Synovial fluid analysis: crystals, cells and culture

• Synovial fluid is the lubricant and nutrient medium of a joint. A sample is taken by sterile aspiration (arthrocentesis) and tested for colour, clarity, viscosity, cell count, differential, Gram stain, culture and crystals.
• Normal fluid is thick, straw-coloured and clear, with under 200 white cells per microlitre and under 25 per cent neutrophils. Blood, pus, crystals or a high cell count turn it abnormal. ^{[2] [1]}
• Crystal identification decides crystal arthropathy: negatively birefringent needle-shaped crystals are monosodium urate (gout); weakly positively birefringent rhomboid crystals are calcium pyrophosphate (pseudogout). ^[2][7]
• White cell thresholds move a fluid along the non-inflammatory, inflammatory, septic and haemorrhagic scale, but no single count proves or excludes infection; a count under 50,000 cells per microlitre does not rule out septic arthritis. ^[4]
• Aspirate before antibiotics where it can be done, and always send Gram stain and aerobic plus anaerobic culture; a negative Gram stain is common in infection and must not reassure. ^[4][5]
• The British Society for Rheumatology (BSR), the American College of Rheumatology (ACR) and the European League Against Rheumatism (EULAR) all treat synovial fluid analysis as the anchor investigation for the hot swollen joint and for crystal disease.

By the end of this page you should be able to:

1. Describe normal synovial fluid and explain why each test on the result sheet is run.
2. Read a synovial fluid white cell count and differential and place the fluid in the non-inflammatory, inflammatory, septic or haemorrhagic category.
3. Identify monosodium urate and calcium pyrophosphate crystals by shape and birefringence and state what each means.
4. Explain the useful and the unreliable biochemical lines on the report, including glucose, and the meaning of a low synovial glucose.
5. Describe the pitfalls of a negative Gram stain and state the manoeuvres that improve culture yield.
6. Decide when a joint must be aspirated and how crystals and sepsis can coexist.
7. Interpret a short synovial fluid result sheet step by step, ending with a one-sentence conclusion and the next step.

Synovial fluid analysis is a group of tests on fluid taken from a joint by needle. The tests are cheap, quick and usually decisive, which is why a hot swollen joint is first aspirated before anything else is ordered. ^[1][5]

The fluid itself is made by the synovial lining and behaves like a dialysate of plasma with added hyaluronan. In health it is clear, straw coloured and very viscous, present in tiny amounts, and it nourishes cartilage while it lubricates the joint. ^[1] Its colour, clarity, viscosity, cell count, differential, Gram film, culture and crystal content therefore mirror the processes happening inside the joint, from mechanical wear to crystal deposition to bacterial infection. ^[2]

This page is about the result sheet rather than the physiology behind each disease. It names the reference intervals published by the sources, states the thresholds that sort fluid into categories, works through three complete (invented, illustrative) result sheets, lists the traps, and closes with key points, a flowchart and check-yourself questions. Regional practice notes are given where a guideline or low-resource setting matters. ^[4]

Each line of the report answers a different question, and mixing the lines up is the commonest beginner error. Read all of them before deciding anything. ^[1]

• Appearance (colour and clarity). Normal fluid is straw coloured, clear or slightly yellow and transparent. Cloudy, opaque or purulent fluid contains cells, fibrin or pus. Blood makes it red or pink; xanthochromia (a yellow supernatant after centrifugation) means older bleeding. ^[1][2]
• Viscosity. Normal fluid is highly viscous because of hyaluronan. In inflammatory arthritis the enzymes of neutrophils break hyaluronan down, so the fluid becomes thin and watery. The traditional (non-standardised) bedside "string sign" (a thread stretched between two fingers) is a crude proxy for this and should not be relied on alone. ^[1][2]
• White cell count. The total leucocyte count per microlitre is the number that sorts fluid into categories. It rises with inflammation and infection and falls with haemorrhage relative to blood. ^[2]
• Differential (percentage of polymorphonuclear cells, PMN, usually neutrophils). This tells you which cells are driving the inflammation. Non-inflammatory effusions keep a PMN percentage under 25 per cent; inflammatory and crystal effusions push it above 50 per cent; septic effusions push it above 75 per cent and often above 90 per cent. ^{[2][4] [3]}
• Gram stain. A rapid, cheap, highly specific test that identifies the shape and Gram reaction of organisms in the fluid. Its sensitivity is variable and can be as low as 26 per cent in some series. ^[4]
• Culture (aerobic and anaerobic). The definitive microbiological test. Yield improves when fluid is also inoculated directly into blood culture bottles. ^[4]
• Crystal examination under polarised light. Compensated polarised light microscopy is the gold standard for detecting crystals and has been since 1961. It separates monosodium urate crystals (negatively birefringent, needle or rod shaped, 1 to 20 micrometres long) from calcium pyrophosphate crystals (weakly positively birefringent, rhomboid or short rod shaped). ^[2][7]
• Glucose. Synovial glucose tracks serum glucose, with normally little difference. A low value is suggestive of bacterial infection but synovial glucose and protein do not reliably distinguish septic arthritis from non-infectious arthritis and are no longer recommended routinely. ^[4]
• Total protein, lactate and LDH. Sometimes added. Synovial lactate 10 mmol/L or more has been proposed as suggestive of infection, but its interpretation depends on assay methodology and isoform reporting, and its evidence base is limited. ^[4]

