Does this patient have bacterial arthritis?

Bacterial arthritis is typically divided into gonococcal and non-gonococcal arthritis (NGA). Non-gonococcal arthritis is of greatest concern, because it is the most damaging; therefore, early diagnosis and treatment is critical. The duration of untreated infection is the most important determinant of joint damage. An infected joint may be destroyed within a few days.

Staphylococcus aureus is the most common organism. This has not changed for decades, however, methicillin-resistant Staphylococcus aureus (MRSA) is clearly increasing in frequency. MRSA septic arthritis is more common in patients who are older, have a greater number of medical co-morbid conditions, and have been hospitalized in the past six months.

Non-gonococcal septic arthritis typically presents as a single red, hot and swollen joint. The knee is most commonly affected (45%), followed by the hip (15%), ankle (9%), elbow (8%), wrist (6%), and shoulder (5%). Septic arthritis of the interphalangeal and metacarpophalangeal joints of the hands are usually the result of fist fights (“fight bites”) or cat bites. Other cardinal signs of infection may be present. The onset is relatively rapid.

Continue Reading

Other joints may be involved, and polyarticular septic arthritis may occur. Patients with polyarticular septic arthritis are often bacteremic, and have a higher risk of death. Involvement of the sternoclavicular, acromioclavicular, and sacroiliac joints is more common in intravenous drug users, for unclear reasons. Historically, septic arthritis in narcotic addicts commonly involved Pseudomonas and other Gram-negative bacteria, due to an epidemic of pentazocine abuse in the 1970s. Pentazocine dissolves in water at room temperature, facilitating contamination with environmental bacteria such as Pseudomonas. Gram-negative infections have become much less common in injection drug users with a decline in pentazocine availability. Gram-negative septic arthritis today is occasionally seen in patients who are older, immunocompromised, or have underlying disease of the urinary or gastrointestinal tracts.

Gonococcal septic arthritis usually occurs as part of the syndrome of disseminated gonococcal infection, in which patients have some combination of urethritis, tenosynovitis, septic arthritis, and a hemorrhagic and pustular rash. Patients with this syndrome are usually younger and have generally good outcomes.

History and physical may lead to the source of the infection. Skin infections are a common cause of Staphylococcus aureus bacterial arthritis. In adults, hematogenous spread is the most common route of infection. Previously injured or inflamed joints are more commonly infected. Transient or persistent bacteremia underlies the importance of blood cultures as part of the evaluation.

Differential diagnoses

The most common differential diagnosis for acute inflammatory monoarticular arthritis is crystalline arthritis. Gout and pseudogout commonly affect the knee, ankles and small joints of foot. The presentation can be identical to that of a NGA septic arthritis. The history and physical and labs may provide clues for differentiation. However, non-gonococcal septic arthritis and crystalline arthritis are not mutually exclusive. Therefore, septic arthritis should be excluded in any case of suspected crystalline arthropathy.

Other conditions may also cause monoarthritis, but usually the joint is not as inflamed. One can be lulled into thinking that a joint is not infected with a damaging organism, based upon this presentation. Likewise, one may attribute the subacute presentation to a patient’s underlying disease (such as rheumatoid arthritis). However, it is important to recall that previously injured or inflamed joints are at higher risk of becoming infected, and uncommon presentations of septic arthritis occur.

The other differential diagnoses for nongonococcal septic arthritis include Lyme arthritis (particularly in endemic areas), viral arthritis, and other bacterial arthritis (including gonococcal arthritis). In tuberculous and fungal arthritis, the affected joint is usually cooler and does not present acutely. Vigilance is necessary to detect this form of infectious arthritis and should be suspected in chronic and particularly destructive forms of arthritis unresponsive to other anti-inflammatory approaches.

