Tuesday, June 27, 2017

What is an infection anyway?

Medical comorbidities and perioperative allogeneic red blood cell transfusion are risk factors for surgical site infection after shoulder arthroplasty

These authors sought to determine surgical site infection (SSI) risk due to medical comorbidities or blood transfusion after primary or revision shoulder arthroplasty. They collected data on medical comorbidities, surgical indication, perioperative transfusion, and SSI were obtained for 707 patients who underwent primary or revision hemiarthroplasty or total shoulder.

For the purpose of this study, SSI was defined in a rather particular manner:  either (1) treatment of a superficial infection within 30 days of surgery with d├ębridement by the treating surgeon or with antibiotics by either the treating surgeon or an infectious disease specialist or (2) treatment of a suspected or confirmed deep infection by return to the operating room for d├ębridement, component exchange, or explantation of components or treatment with therapeutic or long-term suppressive antibiotics by an infectious disease specialist. Positive cultures on return to the operating room were not a requirement for diagnosis of SSI.

Using this set of definitions, the SSI rate was 1.9% for primary hemiarthroplasties and 1.3% for primary total shoulder arthroplasties.

Revision arthroplasty or prior open reduction and internal fixation had higher SSI risk than primary arthroplasties (incidence risk ratio [IRR], 11.4; 95% confidence interval [CI], 3.84-34.0; P < .001).

Among primary arthroplasties, SSI risk factors included male gender (IRR, 60.0; CI, 4.39-819; P = .002), rheumatoid arthritis (IRR, 8.63; CI, 1.84-40.4; P = .006), and long-term corticosteroid use (IRR, 37.4; CI, 5.79-242; P < .001). 

Perioperative allogeneic red blood cell transfusion significantly increased SSI risk and was dose dependent (IRR, 1.68 per unit packed red blood cell; CI, 1.21-2.35; P = .002).

The culture results are shown below.

Comment: This series of cases points to the complex and uncertain relationship between cultures and clinical findings. Patients without the characteristic signs of infection can have positive cultures. Patients with characteristic signs of infection can have negative cultures.

The culture protocol used for these patients is not explained. Specifically, we do not know which cases were cultured, how many specimens were submitted for culture, what culture media were used, and how long the cultures were observed. It is known that unless 5 deep specimens are cultured on three different media and observed for 3 weeks, there is a substantial risk of overlooking Propionibacterium in the wound.

Here are some related posts:

Low-grade infections in nonarthroplasty shoulder surgery

These authors reviewed 35 patients presenting with suspected low-grade infection in which biopsy specimens taken at revision surgery. Patients presented with symptoms akin to resistant postoperative frozen shoulder (persistent pain and stiffness, unresponsive to usual treatments).

Positive cultures were identified in 21 cases (60.0%), of which 15 were male patients (71%). Of all patients with low-grade infection, half were male patients between 16 and 35 years of age.

Propionibacterium acnes and coagulase-negative staphylococcus were the most common organisms isolated (81.1% [n = 17] and 23.8% [n = 5], respectively).  The mean time from index surgery to diagnosis for P. acnes –positive cases was 1.2 years (60.3 weeks).

Not all patients with clinically suspected low-grade infections presentedwith positive microbiologic cultures, despite use of enhanced culture methods. The authors suggest that the suspect organisms may be nonculturable on routine culture media or that the suspect organism was in an area of the shoulder not sampled. Of 14 negative culture cases, 9 were treated with early empirical antibiotics (64.3%); 7 patients reported symptomatic improvement (77.8%). Of 5 patients treated with late empirical antibiotics, 4 stated improvement.  Again, the majority of the patients with negative cultures reported improvement in their pain, stiffness, and range of movement after early empirical antibiotic treatment (doxycycline and amoxicillin, n = 8 [88.9%]; gentamicin and teicoplanin, n = 1 [11.1%]). More than half of these patients (n = 6 [66.7%]) stated that their pain completely resolved

Comment: This series of cases points to the complex and uncertain relationship between cultures and clinical findings. Patients without the characteristic signs of infection can have positive cultures. Patients with negative cultures can respond to antibiotics.

For more, see the below:

These authors point out that while as many as 50% of revision shoulder arthroplasties are culture positive, a consistent, clinically useful definition of a "periprosthetic shoulder infection" is lacking. They conducted a systematic review of the published literature with respect to (1) the definition of a "periprosthetic shoulder infection", (2) the pre-operative evaluation for possible infection, and (3) the harvesting and culturing of specimens at the time of surgical revision.

They found a remarkable lack of consistency in the way different authors defined an 'infection', in the way authors evaluated patients with possible infections before surgery and in the way authors obtained and analyzed specimens obtained for culture harvested at the time of the surgical revision of failed shoulder joint replacements.

This inconsistency makes it very difficult to compare different treatment approaches to failed shoulder joint replacements, recognizing that some of them will have substantial bacteria in the joint, the presence of which may go unrecognized until the culture results are finalized at 2 to 3 weeks after surgery.

Comment: It is critically important not to combine, confuse or commingle data from 

"obvious infections" (i.e. those with swelling, redness, drainage, fever, chills, elevated serum markers of inflammation) where the diagnosis of infection is apparent 

with cases of 

"stealth" presentation (i.e. the unexplained onset of pain and stiffness of the shoulder after a 'honeymoon' of good function in which specimens obtained at revision surgery are strongly positive for organisms such as Propionibacterium).

Here's an example of a stealth presentation:

A 50 year old patient presented desiring a ream and run arthroplasty for severe glenohumeral arthritis

After surgery, the shoulder progressively regained comfort and function. Subsequently, however it started to become stiff and painful without obvious explanation. Eight years after his shoulder arthroplasty, the patient returned to the office with no clinical, laboratory, or radiographic evidence of infection.

A single stage revision was performed (soft tissue releases, prothesis exchange) without any evidence of inflammation, joint fluid, loosening, or osteolysis. Five explant and tissue cultures were sent. The patient was discharge on the yellow protocol (Augmentin) until the results of the cultures were final.

The culture results were
Humeral head explant: 3+ Propionibacterium
Humeral stem explant: no growth
Collar membrane: 1+ Propionibacterium
Humeral periosteum: 1+ Propionibacterium
Joint capsule: no growth

At this point the red protocol (IV ceftriaxone) was started and continued for 6 weeks followed by a 6 month course of Augmentin. The patient has a comfortable shoulder and has regained most of the lost shoulder motion.