Inflammatory markers: CRP, ESR, procalcitonin and ferritin

Inflammatory markers are blood tests that show how active inflammation is in the body. None of them diagnoses a single disease on its own; each measures a consequence of inflammation, and the skill is to read them together, with the clinical picture and the trend over time. This page covers the five that medical students meet most often: C-reactive protein (CRP), the erythrocyte sedimentation rate (ESR), procalcitonin (PCT), plasma viscosity (PV) and ferritin.

By the end you should be able to:

1. Say what each marker measures and name the cytokines or proteins that drive it. ^[1]
2. Quote the typical reference limits, the method dependency of every one of them, and the basic CRP kinetics. ^[1][2][3]
3. Explain why plasma viscosity is more stable than ESR. ^[11]
4. Interpret procalcitonin thresholds and apply the Surviving Sepsis Campaign recommendations on when to use and not to use it. ^[6]
5. Recognise the named patterns (high CRP with normal ESR, very high ESR, ESR greater than 50 mm/h with giant cell arteritis or polymyalgia rheumatica, an ESR that falls in response to treatment). ^[1][4][7][9]
6. Work through three sample result sheets and reach a one-sentence clinical conclusion with the next step for each.
7. Spot the pitfalls: non-inflammatory causes, method mixing, renal failure, obesity and the effect of glucocorticoids. ^[2][8]

Inflammation changes the proteins in the blood within hours to days. The acute-phase response is the name for that change: cytokines released by macrophages and injured tissue, above all interleukin-6, signal the liver to make proteins such as CRP, fibrinogen and serum amyloid A, while albumin falls. ^[1] Different markers in this response move at different speeds and for different reasons, which is the whole point of ordering more than one.

CRP is the fastest and most sensitive of these markers. It is a pentameric protein made by hepatocytes, named because it binds the C-polysaccharide of Streptococcus pneumoniae. ^[1][2] ESR is much older and less specific: it is not a protein at all, but the speed at which red cells settle in a column of anticoagulated blood in one hour, driven by the plasma proteins that make them clump. ^[1][4] Procalcitonin is the marker used to separate bacterial sepsis from viral or non-infectious inflammation when deciding whether antibiotics will help. ^[5][6] Ferritin is the intracellular iron-storage protein that also behaves as an acute-phase reactant, so a normal or raised ferritin does not exclude iron deficiency in an inflamed patient. ^[10] Plasma viscosity is an alternative to ESR that depends on plasma proteins and is not thrown off by anaemia or polycythaemia. ^[11]

The lesson to carry through the whole page is that an inflammatory marker is a thermometer, not a diagnosis. ^[1]

CRP is an acute-phase protein with a very wide dynamic range. In healthy young adult volunteer blood donors the median concentration is 0.8 mg/L, the 90th centile is 3.0 mg/L and the 99th centile is 10 mg/L. ^[2] After a stimulus, plasma CRP can climb from less than 50 microg/L to more than 500 mg/L, a 10,000-fold rise. ^[2] It is produced only by hepatocytes, mainly under transcriptional control by interleukin-6. ^[2]

Kinetics are what make CRP useful at the bedside. Following the onset of the inflammatory stimulus, CRP levels begin to rise within a few hours and peak within 48 hours. Depending on the type and chronicity of the stimulus or treatment, levels may fall rapidly or remain elevated. ^[2] After myocardial infarction the peak at around 48 hours predicts outcome. ^[2] CRP has a short half-life, so when the inflammation stops the value falls by roughly half a day. It shows no diurnal variation and is unaffected by eating. ^[2], so a single result can be compared directly with another.

CRP is non-specific in the best sense: it reflects tissue damage and inflammation without distinguishing infection from sterile inflammation. ^[1][2] High-sensitivity and conventional CRP assays are the same analyte measured to different limits; high-sensitivity simply means the assay can measure CRP at low concentrations, below 3 mg/L. ^[2]

ESR is the height of plasma in millimetres left at the top of a column of anticoagulated blood after one hour of standing vertically. ^[1][4] Red cells settle faster when they aggregate into rouleaux; aggregation depends on plasma proteins, especially fibrinogen. ^[1] ESR therefore lags behind CRP: it reflects a change in plasma protein composition over days rather than hours, and it stays up long after the stimulus has settled.