Culture- and crystal-negative acute inflammatory arthritis may comprise an atypical presentation of an autoimmune arthritis, such as RA or SLE; alternatively, a spondyloarthropathy, such as reactive, psoriatic, enteropathic arthritis, or ankylosing spondylitis, may present this way. Acute rheumatic fever or post-streptococcal arthritis may also produce a culture- and crystal-negative inflamed joint. True bacterial arthritis may sometimes be culture-negative, either due to administration of antibiotics prior to sampling, an inadequate amount of synovial fluid, delays in processing of the sample, or the presence of fastidious bacteria that do not grow well on conventional culture media.

What tests to perform?

Laboratory testing
Synovial fluid culture

The most critical laboratory test for the diagnosis of a NG septic arthritis is the synovial fluid culture. If you get only a drop of fluid, send it for this test. If you have another drop, perform a Gram stain. The remaining fluid can be sent for crystal analysis and cell counts.

Cell counts of greater than 50,000 WBC/mm3 should raise suspicion for NG septic arthritis, although counts of this magnitude can be seen with crystalline arthritis. Moreover, lower cell counts are not helpful in excluding septic arthritis. Hence, culturing the synovial fluid is the most important test.

Gram stain

Gram stains may help provide guidance for empiric antibiotic therapy if positive. Although a negative Gram stain, in a setting of high suspicion for septic arthritis, does not preclude the initiation of empiric antibiotic therapy. Likewise, the identification of crystals does not exclude infection, and empiric antibiotic therapy may be appropriate, based upon the clinical circumstances.

False positive Gram stains may be reported by inexperienced microscopists, as mucin and other debris may be mistaken for Gram-positive cocci. Typically, these pseudo-cocci are much larger than bacterial organisms, when viewed at the same optical power.

Blood cultures

It is important to draw blood cultures and to culture any area found on the PE that may have placed the patient at risk for hematogenous spread. Genital, anal and pharyngeal cultures are needed to rule out gonococcal arthritis in patients in whom that differential diagnosis is being considered. Urine gene probes for GC should be performed in suspected gonococcal arthritis, as Neisseria gonorrhoeae may be difficult to grow in culture.


CBC, ESR, CRP are useful, but not always helpful, and may erroneously lead you away from the diagnosis of NG septic arthritis. They should always be interpreted in the context of the clinical context.


X-rays will give you a sense of structural damage and perhaps urgency for intervention. They may also alert you to the more unusual (adult) occurrence of joint infection from a contiguous structure, such as osteomyelitis. Serial x-rays may be needed to evaluate structural integrity over time. Structural damage is suggested by extensive erosions and loss of joint space. In advanced settings the articulations may be indistinguishable. Cortical or periosteal irregularities may alert you to underlying bone infection. However, adults are more likely to develop primary septic arthritis and secondary osteomyelitis.


MRI may provide important information about the possibility of infection in contiguous areas. It is especially helpful in assessing joints that are difficult to formally evaluate upon purely clinical assessment, such as the sternoclavicular, sacroiliac, and acromioclavicular joints. It may determine if fluid collections are septated or loculated, which require more aggressive techniques to access and drain. Ultrasound by an experienced user may provide similar information and may guide needle placement for aspiration.


For NG septic arthritis, synovial biopsy is not indicated. Certainly, if NG septic arthritis is ruled out and conditions such as tuberculous arthritis, fungal arthritis or atypical Lyme arthritis are still in the differential diagnosis, biopsy with special staining, culture and PCR may be useful.

Interpretation of test results

If the Gram stain is negative, empiric antibiotic therapy is still warranted in situations where there is a reasonable suspicion for septic arthritis. Positive Gram stains will help guide empiric therapy until culture results return (see Table I). It is reasonable to consider Gram-positive cocci in clusters to be MRSA until proven otherwise.

Table I.

Initial Treatment Options for Non-Gonococcal Septic Arthritis Based Upon Clinical Settings*

Besides the importance of obtaining culture of the synovial fluid, there is little uncontroverted evidence that history, physical examination, laboratory studies and x-rays are adequate or even helpful to rule out NG septic arthritis. The experienced clinician must make judgment calls.

