Since its popularization by Sir John Charnley in 1962 and subsequent globalization, total hip arthroplasty (THA) has become recognized as one of the most successful surgical procedures in all of medicine, providing lasting pain relief and functional restoration. Of the several complications associated with joint arthroplasty operations, none is more challenging than periprosthetic joint infection (PJI).
PJI is by no means a new obstacle for joint replacement surgeons. Deep postoperative wound infections have complicated total joint replacement operations since the birth of the procedure. In the late 1960s and early 1970s, the incidence of infection after primary THA was as high as 7% to 10%. After institution of routine prophylactic antibiotics for surgery in the mid-1970s, the infection rate diminished drastically and remained relatively stable at 1.0% to 2.5% ( Fig.74.1 ). Implementation of modern surgical concepts, such as unidirectional laminar flow, ultraviolet lights, body exhaust suits, patient barrier draping, decreased operating room traffic, and fewer persons involved in each case, led to infection rates of 0.5% to 1.0% for THA and 1.0% to 2.0% for total knee arthroplasty (TKA).
Management of infected joint arthroplasty is a challenge for the orthopedic surgeon, infectious disease physician, microbiologist, anesthesiologist, and patient ( Fig. 74.2 ). A major difficulty when dealing with an infected arthroplasty is making a rapid and accurate diagnosis. This may ultimately dictate the success of ensuing treatment strategies. Historically, physicians have used erratic diagnostic criteria when deciding on whether a failed arthroplasty is septic or aseptic. Most physicians use some combination of clinical examination, joint aspirate white blood cell (WBC) count and differential, cultures, laboratory markers such as the erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level, and radiographic evidence to make this diagnosis. Due to the high numbers of false-positive and false-negative results, none of these tests can be relied on alone for the diagnosis of PJI, leading to an ongoing debate about which combination of criteria is ideal.
In this chapter, we discuss the history and physical examination of patients with suspected PJI, review the diagnostic tests available, propose a standardized definition for PJI, and consider future endeavors in diagnostic testing.
Examination and Imaging
Similar to the diagnostic algorithms in all other medical specialties, when evaluating a patient for possible PJI, the first step is obtaining a thorough history followed by a complete physical examination. The history should encompass the preoperative, intraoperative, and postoperative periods.
Preoperative comorbidities are well-known risk factors for PJI. One group examining more than 80,000 Medicare patients concluded that (in decreasing order of significance) congestive heart failure, chronic pulmonary disease, preoperative anemia, and diabetes were the most common independent risk factors for PJI, with obesity and renal disease providing additional risk factors.
In addition to comorbidities, physicians often ask patients about recent dental procedures. The presumption is that invasive dental work may be a considerable risk factor for PJI. However, several orthopedic surgeons, infectious disease specialists, and dentists now think that dental procedures are not a significant cause of PJI and that routine antibiotic prophylaxis may not be warranted in all procedures. Nonetheless, the American Academy of Orthopaedic Surgeons (AAOS) recommends that in patients with previous joint replacement (especially those who are immunocompromised), antibiotic prophylaxis should be considered before undergoing invasive procedures such as dental work.
Several studies, including one by Pulido and colleagues, have shown that intraoperative blood loss with allogenic transfusion, bilateral arthroplasty, increased operative times, and evidence of previous operations on an extremity are important variables in predicting the risk of PJI. Patients should be asked about a history of trauma, injury, falls, drainage, hematoma formation, or wound dehiscence, because these events may point to the need for a thorough infection workup.
An acute onset of continuous pain may be the only hallmark symptom offered by a patient during the examination. Aside from obvious purulent drainage emitting from the wound or a sinus tract, local signs include joint inflammation, erythema, swelling or effusion, and warmth and tenderness to palpation. Systemic signs, including fevers, chills, and night sweats, may or may not be identified. These systemic signs may not be reliable in the immediate postoperative period; for example, one study showed that development of pyrexia was a very poor indicator of PJI. As is the case with most diagnoses, the history and physical examination can provide valuable information and therapeutic guidance, but further diagnostic testing is usually required.
Outpatient Laboratory Data
For most orthopedic surgeons, the next and sometimes final step in the diagnosis of PJI involves obtaining laboratory data from serum specimens and joint aspirates of synovial fluid. Although a complete blood cell (CBC) count can aid the diagnosis of systemic and various local infections, the utility of serum leukocyte counts in PJI is quite unreliable. An increase in the serum WBC count may be seen in as few as 15% of patients with a deep wound infection after total joint arthroplasty. Toosi and co-workers found that the WBC count and percentage of polymorphonuclear cells (PMNs) had very low sensitivity and specificity for PJI, and the investigators advocated abandonment of their use in the workup of PJI.
