The incidence and microbiological profile of surgical site infections following internal fixation of closed and open fractures.

OBJECTIVE: To evaluate the incidence and microbiological profile of surgical site infections (SSIs) associated with internal fixation of fractures and to compare differences in the SSIs observed among patients with closed and open fractures.
METHODS: Retrospective study. Analyzed data included information from all patients who underwent surgery for fixation of closed or open fractures from January 2005 to December 2012 and remained outpatients for at least one year following surgery. Incidence of surgical site infection (SSI) was compared between patients with closed and open infection, as well as polymicrobial infection and infection related to Gram-negative bacilli (GNB). Cumulative antibiograms were performed to describe microbiological profiles.
RESULTS: Overall incidence of SSI was 6%. This incidence was significantly higher among patients with open fractures (14.7%) than among patients with closed fractures (4.2%). The proportions of patients with polymicrobial infections and infections due to GNB were also significantly higher among patients with open fractures. Staphylococcus aureus and coagulase-negative Staphylococcus (CoNS) species were the primary infectious agents isolated from both groups. The overall incidence of MRSA (methicillin-resistant S. aureus) was 72%. A. baumannii was the predominant GNB isolate recovered from patients with open fractures and P. aeruginosa was the most frequent isolate recovered from patients with closed fractures, both exhibited low rates of susceptibility to carbapenems.
CONCLUSIONS: Incidence of SSIs related to the internal fixation of fractures was significantly higher among patients with open fractures, indicating that an open fracture can be a risk factor for infection. Among the bacterial isolates, S. aureus (with a high prevalence of MRSA) and CoNS species were most prevalent. A. baumannii and P. aeruginosa isolates underscored the low rate of susceptibility to carbapenems that was observed in the present study.

Introduction

Surgical site infections (SSIs) associated with internal fixation of fractures are regarded as serious complications. An SSI constitutes a challenge for the entire staff involved in patient's care because it significantly increases recovery time and treatment costs and negatively impacts both functional results and long-term rehabilitation.1, 2 Although incidence of SSIs is expected to be higher in patients with open fractures than in patients with closed fractures, few studies have confirmed this hypothesis.3, 4 Regarding the microbiological profiles of SSIs, Gram-negative bacilli (GNB), most notably Acinetobacter baumannii and Pseudomonas aeruginosa, have been described as having increasingly important roles in these infections, particularly in cases involving high-energy trauma.5, 6, 7, 8

The current study sought to evaluate the incidence and microbiological profile of SSIs associated with internal fixation of fractures and to compare differences in the SSIs observed among patients with closed and open fractures.

Methods

A retrospective study was conducted at the Instituto de Ortopedia e Traumatologia, a tertiary orthopedic academic hospital in São Paulo, Brazil. Analyzed data included information from all patients who underwent surgery for fixation of closed or open fractures from January 2005 to December 2012 and remained outpatients for at least one year following surgery. These data were collected from the database maintained by the institution's Infection Control Board.

In accordance with the institution's protocol, patients with closed fractures who underwent surgery for the internal fixation of their fractures received 24 h of antimicrobial prophylaxis with cefazolin. Patients with type I open fractures according to the Gustilo classification received antimicrobial therapy with cefazolin for 14 days, beginning at their admission. Patients with open fractures of Gustilo types II and III received combination therapy with clindamycin and gentamicin for 14 days, beginning at admission. In addition, patients with open fractures initially had their fractures stabilized by external fixation; internal fixation was subsequently performed after improvement in the soft tissue adjacent to fracture.

Determinations of the incidence of SSIs only included patients who presented with an SSI within 1 year following surgery and were diagnosed based on the criteria established by the Centers for Disease Control and Prevention's National Healthcare Safety Network (CDC-NHSN). To determine the microbiological profile of infections, cumulative antibiogram reports were prepared using established Clinical and Laboratory Standard Institute (CLSI) standards. Only isolates from cultures of bone, deep soft tissue or exudate collected in the operating room after the debridement of devitalized tissues were considered.

Chi-square tests were used to determine how fracture type was associated with the incidences of SSI, polymicrobial infections and GNB infections. These associations were estimated by utilizing bivariate logistic regressions to calculate odds ratios (ORs) and the corresponding 95% confidence intervals. Microbiological findings were only assessed by descriptive analysis.

Results

During the analyzed period, 11,030 patients underwent internal fixation of fractures, including 9143 patients (82.9%) with closed fractures and 1887 patients (17.1%) with open fractures. A total of 664 patients presented with SSIs related to fixation procedures; thus, the overall incidence of infection was 6%. This incidence was significantly higher among patients with open fractures (14.7%) than among patients with closed fractures (4.2%). The proportions of patients with polymicrobial infections and infections due to GNB were also significantly higher among patients with open fractures (Table 1).