Three major factors affect erythrocyte aggregation: the surface free energy of the red cells, their surface charge and the dielectric constant of the suspending plasma. ^[1] Anything that changes the plasma proteins or the red cells changes the ESR:

• Anaemia and polycythaemia influence the ESR, which is why plasma viscosity is preferred where that matters, and red-cell shape and plasma proteins also change the result. ^[11]
• Alterations in the shape of red cells can prevent aggregation and produce a normal ESR despite inflammation. ^[1]
• Plasma viscosity is generally considered to be superior to ESR, being unaffected by age, sex, anaemia or polycythaemia. ^[11]

Because ESR and CRP obey different physics, they move differently in the same illness. ESR and CRP exhibit distinct postoperative kinetics after arthroplasty: ESR remains elevated for a longer duration, while CRP rises rapidly and declines more quickly toward baseline levels. ^[12]

Procalcitonin is a precursor peptide of the hormone calcitonin. Its circulating concentration generally rises during bacterial infections, particularly sepsis, and it responds less to viral infections and sterile inflammation than CRP. ^[5][6] It is detectable at very low concentrations in health but rises rapidly during bacterial infection. ^[6]

Its real value is antibiotic stewardship. Evidence supports the use of procalcitonin mainly for guiding antibiotic discontinuation, rather than for starting antibiotics. ^[5] The Surviving Sepsis Campaign 2021 recommends against using procalcitonin plus clinical evaluation to decide when to start antimicrobials in adults with suspected sepsis or septic shock, and recommends using procalcitonin together with clinical evaluation to decide when to discontinue antimicrobials when the optimal duration is unclear, provided source control is adequate. ^[6] NICE, in its earlier diagnostic guidance, concluded that there was insufficient evidence to recommend routine adoption of procalcitonin testing for sepsis. ^[7] Newer trial evidence (ADAPT-Sepsis) has reopened that question.

Ferritin is an intracellular iron-storage protein, and it also behaves as an acute-phase reactant: in anemia of chronic disease, ferritin is normal or raised because inflammation drives iron sequestration, while it is low in true iron deficiency. ^[10] A normal ferritin therefore does not exclude iron deficiency in an inflamed patient, and a raised ferritin is a non-specific sign of inflammation, infection, malignancy or liver disease. ^[10]

Plasma viscosity is determined by the proteins in the plasma and, like ESR, rises slowly and stays up through chronic inflammation. Unlike ESR, it is not accelerated by anaemia or retarded by polycythaemia, and it is considered superior to ESR in that respect. ^[11] It is also independent of the red cell shape and of tube geometry, which makes it more reproducible. ^[11]

Every reference interval is method-, laboratory- and population-dependent. ^[1][2][4] The intervals below come from the cited sources; the interval printed on the report that your patient actually had is the one to use, and serial results should be read on the same platform and from the same laboratory. ^[1][2][11]

Plasma viscosity is commonly reported in mPa s, and a reference interval is required from the laboratory doing the assay; no single published cut-off exists that transfers between analysers. ^[11] The ferritin interval is assay-dependent and also depends on sex and, in many laboratories, on menopausal status. ^[10]

A practical note on interpretation: a marker is read over time, not once. ^[1] CRP changes within hours, ESR and plasma viscosity over days. A rising CRP is more informative than a single high value, and a falling value shows that treatment is working. ^[1][2]

Severe bacterial infection is usually the most potent stimulator of CRP, and many rheumatic diseases raise CRP during active disease. ^[1] In systemic lupus erythematosus, which is usually not associated with high CRP, a marked CRP rise should push you to look for infection or another complication rather than the lupus itself. ^[1][2] ESR usually rises in acute infections and then falls with recovery; in chronic infection and inflammation it stays up. ^[1]

A high ESR with a normal CRP is not unusual when the sedimentation rate is driven by immunoglobulins or paraproteins rather than by the interleukin-6 driven acute-phase response; polymyalgia rheumatica and giant cell arteritis can occasionally present this way. ^[1][4][9]

Named patterns turn a pair of numbers into a working hypothesis.

