Infectious Disease Compendium

Prosthetic Joint Infection


2012 IDSA Guidelines

They can be acute or chronic, they can be in the joint space or in the stem (often presenting with pain and loosening), they can be hematogenous or acquired at the time of surgery.

Chronic infections tend to be stem infections and present just with progressive pain, night pain, and joint loosening (espcially if occurring in the first one to 2 years). CRaP and ESR are nice if elevated (both up have an 86% chance of infection), but no test is perfect. If a preop tap for revision has a WBC > 1700 or more than 65% PMN's there is a >95% chance there is an infection (PubMed). If three intra-op cultures are positive it also suggets a high likelihood of infection. Duh.

There are a variety of synovial fluid markers for the diagnosis of periprosthetic joint infection including C reactive protein, leukocyte esterase, interleukin6, interleukin1β, α-defensin, and interleukin 17 (PubMed). They are all good with perhaps a nod to alpha defensin as best of breed.

A positive alpha-defensin from joint fluid is reported to have a sensitivity of 100%, a specificity of 98%, a positive predictive value of 96% and negative predictive value of 100% (PubMed).  My rule of thumb is that a test is half as good in the real world, but still looks to be a good test.

On frozen section >10 PMN's per high powered field means infection (PubMed).

Epidemiologic Risks

Intra-operative acquisition, bacteremia (S. aureus has a ~34% chance of seeding a prosthetic joint. Really (PubMed). It is the literature but I don't believe it).

Asymptomatic white cells in the urine is not a risk for infection and is NOT a reason to postpone surgery (PubMed).

And are antibiotics necessary in hip arthroplasty with asymptomatic bacteriuria? Spoiler alert: No (PubMed). Although asymptomatic bacteriuria (ASB) is a risk for prosthetic joint infection, it is a marker since the organism in the joint is NOT in the one in the urine and treating the ASB does nothing (PubMed).


Acute joint: S. aureus, coag negative Staphylococcus (#1 cause), Streptococci, occasionally gram-negative enteric organisms (like E. coli).

Chronic stem: S. aureus, coag negative Staphylococcus (#1 cause), gram-negative enteric organisms (like E. coli), and Propionibacterium acnes for shoulders and P. avidum for hips (PubMed).

It you have a chronic infection/late and the initial operative cultures are negative, have them hold the cultures for two weeks (most labs will only hold the cultures for 5 days) and you will increase the yield, predominantly coag negative Staphylococcus and Propionibacterium (PubMed)(PubMed).

Empiric Therapy

GET CULTURES. It may behoove you to hold the cultures for 14 days as it increases the yield of fastidious organisms like Propionibacterium (PubMed). After cultures are obtained, vancomycin +/- third generation cephalosporins.

If it is an acute joint infection AND streptococcal or methicillin susceptible S. aureus AND you debride the joint with poly exchange (and what is a parrot doing in a joint?) AND treat with 6 weeks (some suggest 2-3, I am not that convinced) of an anti-staph beta lactam IV PLUS rifampin (450 mg po bid) AND follow up with 6 months of po quinolone PLUS rifampin you may salvage the joint. The rifampin is key, despite all the side effects of the medication. One study with small numbers of patients, monotherapy with moxifloxacin for three months had an 80% success rate. I am sticking with combo therapy (PubMed).

High dose daptomycin (8-10 mg/kg) plus rifampicin is also effective (PubMed)(PubMed); maybe push the daptomycin to 10/mg/kg but except a 50% failure rate (PubMed).

Clindamycin and rifampin (PubMed) is also an effective combination. I would still consider it only if there were not other options (PubMed).

Debridement with joint retention has about a 55% chance of working in one large retrospective study (PubMed); oddly there was no long term difference between MRSA and MSSA infection for long term success. I would have bet a worser outcome for MRSA.

If debridement and retention fails, take out the joint, treat 6 weeks with IV, wait a bit (I find the duration is surgeon dependent and there is no data to guide us) then a new hip. In one retrospective study, those who received at least 14 days po antibiotics after reimplantation had fewer infections (PubMed).

There is more data to support better outcomes with a 2 stage procedure over a one stage (PubMed).

Much to my surprise, in a retrospective study ceftriaxone was equal to oxacillin for MSSA infections (PubMed).

With aggressive debridement, IV therapy and long term po with a quinolone, you might salvage (75% probability) gram negative infections (PubMed). If it is ciprofloxicin resistant, the chance of salvage lessens (PubMed).

How long to give oral therapy is not clear. At least 3 months (Pubmed).

Lifetime suppression, especially if that lifetime is not going to be all that long, may be warranted; the goal is function more than cure.  Suppression increases the chance of keeping the joint long term even with S. aureus (PubMed). Pick an antibiotic that is inexpensive and non-toxic. In one series doxycycline was effective (PubMed).

Cure without rifampin is overall maybe 60% (14-83); chance with rifampin is an extra 10-20% in the salvage rate.

Even with streptococcal infections rifampin increases cure, which surprised me (Pubmed)(PubMed). I would haver thought rifampin would add nothing for streptococci.

Despite the long tradition of using antibiotic spacers, the data confirming the use of antibiotics in spacers is still of poor quality (PubMed). And I have seen several cases of aminoglycoside toxicity from spacers (PubMed).


If MRSA, poor soft tissue (sinus tract), loose joint (a stem infection is harder to cure than joint infection), diabetic, or smoker, probably not salvage the joint, but you always try, don't you?

Prosthetic joint infections can be particularly difficult to treat when S. aureus makes small colony variants.

Chronic suppression works and patients who never stop their po antibiotics are more likely to keep their joint.

Don't even try and salvage a gram negative rod prosthetic joint infection. It will not work (PubMed).

There is no reason the treat asymptomatic bacteria in the urine preop: it does NOT decrease the risk of infection, nor does antibiotics for GU procedure (PubMed).

If the prosthetic joint is infected with Enterococcus, monotherapy is equal to combination (PubMed). If the infection is in the stem of the prosthetic joint, rather than the joint space, medical cure is probably impossible.

I prefer to take out the joint, give six weeks of IV therapy, wait a month, and if the joint is not infected, replace the joint. I know there is data for primary exchange, but the risk if repeat for 3 months, could be a death sentence.

Having a positive culture at reimplantation, which occured in an amazing 17% (I can remember two in my storied career) increases failure rate (PubMed).


While suggested by many authorities, there is no data that suggests prophylactic antibiotics are effective in preventing prosthetic joint infections if continued > 24 hours. It only breeds resistance. Actually (the favorite word of a skeptic when beginning a sentence), you do NOT NOT NOT need to do dental prophylaxis for joints. There is a study (PubMed), and I quote, "Dental procedures were not risk factors for subsequent total hip or knee infection. The use of antibiotic prophylaxis prior to dental procedures did not decrease the risk of subsequent total hip or knee infection." Ha!

Curious Cases

Relevant links to my Medscape blog

Any Bug, Any Place

Working with minimal data

A Case From One of the Country's Best

Toxic See Mint

Late Failure

A Bit of Whinging Then a Case

Weird PJI. Part One

Weird PJI 2: Electric Boogaloo

Bog Body

Last Update: 04/03/18.