Are More Specimens Always Better? Microbiologic Diagnosis of Periprosthetic Joint Infections

Jan. 8, 2018

Clinical microbiology labs often receive multiple specimens for periprosthetic joint infections. How many is enough?

Microbiology laboratory technologists working in the Specimen Receiving Area will "phone a friend" for advice occasionally when stumped by a problem. This time, I was called to help guide test set-up on a set of 12 specimens collected from a patient's knee joint. The patient had his knee replaced with a prosthesis a year ago and was experiencing increasing pain in the joint and limited mobility. The surgeon suspected a periprosthetic joint infection. She had ordered at least 6 different tests on each of the 12 specimens collected intra-operatively from the joint, 16S ribosomal RNA gene PCR with sequencing, and various target-specific molecular assays. Therefore, a total of 72 unique microbiology tests were ordered on this patient's joint alone.

Situations such as these can bring a busy lab to a standstill, considering the upfront processing that is involved in setting up this array of tests. Some of these ordered tests involved introduction of the joint fluid into blood bottles; some required sonication of the explanted prosthesis; and others required multiple extractions for a variety of downstream PCR tests. Almost all of the ordered culture tests involved inoculation of a variety of solid and liquid media types.

Many questions were brought up as we poked through the various specimens. Besides the limitation of having enough specimen material for performance of all these tests, what is the evidence supporting this practice? How many specimens should usually be submitted for microbiologic diagnosis of a periprosthetic joint infection? Is the performance of all of these tests on all submitted specimens a cost-effective approach to practicing medicine?

schematic of prosthetic joint infection using a prosthetic hipSchematic showing a total hip arthroplasty in place, with relevant structures highlighted. Source.

Diagnosis of prosthetic joint infection (PJI) is challenging, since no gold standard exists for diagnosis. Various groups have defined PJI, including the Musculoskeletal Infection Society (MSIS), the Infectious Diseases Society of America (IDSA), the American Academy of Orthopaedic Surgeons (AAOS), and the International Consensus Group on Periprosthetic Joint Infection. The IDSA clinical practice guidelines proposed a variety of clinical, histopathologic, or microbiologic criteria which may be used either as definitive evidence of PJI, or merely as one piece of supporting evidence. Microbiologic culture techniques may be either definitive or supporting evidence, with the strength of evidence dependent upon the number of specimens in which microorganisms are detected, the level of detection, and the specimen type. The various definitions are too intricately detailed for this blog post but are further outlined in the guidelines themselves. Various organizations such as the AASO and the International Consensus Group on Periprosthetic Joint Infection are currently updating their guidelines.

Why is it unacceptable to obtain only a single tissue specimen for culture of PJI? The low sensitivity of a single specimen for culture is a concern for PJI diagnosis, but perhaps the biggest reason this should not be done is that it is impossible to differentiate between contamination and a real infection if there's only one tissue cultured and it is positive for an organism of low virulence. The most common organisms associated with PJI include Staphylococcus aureus and Staphylococcus epidermidis, but a long list of other organisms can be involved, many of which are part of the normal skin microbiota. Predictably, the positive predictive value of an organism as a cause of PJI increases as the number of cultures positive for that organism increases.

So, how many specimens should be submitted for microbiologic diagnosis of PJI? A quick answer to this question is that the clinical microbiology laboratory should continue to expect 3 or more specimens collected from each affected joint when PJI is being considered. This question was examined in the 1990s, when studies supported by mathematical modeling demonstrated that 5 or 6 or more perioperative samples were needed to accurately diagnose PJI. However, the current body of evidence suggests fewer. A recent study determined that the optimal number of specimens for PJI diagnosis depended on the manner in which culture was performed. For specimens inoculated into blood culture bottles, three periprosthetic tissue specimens are needed. If standard plate and broth cultures are being performed, four periprosthetic tissue specimens are required for an accurate PJI diagnosis. The standard in the field is to consider two (or more) positive cultures for the same organism as indicative of PJI. There is concern with sending too many periprosthetic specimens for culture to the laboratory; namely, the chances of a contamination event rise with a higher number of specimens which could lead to a false positive result.

Specimen types collected intraoperatively for microbiologic assessment of PJI include tissues taken in and around the area of apparent inflammation, synovial fluid aspiration (intraoperatively and perhaps even pre-operatively), and perhaps even the prosthesis itself. Swabs are never acceptable, and play no role in the diagnosis of PJI. Joint fluid may be inoculated into aerobic and anaerobic blood culture bottles or onto solid and liquid media, once the validation of such specimen type is performed. The explanted prosthesis may be sonicated and the sonicated fluid inoculated onto solid media (for semiquantitative culture). Tissues may be subjected to various sorts of processing. A comprehensive review of the data around these practices, and a general review of PJI overall, is offered by Tande and Patel in a recent Clinical Microbiology Reviews article.

Finally, to answer some of the questions brought up by the technologists for our case patient above: the testing approach is overkill and is not indicated - too many specimens and too many tests! There are little data in the literature concerning cost-effective microbiologic approaches to PJI diagnosis, and studies which address cost impact typically only compare the costs associated with multiple culture specimens. In the future, there is sure to be more research in the area of PJI microbiologic diagnosis. Areas that are ripe for investigation include the optimal number of specimens for microbiologic studies, the role of multiplex PCR panels which identify PJI pathogens, and the advantages of sonication and mechanized agitation for explants and tissues.


Photo credit: Mock up of artificial shoulder implanted by French surgeon Jules Emile Pean in 1893. The impant remained in the shoulder of the patient, a waiter, for two years, when it had to be removed due to infection.

The above represents the opinions of Dr. Audrey Schuetz and not necessarily of the American Society for Microbiology.

Author: Audrey Schuetz 

Audrey Schuetz 
Audrey Schuetz is Associate Professor of Laboratory Medicine and Pathology at Mayo Clinic College of Medicine and Science in Rochester, Minnesota.