The first four lines (appearance, viscosity, count, differential) create the pattern. Gram stain and culture decide infection. Crystals decide crystal disease. Glucose is an adjunct that is increasingly not ordered. ^[4]

Table 1 uses the intervals and thresholds that the cited sources state for adults. They are teaching values: synovial fluid cells, crystals and cultures are interpreted against the laboratory's own method and the clinical picture, and no single threshold is definitively diagnostic. ^[2][4]

Three details sit underneath this table.

Thresholds are not diagnoses. The native joint review states that no single leukocyte threshold is definitively diagnostic, that values must be interpreted within clinical context, and that intermediate results (10,000 to 50,000 cells/microlitre with 50 to 90 per cent PMN) are a grey zone where infection and inflammation cannot be excluded. ^[4]

Antibiotics, and prior treatment in general, shift the numbers. Synovial fluid taken after antibiotics can show lower leukocyte counts and lower PMN percentages. The same review suggests that in that situation lower thresholds (around 16,000 cells/microlitre or more) may improve sensitivity, and that a count below 50,000 does not exclude infection even in the best circumstances. ^[4]

A negative Gram stain is common. Gram stain is highly specific but its sensitivity is variable and can be as low as 26 per cent. Both Gram stain sensitivity and culture positivity fall after prior antimicrobial exposure. ^[4]

Table 2 collects the common directional associations of the synovial fluid tests. It is not exhaustive: a single abnormal line never makes a diagnosis. ^[1][2][4]

The two cleanest rules in this table matter most.

Glucose is unreliable here. The native joint review states that synovial glucose and protein do not reliably distinguish native joint septic arthritis from non-infectious arthritis and should not be routinely ordered. ^[4] When a low synovial glucose is still reported, interpret it against the serum glucose taken at the same time: a fall is suggestive of infection, but false reassurance and false alarm are both common. ^[4]

Crystals and sepsis can coexist. Acute crystal arthritis and bacterial infection look alike and the tables overlap. The presence of crystals confirms crystal disease but does not exclude septic arthritis, and retrospective series report septic arthritis in more than 5 per cent of aspirates that are positive for crystals. ^[4] For that reason crystal analysis should always accompany cell count, Gram stain and culture, never replace them. ^[4]

Once the six parts of the report are read together, the combinations are recognisable. These named patterns are the bridge between the piece of paper and the diagnosis. ^[2][4]

Normal (non-inflammatory) fluid. Clear or straw, high viscosity, under 200 cells/microlitre, under 25 per cent PMN, negative Gram stain and culture, no crystals. This is the fluid of a resting joint, of osteoarthritis, or of a resolved process. It tells you there is no active inflammatory or septic process, but it cannot be reassuring when the clinical picture says otherwise. ^{[2] [1]}

Non-inflammatory effusion (osteoarthritis or meniscal, ligament or cartilage injury). Clear or straw, still viscous, 200 to 2,000 cells/microlitre, PMN under 25 per cent. These are the effusions of wear, tear and structural injury. The count is only modestly raised because the process is mechanical rather than cellular. ^[2]

Inflammatory fluid. Cloudy or yellow, low viscosity, 2,000 or more cells/microlitre (often 2,000 to 100,000), more than 50 per cent PMN, negative Gram stain and culture, no crystals. This is the fluid of rheumatoid arthritis, psoriatic arthritis, reactive arthritis and many of the seronegative spondyloarthritides. It can be hard to distinguish from crystal or septic fluid on the count alone, which is why Gram stain and crystals are indispensable. ^[2]

Crystal fluid. Cloudy or yellow, 2,000 to 50,000 cells/microlitre, more than 50 per cent PMN, negative Gram stain and culture, crystals present. Needle-shaped, negatively birefringent crystals mean gout; rhomboid, weakly positively birefringent crystals mean calcium pyrophosphate deposition. ^[2][7]

Gout confirmed by fluid alone. The ACR/EULAR 2015 gout classification criteria, and the current state-of-the-art review of polarised light microscopy, both say that identification of monosodium urate crystals in a symptomatic joint or bursa is a sufficient criterion for classifying gout. ^[6][7]