For years, the ESR and CRP concentrations have been considered the most important screening parameters available and have been widely used because of their relatively low cost and excellent sensitivity. Of these two parameters, the serum CRP level is sometimes said to be the more important marker because it is considered a more reliable indicator of inflammatory processes. Tests for the ESR and CRP are limited by their very low specificity, and results are greatly affected by age, sex, and underlying medical comorbidities. Because these markers are elevated postoperatively for as long as 60 days, they are often difficult to interpret in patients immediately after surgery.
When PJI is suspected, joint aspiration of synovial fluid can provide a wealth of information about a patient’s infection status. Traditionally, synovial fluid has been cultured to isolate the pathogenic microorganism and guide subsequent antibiotic therapy before and after any surgical intervention. After the discovery of culture-negative PJI, the AAOS recommended keeping patients off of antibiotic therapy for at least 2 weeks before joint aspiration to improve the accuracy of cultures. Traditionally, preoperative prophylactic antibiotics before revision surgery also were withheld, but later studies suggested that the use of antibiotics just before incision would not affect intraoperative cultures and should be given to patients with proven PJI.
There is significant evidence demonstrating that joint fluid should be used for the culture of a potential pathogen and for analysis of the local WBC count and PMN percentage. Several studies, including those by Ghanem and co-workers and Della Valle and associates, have reported that in nonacute infection, aspirate WBC counts higher than 1100 to 3000 cells/μL and PMN concentrations between 60% and 73% have an accuracy of up to 99% in detecting PJI. When dealing with acute infections within the first 4 weeks, a threshold of 10,700 cells/μL for the synovial WBC count and 89% for the PMN fraction has an accuracy of 92% in detecting PJI. With further research, these parameters are likely to become more refined and possibly specific to the hip or knee joint.
Synovial fluid aspirated in the office setting is being studied with novel tests that may increase the specificity and efficiency of diagnosis. The serum CRP level has long been a mainstay for the diagnosis of PJI, but Parvizi and colleagues found the synovial CRP level to be a more sensitive (84.0% versus 76.0%) and specific (97.1% versus 93.3%) diagnostic tool. The investigators think that systemic inflammatory conditions are less influential at the local joint site. To improve diagnostic efficiency and ease, they also introduced the use of simple colorimetric strip detection of leukocyte esterase enzyme secreted by local neutrophils migrating to the synovial infection site. For a leukocyte esterase reading of ++ (categories are negative, trace, +, and ++), the group reported favorable results of 80.6% sensitivity, 100.0% specificity, 100.0% positive predictive value, and 93.3% negative predictive value.
Although radiologic studies are ordered frequently by orthopedic surgeons, they are often of little use without knowledge of when to order them and how to interpret them. Radiologic studies are never the first or last diagnostic test ordered when a patient is thought to have PJI. As the most expensive diagnostic modality, these studies should be ordered only when clinical suspicion yields an intermediate probability of infection and other options have been exhausted.
The standard study is the plain radiograph, which is often routinely ordered at follow-up office visits, regardless of infection suspicion, to evaluate mechanical failure of an implant. Important considerations include obtaining the appropriate views or projections of the joint, ensuring adequate exposure time, and comparing new radiographs with older radiographs to look for interval changes. Unfortunately, the use of radiographs is often not of great use in the diagnosis of PJI because radiographic changes due to infection usually occur only 3 to 6 months after the onset of clinical symptoms. Radiolucent lines surrounding an implant may occur with septic or aseptic failure of a total joint arthroplasty. Specific signs of infection seen late on plain radiographs include a periosteal reaction, rapid migration of the implant, osteolysis in a lacunar shape ( Fig. 74.3 ), and bone resorption at the cement interface.
Bone scans with three-phase scintigraphy are the next most useful radiologic studies in diagnosing PJI. The most commonly used bone scan is technetium-99m methylene diphosphonate ( 99m Tc-MDP), which diffuses mainly to bone hours after injection. The dye concentrates in areas that sustain high rates of bone formation and metabolism. It may be used in patients at least 6 months after surgery who are experiencing a sudden onset of symptoms. The test is highly sensitive in this population, but due to its significant reaction to small differences in bone metabolism, it has a low specificity and yields positive results in many noninfectious conditions, including trauma.
The much more expensive and difficult to administer bone scan is the technetium-99m hexamethyl propylene amine oxime ( 99m Tc-HMPAO) test, which labels mostly neutrophils to differentiate septic (neutrophils) from aseptic (macrophages, lymphocytes, and histiocytes) causes of cell mobilization. This test has false-positive results if performed too early after surgery due to reactive phenomena after surgical trauma. False-negative results occur when dealing with chronic infections characterized by low activity and few neutrophils flooding the chronically infected bone territory.
