What Are the Guidelines for Working Up a Suspected PJI?

    Dr. Bryan Springer answers questions from ICJR about evaluating a total knee arthroplasty patient who may have a deep periprosthetic joint infection, in part 1 of a 2-part series. Part 2 will be published tomorrow.


    ICJR: How common are deep periprosthetic joint infections (PJI) in total knee arthroplasty (TKA) patients, and what can surgeons do to prevent them?

    Bryan D. Springer, MD: Infection following TKA remains one of the most dreaded and difficult complications to treat. The overall incidence of infection in the literature ranges from 0.5% to 2% for primary TKAs and from 2% to 4% for revision TKAs.

    In 2005, 16.8% of all revision TKAs were done because of infection. It is estimated that by the year 2030, 65% of all revision procedures will be performed because of infection.

    Due to the severity and difficulty in treating a patient with an infection, prevention is the key. Several modifiable factors can play a role in diminishing risk, including:

    • Systemic or host factors, such as optimizing weight (BMI less than 40), encouraging smoking cessation, optimizing diabetes control (HGB A1c less than 7), reversing malnutrition, and utilizing methicillin-resistant Staphyloccocus aureus decolonization protocols
    • Environmental factors, such as controlling the operating room environment, making appropriate antibiotic choices, and reducing nosocomial infection
    • Local factors, such as ensuring the skin is clean and using appropriate closure techniques and surgical dressings

    ICJR: How do you work up the patient who presents with a suspected infection?

    Dr. Springer: Every patient who presents with a failed or otherwise painful joint replacement must be evaluated for the presence of a deep PJI. Even if the cause of failure seems obvious, concomitant infection could be present, which would substantially change the treatment plan. If possible, the diagnosis should be made preoperatively for appropriate patient counseling on treatment options.

    In 2010, the American Academy of Orthopaedic Surgeons (AAOS) approved and released its Clinical Practice Guideline (CPG) on The Diagnosis of PerIprosthetic Joint Infections of The Hip and Knee. [1] This CPG provides and algorithmic approach to evaluating a patient with a suspected PJI and also evaluates the utility of each test.

    Relevant recommendations in the CPG are outlined below.

    AAOS CPG Recommendation 1

    Patients should be risk-stratified as “higher” or “lower” probability of developing a PJI based on answers to the questions about the history.

    Strength: Consensus (opinion)

    The patient’s history – including past medical history and the history surrounding the index arthroplasty – is critical in the evalution, as it may lead to a high index of suspicion for infection as the cause of failure or pain.

    Patient risk factors for infection include:

    • Diabetes mellitus
    • History of septic arthritis of the native joint
    • Inflammatory arthritis
    • Skin disorders (eg, psoriasis)
    • Prior ipsilateral joint surgery
    • Malnourished/obesity (they can co-exist!)
    • Renal insufficiency –
    • Immunocompromise

    The key considerations in the patient’s history include the following:

    • Were there wound healing problems at the time of the index arthroplasty?
      • Were antibiotics given for an extended period postoperatively?
      • Was there an extended hospital stay?
      • Was there an early return to the OR?
    • Has there been pain since the index procedure (different than preoperative pain)?
    • Has there been a history of a recent systemic illness or other bacteremia?
      • Suggests a hematogenous infection
      • Specifically ask about dental work or dental problems
    • Is the patient’s pain associated with weight-bearing (mechanical) versus rest?
    • Is there early implant loosening or osteolysis (especially if it has been fewer than 5 years since surgery)?

    AAOS CPG Recommendation 2

    Obtain an erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level to assess for PJI in all patients.

    Strength: Strong

    Patients who have a higher probability of developing a PJI warrant a more vigorous evaluation. Laboratory tests are the easiest and most cost-effective way to screen for infection, which is supported by multiple studies from multiple centers.

    The ESR and CRP are very rarely normal in the face of infection. If both are normal, infection is unlikely. If one is normal and the other is abnormal, further testing is warranted to rule infection in or out.