Pseudogout confirmed by fluid alone. In the same way, identification of calcium pyrophosphate crystals from a joint that is swollen, tender or painful is a sufficient criterion for classifying disease due to calcium pyrophosphate deposition. ^[7]

Septic fluid. Purulent or cloudy yellow, usually over 50,000 cells/microlitre, 90 per cent or more PMN in high probability cases, Gram stain sometimes positive, culture positive in only 60 to 80 per cent of septic cases on table series and often less in recent series. The native joint review divides patients into high probability (purulent fluid, over 50,000 cells/microlitre, over 90 per cent PMN), intermediate probability (10,000 to 50,000 cells/microlitre with 50 to 90 per cent PMN) and a region where neither infection nor inflammation can be excluded. ^[2][4]

Haemorrhagic fluid. Bloody or pink, viscosity variable, variable cell count, PMN in the 50 to 75 per cent range on the published table, Gram stain and culture negative, no crystals. Think trauma, a bleeding tendency, pigmented villonodular synovitis, neuropathic (Charcot) joint or recent surgery. ^[2]

Gonococcal arthritis. The table gives 34,000 to 68,000 cells/microlitre and PMN above 75 per cent, but gonococci are fastidious, Gram stain is often negative in fewer than half of cases, and nucleic acid amplification testing of the fluid and of urogenital and extragenital sites is frequently needed. ^[2][4]

Lyme arthritis. The table gives 3,000 to 100,000 cells/microlitre with more than 50 per cent PMN. In an endemic area with exposure, test the fluid and blood as guided by the clinical picture. ^[2][4]

Mixed or indeterminate fluid. When the count sits in the intermediate band, or when antibiotics have already been given, the probability of infection sits in a grey zone. The response is not to guess but to serialise: repeat the aspiration, take blood cultures, use imaging, involve orthopaedics, and in gonococcal or culture-negative disease use molecular testing. ^[4][5]

The three sheets below are invented for teaching, are labelled as such on every line, and are not real patient data. They use the intervals and thresholds in Table 1. Each sheet is followed by a numbered step-by-step interpretation and a one-sentence conclusion with the next step. ^[2][4]

1. Appearance, colour and viscosity are all normal, so this is not an acute inflammatory or septic effusion. ^[1]
2. White cell count 120 cells/microlitre and 15 per cent PMN sit inside the normal interval, so there is no active cellular inflammation. ^[2]
3. Gram stain and culture are negative, which is the expected finding in a non-infective joint. ^[2]
4. No crystals are seen, so acute crystal arthropathy is not supported. ^[2][7]

One-sentence conclusion and next step: this is a normal non-inflammatory effusion, and no further synovial fluid investigation is needed unless the clinical picture fails to fit.

1. Cloudy yellow fluid of low viscosity with 18,000 cells/microlitre and 75 per cent PMN is inflammatory, not haemorrhagic and not a simple effusion. ^[2][4]
2. The count sits in the crystal band of 2,000 to 50,000 cells/microlitre and the percentage above 50 confirms an inflammatory drive. ^[2]
3. No organisms are seen on Gram stain and culture is negative, so sepsis is not supported, but a negative Gram stain cannot exclude it. ^[4]
4. Needle-shaped, negatively birefringent crystals are monosodium urate, which is a sufficient criterion for gout in a symptomatic joint. ^[6][7]
5. Culture-negative septic arthritis still occurs in culture-negative native joint septic arthritis, and a patient with crystals can still have a concurrent infection, with retrospective series reporting sepsis in more than 5 per cent of crystal-positive aspirates. ^[4]

One-sentence conclusion and next step: this sheet is classic for gout, and the next step is to treat the flare as gout while still arranging blood cultures and re-aspiration if fever, rigors or purulent fluid point to a concurrent septic arthritis.

1. Purulent fluid, a count of 75,000 cells/microlitre and 92 per cent PMN put this in the high-probability septic band. ^[4]
2. The conjugate pair of a count above 50,000 with 90 per cent or more PMN is the classic pattern and raises the post-test probability of infection sharply, although even here it is not 100 per cent. ^[4]
3. Gram stain is positive in this illustrative sheet, but in real life it is as sensitive as its lowest reported value of 26 per cent, so a negative film would not have reassured us. ^[4]
4. The illustrative culture result grew an organism, which is the one finding that settles the diagnosis; a real negative culture does not exclude sepsis, especially after antibiotics. ^[4]
5. No crystals are seen, but crystals and infection can coexist, so their absence does not rescue a very high pre-test probability. ^[4]

One-sentence conclusion and next step: high-probability native joint septic arthritis, and the next step is blood cultures, joint imaging, urgent orthopaedic review, source control and intravenous antibiotics without further delay.

