procalcitonin rises faster (2-4 hours vs. 12-24 hours for CRP), is more specific to bacterial infection (CRP rises with any inflammation including viral illness and autoimmune disease), and falls more predictably with treatment (50% decline per day). CRP is cheaper and more widely available, but PCT provides more actionable information for antibiotic stewardship decisions. Many clinicians use both.

Can procalcitonin be elevated without infection?

Yes. Procalcitonin can be falsely elevated after major surgery (for 1-3 days), in severe trauma or burns, in neonates during the first 48-72 hours of life, in chronic kidney disease (reduced clearance), in medullary thyroid carcinoma, and in small-cell lung cancer. Some fungal infections and parasitic diseases produce moderate PCT elevations. Clinical context and serial measurements help distinguish infectious from non-infectious causes.