    AAOS CPG Recommendation 3

    If the ESR and CRP are abnormal, aspirate the joint and send the fluid for culture, synovial fluid white blood cell (WBC) count, and differential.

    Strength: Strong


    • Ensure the patient has been off antibiotics for at least 2 weeks (Recommendation 8; Strength: Moderate); improves the accuracy of the culture results
    • Send for aerobic and anaerobic cultures
    • Consider fungal and acid-fast bacilli testing in patients with prior negative aspirates
    • Repeat aspiration if clinical suspicion of an infection is high but initial cultures are negative
    • Consider decanting into blood culture bottles to increase culture yield, especially if prior aspiration was negative

    Synovial Fluid WBC Count

    • Inexpensive, ubiquitous, and objective; has been shown to be arguably the best single test for PJI
    • Can be done pre- or intraoperatively, with results in 30 minutes or so
    • Optimal cut-off value is controversial; most data suggest about 3000 WBC/microliter

    Synovial Fluid WBC Count Differential

    • Has been shown to be a very good test
    • Optimal cut-off value of 60% to 80%
    • A value greater than 90% is very suspicious for PJI

    A slightly less-aggressive approach is recommended for the hip (Recommendation 4; Strength: Strong) given the increased morbidity of a hip aspiration and a high rate of false-positive cultures if done routinely.

    AAOS CPG Recommendation 9

    Nulcear imaging is an option for the diagnosis of PJI.

    Strength: Weak

    Plain x-rays are rarely abnormal in patients with PJI. However, periosteal reaction/early lytic lesions are very suspicious and should be assumed to be associated with infection and not bearing surface wear if seen within the first few years postoperatively.

    Nuclear medicine studies are expensive, and their value in routine use is unclear. They can be used, however, as a secondary test for patients for whom revision is not indicated or for whom synovial fluid aspiration is not available (most commonly seen in patients with a stiff TKA).

    The best nuclear medicine test to diagnose PJI is the indium111 + sulfur colloid marrow scan.  The sulfur colloid scan corrects for marrow packing. Negative tests seem reliable in ruling out infection

    AAOS CPG Recommendation 11

    The use of gram stains to diagnose PJI is not recommended, as they are not reliable.

    Strength: Strong

    Gram stain may be helpful in cases in which gross purulence is identified to guide empiric treatment. However, this cannot be the only test used to evaluate a patient for infection because it is not sensitive enough.

    In fact, gram stains can be falsely positive. Crystal formation in Gram’s dye can look like gram-positive cocci, and bacteria can actually grow in the reagents.

    AAOS CPG Recommendation 12

    Intraoperative frozen sections can be helpful in diagnosing PJI when infection has been neither ruled in nor ruled out.

    Strength: Strong

    Intraoperative frozen sections can be reliable, but the reliability depends on an interested and dedicated pathologist. Surgeons should request the same pathologist for interpreting all their cases. Surgeons should also meet with the pathologist to decide on the criteria for infection. This will ensure both parties are “speaking the same language.”

    Criteria for infection are controversial:

    • Average of 10 polymorphonuclear leukocytes (PMNs) in the 5 most cellular fields (Lonner criteria)
    • Use of an average of 5 or 10 PMNs does not seem to make a difference in testing utility
    • PMNs in fibrin do not count

    AAOS CPG Recommendation 13

    Obtain multiple cultures at the time of revision surgery.

    Strength: Strong

    Obtaining multiple cultures greatly enhances the likelihood of receiving true positive culture results and differentiating falsely positive cultures. I recommend an odd number – 5 is generally adequate. Tissue and fluid cultures are preferred over swabs.

    The surgeon should request aerobic and anaerobic cultures. In addition, acid-fast bacilli and fungal cultures can be done if prior preoperative cultures have been negative and the synovial fluid WBC count and/or differential are suggestive of infection.

    AAOS CPG Recommendation 14

    If there is a high degree of suspicion for infection, hold antibiotics until the joint has been aspirated. 

