PJIs and surgical site infections (SSIs) after TKA are extremely problematic and have been associated with significant morbidity. Given the degree of medical, emotional, and financial impact on the patients, surgeons, and health care systems, substantial efforts have been put forth by the orthopedic community in attempts to limit the risk of infection after elective total joint replacement. Furthermore, the recent implementation of the bundled payment initiatives aiming to improve patient care while reducing costs has put considerable pressure on surgeons and health care systems to minimize complications such as PJIs, which have been shown to have a significant impact in the setting of a value-based system.⁴

The diagnosis of infection after TKA is often challenging and sometimes requires complex decision-making. The ability of the surgeon to make an accurate diagnosis is paramount, given that the treatment algorithms are often vastly different from those of other failure mechanisms. Early steps to diagnosing infection begin with the understanding of risk factors and a focused history with attention to the patient’s relevant medical, surgical, and social history followed by a physical examination. An index of suspicion is then established which will help interpret subsequent testing in accordance with Bayesian theory.

When evaluating a painful TKA, it is important for the clinician to have a high index of suspicion for infection based upon the patient’s risk factors, history, and examination and to ensure that proper workup is established to rule out or diagnose infection. Diagnosis of infection, especially in the acute postoperative period, can be particularly challenging, given that most tests evaluate systemic inflammatory measures and not infection directly. Acute PJI is defined as an infection that occurs within 90 days of surgery, whereas chronic PJIs occur at 90 days or later after surgery. Laffer et al proposed that 45% of infections present early, 23% delayed, and 32% represent late infections.⁵,⁶ Infections originate from bacterial contamination of the implant either during surgery for acute infections or later through hematogenous transfer or local dissemination.⁷ In the early postoperative period, symptoms can be confused with normal healing and inflammatory markers are usually elevated due to the systemic inflammatory response of surgery.⁸,⁹ All patients with a suspected PJI should have a physical examination, appropriate imaging, and inflammatory markers including ESR and CRP. Positive values in the setting of a high clinical suspicion merit an aspiration for further evaluation.

A number of factors that can alter the risk of infection after TKA have been identified; some are associated with increased risk while some may be protective. The use of antibiotic prophylaxis in addition to modern surgical and sterilization techniques has helped reduce the incidence of PJI tremendously over the past few decades.⁹,¹⁰,¹¹ However, PJIs remain a major problem. Factors that appear to reduce the rate of PJI include shorter duration of surgery, prophylactic antibiotics given within 60 minutes of the surgical incision, skin/nasal decolonization, use of antibiotic-impregnated methylmethacrylate fixation cement, preoperative chlorhexidine wash, and directed laminar airflow in the operating room.¹⁰,¹¹,¹²,¹³,¹⁴ On the other hand, multiple studies have identified risk factors that are associated with increased risk of PJI. Pulido et al identified higher American Society of Anesthesiologists (ASA) score, morbid obesity, bilateral arthroplasty, allogenic transfusion, postoperative atrial fibrillation, myocardial infarction, urinary tract infection, and longer hospitalization as independent predictors for PJI based on a 9245 patient analysis study.¹⁵ Similarly, Namba et al identified a number of risk factors associated with deep surgical site infection after primary TKA based on an analysis of 56,216 knees.¹⁶ Their study showed that body mass index (BMI) of ≥35 (hazard ratio [HR] = 1.47), diabetes mellitus (HR = 1.28), male sex (HR = 1.89), an American Society of Anesthesiologists (ASA) score of ≥3 (HR = 1.65), and a diagnosis of posttraumatic arthritis (HR = 3.23) all had an increased incidence of PJI. Furthermore, inflammatory arthritis, sickle cell disease, psoriasis, malnutrition, a compromised immune system, and previous surgery to the limb have all been linked to PJI postoperatively¹⁷,¹⁸,¹⁹,²⁰,²¹ (Table 73-2).

In addition to the recognition of obesity and diabetes being linked to PJI, nutritional status has become increasingly important in optimizing surgical outcomes and preventing postoperative infections and wound complications. Serum albumin, prealbumin, and transferrin levels are well-recognized biologic markers of the patient’s nutritional status, with serum transferrin levels having been shown to be more sensitive than both lymphocyte counts and albumin levels in predicting postoperative infection rates.²²,²³,²⁴,²⁵ Patients with low nutritional markers should raise the index of suspicion for infection, and malnourished patients should be nutritionally optimized prior to surgery.²⁶

Inflammatory arthritis is also recognized as a well-known risk factor for the development of infection after TKA.¹⁷,²⁷ Rheumatoid arthritis (RA) patients have been shown to be twice at risk of developing infection, regardless of the site, than the non-RA population.²⁷ Comparing osteoarthritis with RA in more than 2000 total knee arthroplasties, the rate of infection was 2.4 times greater in patients with RA with conventional designs and 2.5 times greater for hinged total knee designs.²⁸,²⁹ Wilson et al reviewed 4171 total knee arthroplasties performed between 1973 and 1987 and found that 67 (1.6%) developed PJI. The incidence of infection was significantly higher for knees affected by RA (45 [2.2%] of 2076) than those affected by osteoarthritis (16[1%] of 1857).³⁰ Similarly, patients with psoriatic arthritis
have an increased infection rate after TKA. Twenty-four arthroplasties performed in 16 patients with established long-standing psoriatic arthritis were reviewed. Four (17%) subsequently developed a deep infection. Two of the infections occurred at 1 and 6 months postoperatively; the other two occurred 3 and 5 years postoperatively.³¹ Recent studies are showing better results with the emergence of biologics and new psoriasis treatment modalities.³² Diagnosis of infection can be more difficult in patients with inflammatory arthritis, given the clinical picture, symptomatology, and presence of elevated inflammatory markers at baseline. However, Della Valle et al found that the utility of all serum and synovial tests for predicting chronic periprosthetic joint infection was similar for patients with noninflammatory and inflammatory arthritis and therefore advocate for using similar ESR, CRP, and synovial fluid white blood cell count with differential cutoff values in patients with inflammatory arthritis for the diagnosis of PJI.¹⁷