• CRP under 1 mg/dL (about 10 mg/L) with a normal ESR: no current acute-phase response. Does not exclude disease: early infection may not yet have raised either. ^[1]
• CRP rising within hours, ESR still normal: early acute inflammation. CRP begins to rise within a few hours and peaks within 48 hours, while ESR lags behind it. ^[1][12]
• Both CRP and ESR raised: ongoing inflammation of any cause. CRP usually tracks disease activity more closely than ESR in this situation. ^[2]
• CRP settles before the ESR: a known pattern in recovery and after surgery. CRP rises rapidly and declines more quickly toward baseline; ESR remains elevated for a longer duration. ^[12]
• ESR of 50 mm/h or more: a 1990 ACR research criterion for giant cell arteritis, and such high values are typical in giant cell arteritis and polymyalgia rheumatica, but normal inflammatory markers do not exclude them. ^[9] Giant cell arteritis can very occasionally occur with entirely normal inflammatory markers, so do not let a normal ESR close the diagnosis when the clinical picture is strong. ^[9]
• CRP raised and ESR under 50 mm/h in a patient with vascular or visual symptoms: consider giant cell arteritis none the less; CRP may be the more sensitive flag. ^[9]
• High ESR with a normal or only mildly raised CRP: think of paraproteinaemia, monoclonal gammopathy, pregnancy-related plasma protein changes and some connective tissue diseases, where plasma proteins drive sedimentation rather than acute-phase protein switching. ^[4][11]
• Procalcitonin under 0.1 ng/mL in a sick patient: a serious bacterial infection is unlikely, but not excluded; do not withhold treatment on the number alone. ^[5][6]
• Procalcitonin 0.25 to 0.5 ng/mL: an indeterminate band; the number must be read with imaging, microbiology and severity, not as a diagnosis. ^[5]
• Procalcitonin 0.5 ng/mL or more together with a consistent clinical picture: supports treating as systemic bacterial infection. ^[5]
• Ferritin normal or raised in an anaemic, inflamed patient: iron deficiency may still be present; add a transferrin receptor test or a bone-marrow iron stain to decide. ^[10]
• Ferritin below about 15 to 30 microg/L in a patient who is not inflamed: carries more weight for iron deficiency than the same number in someone with raised CRP or ESR.
• Plasma viscosity high while both ESR and CRP behave unexpectedly: paraproteinaemia and chronic inflammation are the usual explanations. ^[11]
• Procalcitonin normal after antibiotics have already started: procalcitonin falls with successful treatment; a falling trend supports stopping early. ^[5][6]

The practical algorithm is simple and is the one the major guidelines support. ^[6]

Procalcitonin concentrations are not all-or-nothing: cut-offs such as 0.25 and 0.5 ng/mL are operationally defined trial values, not diagnostic absolutes. ^[5] Serial measurements and the trend matter more than a single value, particularly as justification for stopping rather than starting antibiotics. ^[5][6]

The values below are invented for teaching and are labelled as illustrative; they are not real patient data, and they are not a diagnosis. Every range, threshold and rule used in the interpretation is taken from the cited sources.

1. CRP of 0.6 mg/L sits below the median (0.8 mg/L) and well below the 90th centile of 3.0 mg/L for healthy donors. ^[2]
2. ESR of 11 mm/h is within the Westergren upper limit of 15 mm/h for a young man. ^[1]
3. Ferritin and plasma viscosity are both inside their laboratory's own intervals.
4. There is no evidence of an acute-phase response here, so no further investigation for inflammation is warranted.

Conclusion: this is a normal inflammatory marker panel in an asymptomatic adult; the next step is routine care and retesting only if symptoms develop.

1. CRP of 48 mg/L is well above the MedlinePlus usual normal figure of about 10 mg/L, so an active inflammatory process is present. ^[3]
2. ESR of 96 mm/h far exceeds the female young-adult upper limit of 20 mm/h and is in the very high range that polymyalgia rheumatica or giant cell arteritis are nearly always associated with. ^[1][9]
3. Procalcitonin is low, arguing against acute bacterial sepsis and making sepsis a less likely explanation for the pyrexia than giant cell arteritis. ^[5][6]
4. Ferritin is raised as an acute-phase reactant, which is expected. ^[10]

Conclusion: this pattern supports an active medium-vessel or large-vessel vasculitis such as giant cell arteritis, so the next step is immediate high-dose glucocorticoid treatment directed by a rheumatologist, with temporal artery biopsy arranged without delaying therapy when vision is threatened.