    Strength: Strong

    This is quite possibly the most important recommendation in the CPG: The use of antibiotics prior to culturing the joint ruins the likelihood of correctly making a diagnosis of PJI and identifying the infecting pathogen – which may delay appropriate care.

    Share this recommendation with physicians in physical medicine and rehabilitation, internal medicine, and the emergency department, as well as any other physicians and healthcare providers who assist you in caring for joint replacement patients.

    AAOS CPG Recommendation 15

    Prophylactic antibiotics should not be withheld prior to routine revisions if the suspicion for infection is low or if the diagnosis of PJI has already been made (in other words, most of the time).

    Strength: Moderate

    At the time this CPG was written, only 1 study was available to support this recommendation. A second Level I study has since been completed. If the guideline were to be examined again, this would receive a strong reccomendation.

    Areas for Future Research

    Alpha Defensin (Synovasure)

    Alpha defensin is a biomarker that can be measured in synovial fluid, and it is presently available as a commercial test combined with the measurement of synovial fluid CRP.

    Alpha defensin testing has been shown to have excellent sensitivity and specificity. [2] It appears to normalize following treatment, making it useful for predicting infection eradication as part of a 2-stage exchange. It even appears to be accurate in the face of antibiotic treatment.

    I presently use alpha defensin testing in my practice to confirm the presence of infection in cases that appear to be culture negative.

    Leukocyte Esterase Reagent Strips (Urinalysis Strips)

     Leukocyte esterase is present in activated PMNs and is found in infected bodily fluids. The pad of the reagent strip contains a detergent that lyses PMNs, releases leukocyte esterase, and combines with a salt to create a purple color.

    Leukocyte esterase reagent strips are commonly used to diagnose urinary tract infection, chorioamniotis, and peritonitis. Their use in PJI was first described by Parvizi at the AAOS meeting in 2010. [3] Subsequent research by Wetters et al [4] showed excellent sensitivity, making this a good rule-out test.          

    In summary, about 90% of infections can be diagnosed via the ESR/CRP and aspiration. As the CPG notes: [1]

    • Obtain an ESR and CRP prior to all revisions and in any cases in which PJI is suspected or the patient has a failed or painful hip or knee arthroplasty.
    • If these labs are abnormal and PJI is suspected, aspirate the joint. Send the aspirate for culture, synovial WBC count, and differential.
    • Withhold antibiotics until the patient is diagnosed.


    1. The Diagnosis of PerIprosthetic Joint Infections of The Hip and Knee: Guideline and Evidence Report, published by the American Academy of Orthopaedic Surgeons, © Available at https://www.aaos.org/research/guidelines/PJIguideline.pdf; accessed August 17, 2017.
    2. Deirmengian C, Kardos K, Kilmartin P, Cameron A, Schiller K, Parvizi J. Combined measurement of synovial fluid α-Defensin and C-reactive protein levels: highly accurate for diagnosing periprosthetic joint infection. J Bone Joint Surg Am 2014 Sep 3;96(17):1439-45. doi: 10.2106/JBJS.M.01316.
    3. Parvizi J, Azzam K, Jacovides C,Antoci V, Ghanem E. Diagnosis of periprosthetic joint infection: the role of a simple, yet unrecognized enzyme. Presented at the annual meeting of the Academy of Orthopaedic Surgeons AAOS, New Orleans, Louisiana, March 10-13, 2010.
    4. Wetters NG, Berend KR, Lombardi AV, Morris MJ, Tucker TL, Della Valle CJ. Leukocyte esterase reagent strips for the rapid diagnosis of periprosthetic joint infection. J Arthroplasty. 2012 Sep;27(8 Suppl):8-11. doi: 10.1016/j.arth.2012.03.037. Epub 2012 May 17.

    About the Expert

    Bryan D. Springer, MD, is Fellowship Director at OrthoCarolina Hip and Knee Center, Charlotte, North Carolina.


    Dr. Springer has no disclosures relevant to this article.