UW Shoulder and Elbow Academy: Propionibacterium - does preoperative systemic antibiotic prophylaxis work?

Preoperative doxycycline does not decolonize Propionibacterium acnes from the skin of the shoulder: a randomized controlled trial

These authors point out that Propionibacterium is frequently cultured in patients undergoing both primary and revision shoulder surgery. They conducted a prospective, randomized controlled trial of male patients undergoing shoulder arthroscopy to evaluate the efficacy and safety of preoperative oral administration of doxycycline in decreasing the colonization of skin around the shoulder by P. acnes.

Patients were randomized to receive oral doxycycline (100 mg twice a day) for 7 days or to the standard of care (no drug). Before skin incision, 2 separate 3-mm punch biopsy specimens were obtained from the sites of the anterior and posterior arthroscopic portals and were sent for culture in anaerobic and aerobic medium held for 13 days.

To serve as a “negative control,” while wearing sterile gloves, 20 consecutive sterile swabs were opened in the operating room, wiped through the air, and sealed in a specimen container. 20% of these cultures were positive.

In the overall cohort (74 patients), 38 patients (51.3%) had at least 1 positive culture for Propionibacterium; 22 patients (29.7%) had positive cultures from both the anterior and the posterior portal sites and 16 patients (21.6%) had positive cultures from just 1 site. All patients with a positive culture from the posterior portal also had a positive culture from the anterior portal. The mean time to isolate Propionibacterium from cultures was 5.9 days (range, 2-9 days).

22 of 37 (59.5%) patients in the no-drug group and 16 of 37 (43.2%) patients in the doxycycline group had at least 1 dermal culture positive for P. acnes (P = .245). The number of shoulders with negative cultures was greater in the Doxyclycline group.

The authors concluded that that a 7-day course of oral doxycycline administration was safe but was only marginally effective in reducing the colonization rate of P. acnes in the dermal layer of the shoulder in male patients.

Comment: Over half of the dermal cultures obtained from the surgically prepared dermis of shoulder skin of male patients were positive for Propionibacterium. This finding is consistent with prior studies showing that skin surface preparation does not eliminate Propionibacterium from the dermis. As the authors point out, although Propionibacterium is usually sensitive to doxycycline, preoperative administration of this antibiotic had only a modest effect in reducing the positive culture rate. This may be a result of the fact that the sebum in sebaceous glands that harbor Propionibacterium is relatively out of reach of the systemic circulation.

The 20% positive culture rate for Propionibacterium in the 'negative controls' is of interest. It is difficult to know how to factor in this rate of positive 'control' cultures when assessing the positive culture rate of the dermal specimens. The authors did used 'semiquantitative methods' for reporting the bacteria load in positive dermal cultures (see below), it would be of interest to know the bacterial load found in the 'negative control' cultures.