Certainly, not every joint effusion requires an aspiration and culture. However, when there is potential for septic arthritis, the experienced clinician is not dissuaded from culturing synovial fluid and treating with empiric antibiotics in the presence of mixed or unsupportive clinical findings or diagnostic tests. The consequences of a missed diagnosis can be severe joint destruction.

How should patients with bacterial arthritis be managed?

Empiric treatment may be guided by the Gram stain results, if they are available. A negative Gram stain or insufficient fluid to perform a Gram stain would require empiric therapy based upon the clinical setting. Modifications in treatment may be made after the culture results have returned (see Table I).

If no Gram stain results are available and empiric therapy is warranted, coverage for MRSA is advised. This approach will also cover most streptococcal species. In special populations, such as the immunosuppressed or patients with chronic infections (such as UTI), or colon inflammation/infections or cancer, the addition of coverage for Gram-negative organisms is needed. In younger, sexually active individuals coverage for gonococcal arthritis, in addition to MRSA, should be considered.

If the Gram stain shows Gram-positive cocci in clusters, coverage for MRSA is wise until culture results return. Streptococcal infections will be empirically covered by most antibiotics for MRSA.

Gram-negative rods on Gram stain require suitable coverage (see Table I), but it may be wise to also cover for Staphylococcus aureus until the culture results have returned. The culture and sensitivity results will allow refinement of antibiotic therapy.

Duration and route of therapy has been anecdotally defined with reasonable agreement. There is a paucity of high quality studies in the adult literature that address this. Meta-analyses are hampered by small studies of poor or inconsistent design. Despite this, there is reasonable agreement on duration and route of therapy.

With variation based upon the clinical setting, a total of 6 weeks of antibiotic therapy is usually adequate to eradicate the infection. Typically, the first 2-3 weeks is given intravenously.

Although high quality studies are lacking, there is strong agreement that drainage of joints with high WBC counts is warranted to prevent damage and destruction to cartilage and other tissues. There appears to be a subtle preference for open drainage (arthrotomy), however, it is perfectly reasonable to serially aspirate joints that are easily accessible. Open drainage or arthroscopy may be needed for septated or highly loculated fluid, or for fluid that is difficult to access by joint aspiration.

Again, there is little evidence, but much agreement that drainage should continue until the WBC count consistently fall below 50,000 WBC/mm3.

Simple ROM exercise is wise (within the limits of the setting). However, more aggressive physical therapy should be reserved until after the acute inflammation has resolved. The extent of weight bearing can be determined largely by the patient, as long as the joint is stable.

What happens to patients with bacterial arthritis?

The duration of untreated infection is the most important determinant of outcome in NG septic arthritis. Early diagnosis and treatment improves outcomes. An untreated NG septic joint can be destroyed in a matter of days. Certain joints rendered terminal by septic arthritis may be amenable to total joint arthroplasty, once the infection has been eradicated.

How to utilize team care?

Questions regarding choice of antibiotic therapy, particularly in complicated settings, may require the assistance of an infectious disease consultant. A rheumatologist or interventional radiologist may be needed to access more difficult joints. Orthopedists may be needed to access deep joints or joints that are septated or where fluid is loculated or sequestered. Orthopedists are also needed for open drainage (arthrotomy), arthroscopy or rarely bone biopsy of contiguous areas. Orthopedists may perform total joint arthroplasty after the infection has been eradicated.

Close attention to wound drainage equipment is required.

Physical and/or occupational therapy is limited to ROM during the acute phase. Afterwards, attention is given toward optimizing ROM and function; with strengthening of muscles surrounding the affected joint. PT and OT play important roles in preparing the patient for upcoming total arthroplasty as well as during the rehabilitation period.

Are there clinical practice guidelines to inform decision making?

There are no clinical practical guidelines.