Three-phase scintigraphy with indium-labeled WBCs ( Fig. 74.4 ) is frequently used by the Japanese and is popular among some physicians in the United States when the diagnosis is uncertain. The described testing modalities often are used in concert to aid in the diagnosis of indolent PJI.
Although research continues in the use of other radiologic techniques, such as ultrasound, computed tomography, and magnetic resonance imaging, there is no significant role for these modalities in the diagnosis of PJI aside from improving the accuracy of joint aspiration procedures. Because of the questionable specificity of radiologic studies, even when a positive result is obtained, continued workup is usually necessary.
Intraoperative and Perioperative Data
Perhaps the final opportunity to diagnose a patient with PJI occurs during the time of revision surgery in the operating room. Open arthrotomy into the joint allows visualization of infectious or inflammatory processes around the joint capsule, implants, or bone. Visual evidence of purulence surrounding the prosthesis and identification of a communicating sinus tract are two important criteria used by many institutions to diagnose PJI.
Perioperative culture is often essential for identification of an organism and subsequent antibiotic therapy. Although there is definite agreement that intraoperative specimens for culture are necessary, there is less certainty about the type of specimen needed, the quantity of specimens required, and the ideal duration of incubation for cultures. Tissue specimens are considered superior to swab specimens, but many community physicians continue to use swabs for culture due to their ease and efficiency. Three to five tissue specimens have been considered sufficient for culture accuracy with 99.6% specificity; however, some physicians still recommend taking five or more tissue specimens during surgery. No standard exists for the duration of culture. Shorter incubation periods may grow only the most virulent organisms, such as Staphylococcus species, and may lead to many false-negative results. Longer incubations of up to 14 days may be required for low-grade infections with organisms such as Propionibacterium . Several other factors affecting culture results, such as antibiotic administration, biofilm formation, and contaminants, continue to be investigated.
Another perioperative diagnostic test involves counting the number of neutrophils per high-power field in histologic frozen specimens. Theoretically, the test is useful in differentiating real infection from contamination, but in practice, it is highly observer dependent, and results vary greatly based on the experience of the pathologist. The AAOS recommends against routine use of frozen section specimens except when PJI has yet to be discounted or confirmed. When frozen sections are used, the general consensus for positive test results is identification of 5 to 10 PMNs in at least five high-powered fields at ×400 magnification.
Previously used on a regular basis, the Gram stain has been essentially cast aside as useless in diagnostic testing of PJI due to its low sensitivity, a conclusion stemming from the incredibly high number of false-negative results it yields. During a revision procedure, visualization into the joint and collection of specimens for culture provide important information in the last phase of diagnosing PJI.
PJI continues to pose one of the more serious threats to a successful outcome after total joint arthroplasty. The accurate, efficient, and consistent diagnosis of PJI is a challenge. There is no shortage of diagnostic tests that may be done in evaluating patients with suspected PJI, but there are definite shortcomings in relying on any one of these tests alone. In obvious cases of PJI, various combinations of diagnostic tests may be successfully used, but in many of the less clear cases of infection, the lack of a gold standard for diagnosis raises concerns among the medical community about patient treatment and consistency.
A workgroup convened by the Musculoskeletal Infection Society (MSIS) in 2011 met to analyze available evidence and propose a new definition for PJI. The goal was to provide a definition for PJI that could “be universally adopted by all physicians, surveillance authorities (including the Centers for Disease Control, medical and surgical journals, the medicolegal community), and all involved in the management of PJI”.
In 2011, Parvizi and colleagues described the diagnosis of PJI using the following criteria :
A sinus tract communicates with the prosthesis, or
A pathogen is isolated by culture from two separate tissue or fluid samples obtained from the affected prosthetic joint, or
When four of the following six criteria exist:
Elevated serum ESR and CRP concentration
Elevated synovial WBC count
Elevated percentage of synovial PMNs
Purulence in the affected joint
Isolation of a microorganism in one culture of periprosthetic tissue or fluid
More than five neutrophils per high-power field in five high-power fields observed from histologic analysis of periprosthetic tissue at ×400 magnification
Based on these proposed criteria, it may be possible to standardize the diagnosis of PJI across the orthopedic community. However, many of the criteria that comprise the new definition are vague about definitive quantitative thresholds for elevated laboratory markers. Current studies are looking into refinement of these thresholds to make it easier for all physicians to interpret the results of diagnostic studies.