1. CRP started high and is now falling: this is the expected early peak-and-fall pattern, and 48 hours is the usual time to peak for CRP. ^[1][2]
2. ESR is barely different because it lags and remains elevated for a longer duration. ^[1][12]
3. Procalcitonin is falling from a clearly raised level, which supports continued treatment and is the kind of trend that justifies stopping antibiotics early when the patient is clinically ready. ^[5][6]
4. Clinical assessment (observations, source control, imaging) still governs discharge, not the marker. ^[6]

Conclusion: the inflammatory response is resolving on treatment, so the next step is continued antibiotics with daily reassessment, de-escalating when the clinical picture and the falling procalcitonin agree, rather than a fixed 7-day course.

• Treating a CRP as if it were a diagnosis. CRP reflects tissue damage and inflammation; a high value does not tell you what is wrong, where it is, or whether it is infection. ^[1][2]
• Reading one result in isolation. Acute injury is a rise and fall, not a number. A single CRP can be made by an early infection or an old inflammation. Serial samples read on the same assay are the only safe way to judge. ^[1][2][12]
• Mixing methods or laboratories. ESR and CRP behave differently and must not be compared to each other; nor may serial results from two different assays or laboratories be trended. ^[1][2][11]
• Ignoring anaemia and polycythaemia. Both accelerate or slow the ESR, so an ESR in an anaemic patient may be less impressive than it looks and a normal ESR in a polycythaemic one can be falsely reassuring. Plasma viscosity is unaffected by these. ^[1][11]
• Forgetting red-cell shape. Sickle cells, target cells and echinocytes may prevent aggregation and give a normal ESR in active inflammation. ^[1]
• Trusting a normal ESR. Giant cell arteritis has been documented with normal inflammatory markers, and a patient with compatible symptoms and signs needs treatment while the investigation proceeds. ^[9]
• Confusing high-sensitivity and conventional CRP. They are the same analyte on different measuring ranges. Quoting a cardiovascular-risk cut-off from a high-sensitivity report for an acutely infected patient, or vice versa, is a category error. ^[2]
• Overinterpreting procalcitonin. The cut-offs are trial operational values, not diagnostic absolutes. A low level does not exclude localised or early infection and must not be used to withhold antibiotics in a sick patient. ^[5][6]
• Forgetting glucocorticoids. Prednisolone and other glucocorticoids suppress CRP and ESR, so in polymyalgia rheumatica a relapse can occur with only mildly raised markers. Do not equate normal markers with remission once treatment has started. ^[8]
• Ignoring IL-6 blockade. Tocilizumab and similar agents lower CRP specifically because they block the pathway that produces it. A normal CRP on therapy does not mean the disease is inactive. ^[8]
• Reading ferritin as iron. Ferritin is an acute-phase reactant, so a raised or even normal ferritin does not exclude iron deficiency during active inflammation. ^[10]
• Using obesity- and contraception-driven CRP as inflammation. A raised baseline CRP in an obese or oral-contraceptive user is a real association, not acute inflammation. Do not launch an infection work-up on the number alone. ^[2]
• ESR normal limits by population. Apply the age- and sex-specific reference intervals printed by the laboratory. ^[4]

• CRP is the fastest and most responsive marker of the acute-phase response; it is made by hepatocytes under interleukin-6 control and has no diurnal variation. ^[1][2]
• ESR is a slow, non-specific measurement of red-cell sedimentation driven by plasma proteins, especially fibrinogen; it lags behind CRP and stays up longer. ^[1]
• CRP begins to rise within a few hours of the inflammatory stimulus and peaks within 48 hours; the size and speed of the rise, read against the laboratory's own interval, matter more than any single cut-off. ^[2]
• Plasma viscosity measures the same biology more cleanly than ESR and is unaffected by anaemia and polycythaemia, which is why it is often the preferred test where it is available. ^[11]
• Giant cell arteritis and polymyalgia rheumatica often show high ESR, and an ESR of 50 mm/h or more is a 1990 ACR research criterion for giant cell arteritis, but normal markers do not exclude either condition. ^[9]
• Procalcitonin is a better marker for systemic bacterial sepsis than CRP or ESR, but it should be used mainly to decide when to stop antibiotics, not whether to start them. ^[5][6]
• Ferritin is both an iron-storage protein and an acute-phase reactant, so its value must be read with the CRP and ESR, and preferably with a transferrin-receptor test, in an inflamed patient. ^[10]
• Inflammatory markers are never a diagnosis: read them serially, on the same assay and laboratory, and always alongside the clinical picture. ^[1][2]