A history of old infection in the knee prior to TKA has demonstrated an increased susceptibility to the development of deep infection after TKA. In a review of 65 total knee arthroplasties with a history of infection, deep infection occurred in five (7.7%) overall. In 20 of these 65 patients who demonstrated prior history of both bone and joint infection, three (15%) developed deep infection. In the other 45 patients who had a prior history of only joint sepsis, two (4%) developed deep infection.³³ Moreover a history of PJI in a different joint, despite being adequately treated, increases the risk of PJI after TKA.³⁴

Timing of intra-articular steroid injections in the knee has been linked to postoperative infection. Browne et al showed that there was no significant difference in patients who underwent TKA more than 3 months after injection based on a national database study.³⁵ The incidence of infection within 3 months (2.6%, OR 2.0 [1.6-2.5], P < .0001) and 6 months (3.41%, OR 1.5 [1.2-1.8], P < .0001) after TKA within 3 months of knee injection was significantly higher than the control cohort. Meanwhile there was no significant difference in patients who underwent TKA more than 3 months after injection.³⁶,³⁷ In contrast, another large database study performed by Bedard et al showed that there is an increased risk of PJI if the knee injection occurred within 7 months (OR 1.38-1.88; P < .05), but the risk returns to baseline after 7 months.³⁸

Numerous additional endogenous or host risk factors have been identified that may predispose certain patients to postoperative septic complications. These include smoking, alcohol abuse, opiates use disorder, psychiatric illnesses, substance abuse disorder, chronic renal failure, infection at a remote site from TKA, malignancy, oral steroid use, and development of postoperative hematoma, as well as increased operative time >121 minutes.¹⁸,³⁹,⁴⁰,⁴¹,⁴²,⁴³,⁴⁴,⁴⁵

Patient risk factors can be classified as modifiable and nonmodifiable risk factors. Many preoperative risk factors for the development of PJI are related to patient comorbidities and are, therefore, modifiable. Many studies to date highlight the need for preoperative optimization of modifiable risk factors to further reduce the risk of complications after surgery.²⁰,⁴⁶,⁴⁷,⁴⁸

Patients often present with nonspecific symptoms of knee pain following TKA. Though this is a common complaint, particularly in the postoperative period, symptoms concerning for PJI include constitutional symptoms (fevers, chills, malaise), drainage, and/or new or progressive pain and decline in function.

A patient history should include the types of prostheses, date of implantation, past surgeries on the joint or limb, history of wound healing problems following surgery, remote infections, current clinical symptoms, comorbid conditions, prior and current microbiology results from aspirations and surgeries if applicable, and antimicrobial therapy for the PJI including local antimicrobial therapy. While being febrile is traditionally associated with infection, it is important to distinguish fever resulting from infection from a normal postoperative course.⁴⁹,⁵⁰ A recent study of 100 patients undergoing total hip arthroplasty (THA) and 100 patients undergoing TKA showed that the normal postoperative febrile response peaked on the first postoperative day and normalized by the fifth day. In 19% of patients, the maximum body temperature was between 39°C and 39.8°C.⁵¹ Febrile episodes can be a sign of other complications with significant morbidity and mortality, such as atelectasis, hematoma, urinary tract infection, fat emboli, or venous thromboembolism (deep vein thrombosis/pulmonary embolism).⁴⁹

Unremitting pain with or without swelling and fever in the acute (within 90 days) postoperative period is uncommon, and it should suggest the possibility of deep infection. Similarly, new-onset pain with or without swelling and fever, preceded by a quiescent period, is suggestive of late hematogenous infection. In either case, AAOS clinical practice guidelines recommend against initiating antibiotic treatment in patients with suspected periprosthetic joint infection until after an appropriate workup, including cultures from the joint, has been obtained.⁵²

Subclinical infections, with no overt physical findings of inflammation, swelling, or drainage, are more diagnostically challenging. Host factors, the infecting organism, or previous administration of antibiotics may contribute to the maintenance of a subclinical infection. Although certain mechanical factors may be excluded by physical examination, pain may be the only consistent finding in the presence of deep joint infection. In a study of 52 patients with infection after total knee replacement, pain was present in 100% of the patients. Seventy-seven percent had swelling of the knee, whereas 27% had active drainage.⁵³,⁵⁴

Prolonged drainage or delayed wound healing may also aid in the diagnosis of infection as they are recognized as a risk factor for the development of deep periprosthetic infection.⁵⁵ In the acute postoperative period, a small amount of drainage is not uncommon. However, prolonged wound drainage can effectively serve as a conduit through which bacteria can enter the wound. Drainage occurs in approximately 25% of knee arthroplasty cases and may be further classified culture-negative or culture-positive.⁵⁴ Cultures taken from wound drainage are generally unreliable, and it is not recommended to base antibiotic treatment solely on this result.