Correct!
5. All of the above
Common sources for iatrogenic S. aureus infection include venous catheters, urinary catheters, and aspiration – which in our patient might potentially have occurred during sedation for endoscopy(1). It should be noted that although S. aureus does not naturally colonize the urinary tract or typically cause community-acquired urinary tract infections, it is an important cause of catheter-associated UTI in hospitalized patients (2).
Less common iatrogenic sources of S. aureus infection were also considered. Bacteremia with S. aureus and other bacteria have been described as a complication of upper and lower GI endoscopy, and has resulted in serious infections including endocarditis, septic arthritis and brain abscess (3). There are only three reported cases of purulent pericarditis presumed secondary to endoscopy, two of which involved esophageal perforation and neither of which involved S. aureus (4,5).
S. aureus is the most common bacterial contaminant of propofol – used for sedation during the patient’s endoscopy. Such contamination is typically associated with environmental exposure of opened vials or syringes, and have been reported to have led to fatal infections (6). S aureus can also rarely contaminate packed red blood cells (7), and platelet concentrates (8-9) – the latter of which our patient did not receive –causing potentially fatal transfusion-related sepsis. The exact origin of our patient’s MRSA bacteremia was never determined.
The most striking manifestation of this multisystem infection was the presentation and course of purulent pericarditis. The first symptom – pleuritc chest pain – occurred five days before the patient experienced tamponade. The patient never had a documented fever or pericardial rub. When admitted with sepsis, rapid enlargement in her cardiac silhouette over the preceding 48 hours, and EKG changes of pericarditis were not immediately recognized. The admitting team violated a dictum taught by Dr. Tom Bajo (deceased): “never order blood cultures without starting antibiotics first.” However, the patient recovered with emergency pericardiocentesis and adequate antibiotics.
Purulent pericarditis is diagnosed upon pericardiocentesis upon the finding of gross pus in the pericardial space due to infection. It was a relatively common complication of pneumococcal pneumonia in the pre-antibiotic era, but is now typically nosocomial in origin, complicating cardiothoracic surgery or catheter-associated blood stream infections (10-12).
The most common pathogens include S. aureus and oral anaerobes and S. pneumoniae; less commonly Candida spp. and gram negatives including Salmonella. The initial presentation of fever, cough and chest pain is non-specific. Fever – absent in our patient – is said to be virtually universal. About 30% have chest pain which may or may not be pleuritic, and 40% have a pericardial friction rub. The observation in autopsy series that only 50% of cases are diagnosed premortem suggests the diagnosis is difficult to make, perhaps due to absence of the typical signs of pericarditis and attribution of the constitutional symptoms to other infections, as in our case. Disease progression is rapid with a mean duration of three days between first symptoms and hospitalization. About 50% of patients with purulent pericarditis develop tamponade, which due to rapid progression can occur with relatively small volumes of pericardial pus. Treatment should include drainage and antibiotics. The later can be targeted as in our case, or empirical, in which case vancomycin 15-20mg/Kg Q12 hourly, ceftriaxone 2gm daily and fluconazole 400 IV daily (in immunocompromised patients) are suggested (10-12). Ceftaroline and daptomycin are options in patients with vancomycin treatment failure (13).
References