Table 1

Patient population of the study.

	Total	Closed fractures	Open fractures	Comparison between the groups with closed and open fractures
Patients	11,030	9143	1887	–
Patients with SSIs	664 (6%)	386 (4.2%)	278 (14.7%)	p < 0.001
Patients with polymicrobial infections	199 (1.8%)	90 (1%)	109 (5.8%)	p < 0.001
Patients with GNB infections	293 (2.65%)	130 (1.42%)	163 (8.64%)	p < 0.001

The microbiological findings indicated that a total of 529 infection-associated bacterial isolates were recovered, including 357 isolates from patients with open fractures and 172 isolates from patients with closed fractures. In both groups, the predominant isolates were Gram-positive cocci, which accounted for 53% of the observed infections. GNB accounted for 45% of the isolates; both patient groups exhibited a similar incidence of GNB, although a greater absolute number of isolates were recovered from patients with open fractures than from patients with closed fractures.

Staphylococcus aureus and coagulase-negative Staphylococcus (CoNS) species were the primary infectious agents isolated from both groups. The overall incidence of MRSA (methicillin-resistant S. aureus) was 72%; this incidence was 75% among patients with open fractures and 66% among patients with closed fractures. For the group of patients with open fractures, in addition to CoNS species, Enterococcus spp. were the second most agent that was isolated; 76% of the Enterococcus isolates exhibited susceptibility to vancomycin.

A. baumannii was the predominant GNB isolate recovered from patients with open fractures. P. aeruginosa was the most frequent isolate recovered from patients with closed fractures. Both A. baumannii and P. aeruginosa exhibited low rates of susceptibility to carbapenems (susceptibilities to imipenem of 57% and 47%, respectively). Anaerobic bacteria and fungi accounted for 2% of the isolates. Table 2 summarizes the microbiological findings described in this study.

Table 2

Comparison of the main microbiological findings in the groups of patients with closed and open fractures.

Closed fractures	Open fractures
Total number of isolates
357	172
Primary agents
S. aureus (56 isolates – 33%) 66% MRSA	S. aureus (83 isolates – 23%)75% MRSA
CoNS species (24 isolates – 14%)	CoNS species (51 isolates – 14%)Enterococcus spp. (51 isolates – 14%)76% sensitivity to vancomycin
P. aeruginosa (18 isolates – 10%) 47% sensitivity to carbapenems	A. baumannii (44 isolates – 12%)57% sensitivity to imipenem

Discussion

The incidence of SSI was significantly higher in patients with open fractures. Although this finding is to be expected due to the high degree of contamination observed in these high-energy injuries, few studies have proven the higher incidence of SSI in this population, and none of them featured a sample as large as that of the present study. There was a predominance of GNB infections and polymicrobial infections in the group of patients with open fractures. Other studies, conducted mainly in soldiers stationed in the Middle East, corroborate these findings.1, 5, 6, 7

Among all agents that were isolated, S. aureus and CoNS species were the primary causative agents of infection; MRSA was highly prevalent. Among patients with open fractures, Enterococcus spp. were also important causative agents of SSIs; only 76% of these Enterococcus isolates were susceptible to vancomycin. A. baumannii and P. aeruginosa were the main GNB isolated from SSIs in the group of patients with open fractures and in the group with closed fractures, respectively. Both of these species exhibited low rates of susceptibility to antimicrobial agents, including carbapenems. These findings are consistent with the results of Torbert et al., who previously analyzed 214 cases of SSIs related to fracture fixation. Other studies that analyzed severe open fractures in combat environments have also found that A. baumannii was highly prevalent.5, 6, 7 A high prevalence of P. aeruginosa has also been reported by a study that evaluated the infectious complications of traumatic amputations.

Conclusion

The incidence of SSIs related to the internal fixation of fractures was significantly higher among patients with open fractures, indicating that an open fracture can be a risk factor for infection. The current study also indicated that there were a significantly elevated number of polymicrobial and GNB infections among patients with prior open fractures. Further studies are needed to evaluate potential variables that could be relevant to these findings.

Among the bacterial isolates, S. aureus (with a high prevalence of MRSA) and CoNS species were most prevalent. Enterococcus spp. (with a low rate of sensitivity to vancomycin) and A. baumannii were also prevalent among the isolates recovered from patients with open fractures. In addition, P. aeruginosa was one of the more significant isolates recovered from patients with closed fractures. The A. baumannii and P. aeruginosa isolates underscored the low rate of susceptibility to carbapenems that was observed in the present study.

Conflicts of interest

The authors declare no conflicts of interest.