To investigate factors pertaining to positive microbiological cultures of pleural fluid post-operatively.
Six hundred fifty eight patients underwent surgery for non-infectious conditions over a 24 month period in a 3 thoracic surgeon unit; 224 (34%) had pleural fluid cultured. Twenty patients (3%) had positive pleural fluid cultures. Retrospective chart review recording length of hospital stay, duration of chest drainage, diagnosis, operation, factors predisposing to and clinical signs of infection, organisms cultured and antibiotic administration.
Of the 20 patients with positive pleural fluid culture, 15 underwent pulmonary resection (10 lobectomies, 4 sublobar); 5 talc pleurodeses. Sixteen had underlying malignancies, 4 benign pathology. There were 14 thoracotomies and 6 VATS procedures. Median time of chest drainage was 11 days (range 2-67) and median hospital stay was 15 days (range 3-133). Positive pleural culture occurred at a median of 7 days (1-14). One patient was diabetic, 5 were on steroids and 11 were hypoalbuminemic pre-operatively. Thirteen patients were febrile when pleural fluid was sent. All patients had elevated white counts and/or pyrexia, except one patient on steroids who was neither. Empirical antibiotic treatment was initiated in 15 patients (with fever or elevated white counts). This was modified according to culture results when appropriate. The other 5 patients received culture-directed antibiotics after these results became available. Microscopy (with Gram staining) only detected organisms in 50% of culture positives. Most common organisms cultured were S. epidermidis (7) followed by Pseudomonas spp. (4) and E. faecalis (4). Of the 30 organisms isolated 15 were skin flora, 8 bowel commensals and 7 respiratory tract organisms.
Pre-operative steroid administration and poor nutritional status (reflected by hypoalbuminemia) seem to be significant risk factors for post-operative pleural sepsis when chest drains remain in situ. These patients warrant close clinical surveillance.
Pleural fluid culture is an investigation with a low yield if sent routinely. It is indicated in patients with pyrexia and/or neutrophilia. Microscopy is unreliable for predicting positive cultures.
L. Tang, None.