Other considerations

  • 711.0 Pyogenic arthritis

  • 716.6 Unspecified monoarthritis

  • 716.5 Unspecified polyarthropathy or polyarthritis

  • 711.9 Unspecified infective arthritis

  • 711.4 Arthropathy associated with other bacteria diseases (not gonococcus or menningococcus)

  • 81.91 Arthrocentesis

For most uncomplicated cases of septic arthritis, the first few days of intravenous antibiotic therapy is given in the hospital setting. The patient is usually discharged with arrangements for home IV therapy as soon as they are clinically improved, and the WBC count in the affected joint falls below 50,000 WBC/mm3.

What is the evidence?

Ross, JJ.. “Septic arthritis of native joints”. Infect Dis Clin North Am. vol. 31. 2017. pp. 203-218. (This recent update addresses clinical presentation, identifying features and patterns.)

Kim, H,, Kim, J,, Ihm, C. “The usefulness of multiplex PCR for the identification of bacteria in joint infection”. J Clin Lab Anal. vol. 3. 2010. pp. 175-81. (This emerging technology has the potential to shorten the interval between diagnosis and treatment, and the potential for improved outcomes resulting from early and specifically directed intervention.)

Chander, S,, Coakley, G. “What’s new in the management of bacterial septic arthritis?”. Curr Infect Dis Rep. vol. 5. 2011. pp. 478-84. (The current condition and future management trends are covered in this review of epidemiology, diagnosis, and treatment.)

Butt, U,, Amissah-Arthur, M,, Khattack, F,, Elsworth, CF. “What are we doing about septic arthritis? A survey of UK-based rheumatologists and orthopedic surgeons”. Clin Rheumatol. vol. 5. 2011. pp. 707-10. (This paper compares and contrasts opinions about closed versus open drainage of septic joints in the UK. It is not generalizable and reaches no clear conclusions, but is an example of one of several areas in septic arthritis where opinion dominates evidence in decision making due to a lack of well conducted trials.)

Johns, BP,, Loewenthal, MR,, Dewar, DC.. “Open compared with arthroscopic treatment of acute septic arthritis of the native knee”. J Bone Joint Surg Am. vol. 99. 2017. pp. 499-505. (In this study, outcomes were superior in patients with bacterial arthritis of the knee treated with arthroscopy, rather than open surgical drainage. However, generalizability is limited by the single-centre, retrospective, non-randomized design.)

Garcia-De La Torre, I,, Nava-Zavala, A. “Gonococcal and nongonococcal arthritis”. Rheum Dis Clin North Am. vol. 1. 2009. pp. 63-73. (A broad review that allows the reader to compare and contrast these two forms of bacterial arthritis.)

Ross, JJ,, Davidson, L.. “Methicillin-resistant Staphylococcus aureus septic arthritis: an emerging clinical syndrome”. Rheumatology (Oxford). vol. 44. 2005. pp. 1197-8.

Al-Nammari, SS,, Bobak, P,, Venkatesh, R. “Methicillin resistant Staphylococcus aureus versus methicillin sensitive Staphylococcus aureus adult haematogenous septic arthritis”. Arch Orthop Trauma Surg. vol. 7. 2007. pp. 537-42. (These retrospective studies compare septic arthritis due to methicillin-sensitive and -resistant Staphylococcus aureus, including differences in comorbid conditions, presentations, treatments and outcomes.)

Barcia-Arias, M,, Balsa, AS,, Mola, EM. “Septic arthritis”. Best Pract Res Clin Rheumatol. vol. 3. 2011. pp. 407-21. (This paper is a recent review of diagnostics and therapeutics for septic arthritis.)

Mathews, CJ,, Coakley, G. “Septic arthritis: current diagnostic and therapeutic algorithm”. Curr Opin Rheumatol. 2008. pp. 457-62. (This review highlights the lack of high-quality evidence based medicine to guide clinicians in the management of the inflamed joint.)
**The original author for this chapter was Dr. Raymond Pertusi. The chapter was revised by Dr. John J. Ross.