These are the mistakes that turn a good test into a wrong decision. ^[4][5]

1. Aspirating after antibiotics. Whenever a septic joint is suspected, obtain synovial fluid before antimicrobial therapy if it can be done safely; antibiotics lower the cell count, the PMN percentage, and the yield of Gram stain and culture. ^[4][5]
2. Trusting a negative Gram stain. Gram stain is highly specific, but its sensitivity is variable and can be as low as 26 per cent. Prior antibiotics reduce both Gram stain sensitivity and culture positivity. ^[4]
3. Missing atypical organisms. Gonococci are difficult to recover by conventional culture, and neutrophil-poor or fastidious infections such as those caused by Kingella kingae in children, Brucella, mycobacteria or Borrelia may need nucleic acid amplification, specialist media or longer incubation. ^[3][4]
4. Thinking crystals exclude sepsis. The presence of crystals confirms crystal disease but does not exclude infection; sepsis is reported in more than 5 per cent of crystal-positive aspirates. ^[4]
5. Treating crystal counts as absolute. The inflammatory and septic bands overlap. A count of 30,000 cells/microlitre with 80 per cent PMN is an intermediate-probability result, not a verdict, and needs the clinical picture, imaging and repeat aspiration. ^[4]
6. Using glucose as a discriminator. Synovial glucose and protein do not reliably distinguish septic from non-septic arthritis and are not recommended routinely. ^[4]
7. Dry tap panic. If no fluid is obtained, septic arthritis is not excluded. Repeat image-guided arthrocentesis, and keep the diagnostic process moving. ^[4]
8. Post-aspiration care. Apply cold for 24 to 36 hours after the procedure and advise a return to activity based on the underlying diagnosis. ^[1]
9. Sample handling. Inoculate fluid into aerobic and anaerobic blood culture bottles to improve yield, and avoid swabs, which have inferior diagnostic sensitivity. ^[4]
10. Missing the indications. A joint that is hot and swollen, or any large effusion without a clear cause, deserves aspiration; removal of fluid can also relieve symptoms. ^[1][5]

When to aspirate (the short version): whenever infection, crystals or an unexplained effusion are in the differential; before antibiotics in suspected sepsis if it can be done safely; under imaging guidance when the joint is deep (hip, shoulder) or the volume is small. ^[1][4][5]

• Synovial fluid analysis is the anchor investigation for the hot swollen joint, for crystal disease and for unexplained effusions, and it is usually performed before imaging or treatment decisions. ^[1][5]
• Normal fluid is straw coloured, clear and viscous, with under 200 cells/microlitre and under 25 per cent neutrophils. ^[2]
• Non-inflammatory fluid (for example osteoarthritis) runs at 200 to 2,000 cells/microlitre with PMN under 25 per cent. ^[2]
• Inflammatory fluid runs at 2,000 to 100,000 cells/microlitre with PMN above 50 per cent. ^[2]
• Crystal fluid (gout or pseudogout) runs at 2,000 to 50,000 cells/microlitre with PMN above 50 per cent. ^[2]
• High-probability septic fluid has over 50,000 cells/microlitre and 90 per cent or more PMN; intermediate (grey zone) fluid has 10,000 to 50,000 cells/microlitre with 50 to 90 per cent PMN. ^[4]
• Needle-shaped, negatively birefringent crystals are monosodium urate (gout); rhomboid, weakly positively birefringent crystals are calcium pyrophosphate (pseudogout). ^[2][7]
• The identification of monosodium urate crystals in a symptomatic joint or bursa is sufficient for classifying gout; calcium pyrophosphate crystals in a swollen, tender or painful joint are sufficient for classifying calcium pyrophosphate deposition disease. ^[6][7]
• Gram stain is highly specific but its sensitivity can be as low as 26 per cent; a negative stain never excludes sepsis. ^[4]
• Sterile technique, direct inoculation into blood culture bottles and the avoidance of swabs improve yield; prior antibiotics reduce both Gram stain and culture. ^[4]
• Glucose and protein do not reliably separate septic from non-septic arthritis and are not recommended routinely; lactate 10 mmol/L or more has been proposed as suggestive of infection but is not stand-alone. ^[4]
• Crystals and infection can coexist; sepsis is reported in more than 5 per cent of crystal-positive aspirates. ^[4]
• Aspirate before antibiotics in suspected sepsis if it is safe to do so, and involve orthopaedics, infectious diseases and rheumatology early. ^[4][5]
• No single threshold is definitively diagnostic; always read counts, differential, Gram stain, culture and crystals alongside the whole patient. ^[4]