Total hip or knee joint replacement has been one of the great gifts of medicine to mankind. Millions of individuals who would otherwise spend the remainder of their life inactive, limping painfully with a stick or crutches or wheelchair-bound, can now live relatively painless, enjoyable, productive lives. However, the physical and emotional devastation that is wreaked if an infection involving the total joint replacement occurs is almost immeasurable1. Morbidity is substantial and may be permanent. In addition, if revision arthroplasty is required, the mortality rate is elevated and the economic cost to the hospital averages almost four times that of a primary joint replacement2.
On the average, a total hip or knee replacement endures about one million cycles of use per year. As a result, these mechanical devices would be expected to eventually loosen and/or wear out. This gradual process may take one or more decades. Bone cement may loosen, and high-density polyethylene spacers may wear down to create pseudo-laxity of the ligaments. The polyethylene wear-particle debris will also excite an erosive inflammatory reaction that further exacerbates the wear process. A treating surgeon then faces a dilemma—is the etiology of the resulting painful, loose total joint replacement due to the natural mechanical wear process, or is it due to an infection? How should the joint revision process proceed? What steps are appropriate during the preoperative workup so that the surgeon does not inappropriately revise an infected joint and thus create another infected joint? An algorithm is presented by Dr. Parvizi to help in this decision-making process.
In Dr. Parvizi's superb article, some of the decision-making is straightforward. According to the criteria used at the authors’ institution, a prosthetic joint is regarded as infected if: (1) there is an open wound or sinus tract; (2) purulence is encountered in the joint; (3) preoperative or intraoperative tissue or synovial fluid cultures are positive; or (4) at least three of the following four markers in the serum or aspirated synovial fluid are elevated: erythrocyte sedimentation rate (ESR) > 30 mm/hr, serum C-reactive protein (CRP) level > 10 mg/L, synovial white blood-cell (WBC) count > 1700 cells/μL, and synovial polymorphonuclear (PMN) percentage (differential) > 64%.
In spite of these straightforward guidelines, a substantial group of patients still falls in a “gray area” in which diagnosis of the etiology of a painful, loose prosthetic joint remains indeterminate. Patients with active autoimmune disease, patients being treated with corticosteroids, and patients who have previously been treated with antibiotics for an unrelated condition (e.g., urinary tract infection, pneumonia, or an abscessed tooth) may all have a periprosthetic joint infection and yet meet none of the criteria spelled out by Dr. Parvizi. For instance, if a patient was recently treated with oral antibiotics when an abscessed molar was extracted, attempts to culture bacteria from the joint aspirate may fail. If a patient with asthma or rheumatoid arthritis is being treated with prednisone, the ESR value may remain in the normal range.
A surgeon performing revision total joint arthroplasty must always be alert to the possibility that unrecognized infection is present. Frozen sections or a “stat” (rush) Gram stain are frequently employed intraoperatively to help in the decision-making process, but this adds time and cost to the procedure. (A technetium-99 scan is expensive and will not help to determine the etiology of loosening, as the result will be positive regardless of whether the prosthesis is loose because of mechanical factors or infection.) The leukocyte esterase strip test proposed by Dr. Parvizi is inexpensive and highly accurate, and it can yield a result within one to two minutes. The test strips are so inexpensive that a three-pack can be ordered for less than $2.00. A ++ test reading is ∼80% sensitive and 100% specific, and a negative or trace reading has a negative predictive value of >97%. Thus, if the preoperative workup is equivocal and the reading on a leukocyte esterase strip test performed intraoperatively is negative or trace, it is highly likely that the surgeon's problem will revolve solely around the technical issues of revision arthroplasty. A result of ++ on the leukocyte esterase test, on the other hand, would mean that it is highly probable that infection is present, resulting in an entirely different treatment algorithm.
As pointed out by Dr. Parvizi, further confirmation of this diagnostic test is needed. Are there clinical situations that could lead to false positives or false negatives? What should the surgeon do when faced with a + reading rather than a ++ reading? Nevertheless, this article has provided us with a great tool that is both simple and inexpensive and can assist very quickly with intraoperative decisions during revision arthroplasty.
1. Whitehouse JD, Friedman ND, Kirkland KB, Richardson WJ, Sexton DJ . The impact of surgical-site infections following orthopedic surgery at a community hospital and a university hospital: adverse quality of life, excess length of stay, and extra cost. Infect Control Hosp Epidemiol. 2002;23:183-9.
2. Sculco TP. The economic impact of infected total joint arthroplasty. Instr Course Lect. 1993;42:349-51.