This article is more than 5 years old. Information may no longer be current.
Review of infected TKA cases reveals a few lessons and caveats for diagnosis
Investigator explains when to take intraoperative cultures, how to handle false positives and more.
Click Here to Manage Email Alerts
In the hopes of nearing established guidelines for infection diagnosis, orthopedic surgeons at three centers evaluated the handling of infection in nearly 200 infected total knee arthroplasty patients.
Investigators collected data on 889 total knee arthroplasty (TKA) cases, including 197 infected cases, from databases at Tulane University Medical Center in New Orleans, the Rothman Institute in Philadelphia and the Washington University School of Medicine in St. Louis, according to Robert L. Barrack, MD.
“Early, accurate diagnosis is critical in managing the infected knee, yet there is not uniform agreement in what even constitutes an infection,” Barrack said. “A number of challenges in diagnosing infection persist because there is not a single test that’s 100% accurate.”
Data collection
|
Barrack and his colleagues analyzed the 889 patients’ joint fluid for cell count and differential gram stain, routinely at one center and selectively at two centers.
In cases of conflicting data in preoperative tests, the investigators also obtained intraoperative frozen sections.
All three databases routinely recorded patients’ white blood cell count, erythrocyte sedimentation rate (ESR), intraoperative cultures and gram stains, Barrack said.
The investigators found 692 noninfected and 197 infected cases.
Barrack presented the study results at a recent Current Concepts in Joint Replacement meeting.
Lower infection threshold
Investigators first discredited the theory that an infection is present if a patient has greater than 50,000 white cells and greater than 90% polymorphonuclear leukocytes (PMN). The threshold should be lower, Barrack explained.
“In this series, noninfected knees averaged less than 500 cells with less than 30% PMNs. So the best threshold value to pick is only about 75% PMNs with 2,000 white cells. That would give you a positive predictive value of 96%, which is actually as good or better than the results of the intraoperative culture itself.”
The investigators also found that intraoperative cultures may not be an absolute gold standard after they found 12 false negative cultures among infected patients. “There is 6% of cases that will not grow anything in spite of the fact that you know they’re infected – even with gross purulence,” Barrack said.
Also in this series, 41 uninfected patients presented with false positive cultures.
|
Images: Barrack RL |
Investigators classified 29 cultures as definitely false, based on the following criteria: a single positive culture, growth on the broth only, single organism that was a nonpathogen and nonvirulent, a “rare” growth, and no preoperative evidence of infection. Seven had other evidence of infection.
Out of the 41, only 16 received intravenous antibiotics, including 11 of the 12 patients with other evidence of infection and five without any other evidence of infection, Barrack said.
“A single positive intraoperative culture does not require treatment in the absence of any other evidence of infection,” Barrack said.
He recommended obtaining intraoperative cultures when there is preoperative suspicion of infection or suspicious tissue.
But Barrack warned against obtaining a single culture, and said to obtain five cultures or none.
Antibiotic use
In half of the cases, surgeons routinely gave patients preoperative antibiotics, Barrack said.
However, 90% of the time preoperative antibiotics did not prevent demonstrating the same organism as the preoperative aspirate; withholding the antibiotics demonstrated a new organism less than 5% of the time.
Despite this, Barrack said, “Withholding antibiotics should rarely be done … You should withhold antibiotics when you do not have an organism on an aspirate where you think that it might be infected.”
Barrack said surgeons should routinely aspirate, because it provides useful information in almost every case. Surgeons can avoid aspirating if the ESR and C-reactive protein are “absolutely normal,” and if the patient is more than 5 years postoperative and has no clinical or radiographic evidence of infection, Barrack said.
Although diagnosis of infection is still an unsolved problem in some cases, Barrack said that understanding the available tests will provide accuracy 90% to 95% of the time.
He added: “With new technology, this promises to improve to 98% or 99% in the near future.”
For more information:Reference:
- Robert L. Barrack, MD, can be reached at Washington University School of Medicine, 660 S. Euclid Ave., Campus Box 8233, Department of Orthopedic Surgery, St. Louis, MO 63110; 314-727-2592; e-mail: barrackr@wustl.edu. He is a consultant for Smith & Nephew and is also a member of the company’s advisory committee/review panel.
- Barrack RL. The infected knee: Diagnosing an unsolved problem. #92. Presented at the 23rd Annual Current Concepts in Joint Replacement Winter 2006 Meeting. Dec. 13-16, 2006. Orlando, Fla.