Standardized practice is associated with low rate of surgical site infection in orthopaedic trauma

1. Introduction

Staphylococcus aureus colonization is a known independent risk factor for the development of surgical site infection (SSI). It is the most common causative pathogen of SSI in orthopaedic surgery, and its incidence has been increasing in recent years.^{1, 2, 3, 4, 5} The organism's ability to form a biofilm resistant to antibiotic treatment is of particular importance in orthopaedics due to the use of metal implants. Furthermore, liberal use of broad-spectrum antibiotics over time has resulted in increased numbers of antibiotic-resistant strains of pathogens, such as Methicillin-resistant Staphylococcus aureus (MRSA), of particular concern after joint arthroplasty and orthopaedic trauma surgery.

In an effort to reduce SSIs, particularly MRSA, many investigators have explored standardized protocols aimed at eliminating or minimizing bacterial colonization in patients preoperatively. Although there remains debate regarding its efficacy, it is generally recommended that patient's undergoing joint arthroplasty procedures use a pre-admission chlorhexidine wash prior to surgery in an effort to minimize SSI as a part of a standard preoperative protocol as the effect of chlorhexidine is cumulative and increases with multiple administrations and duration of use.^{6, 7, 8} Similarly, preoperative eradication of nasal MRSA colonization using mupirocin nasal ointment is performed in the preoperative setting as part of many standardized total joint arthroplasty protocols.^{9, 10, 11} However, both of these pre-operative measures to reduce infection can potentially be impractical as part of a standardized protocol among orthopaedic trauma patients, due to the frequent emergent or urgent nature of trauma surgery. A modified approach to minimizing the risk of infection in such patients involves a preoperative scrub protocol employed immediately prior to the final preoperative skin preparation. Anecdotally, such a practice is employed by surgeons at a number of institutions, but to our knowledge the effects of such a protocol have not previously been reported. In addition to preoperative bacterial decolonization, minimizing deep SSIs has also been associated with standardization of preoperative prophylactic antibiotic administration,^{12, 13, 14, 15} methods of venous thromboembolism (VTE) prophylaxis,^16,17 and wound closure techniques.¹⁸

Standardizing care has been shown to decrease the rate of infection among numerous surgical procedures, including among colorectal and neurosurgical procedures.^{19, 20, 21, 22} Given this, the orthopaedic trauma surgeons at our institution have sought to standardize care as much as possible despite training at different residencies and fellowship programs. Through a series of informal and formal meetings, we have developed uniform protocols for skin preparation, antibiotic prophylaxis, wound closure, and VTE prophylaxis. All protocols are potentially modifiable on a patient-by-patient basis, with the highest rate of variation being for venous thromboembolism prophylaxis targeted to a specific clinical scenario. The entire orthopaedic trauma team including nurses, midlevel providers, and residents are educated on these protocols. Further, feedback from these individuals aided in the design of the protocols by the attending surgeons.

The aim of this study was to determine the rate of deep SSI following open surgery for orthopaedic trauma in such a standardized practice setting. Our hypothesis was that by employing standardized practices in the treatment of these relatively high-risk patients, we would be able to achieve an infection rate more similar to elective joint arthroplasty than to historical rates of infection among orthopaedic trauma patients.

2. Methods

After institutional review board approval, we prospectively studied all adults undergoing open orthopaedic trauma surgery at an urban level 1 trauma center between January 1, 2014 and December 31, 2014. The final cohort for the review consisted of 468 trauma cases after exclusions for pre-existing infection (18), non-traumatologist surgeon (14), percutaneous procedure (11), unknown MRSA status (22), and insufficient follow-up (99). Standardized protocols were in place throughout the study period for antibiotic prophylaxis, skin preparation, wound closure, and VTE prophylaxis. All patients underwent regional decolonization consisting of a stepwise chlorhexidine mechanical scrub followed by alcohol wipe of the operative extremity prior to final preparation and draping in the operating room. Patients received cefazolin (weight-based dosing) within 1 h prior to incision unless allergy or MRSA status triggered the use of vancomycin. Standard wound closure employed the use of interrupted/buried 2-0 monocryl sutures for subcutaneous closure followed by interrupted, vertical mattress 3–0 nylon sutures for skin closure. Staples were not utilized. Wounds associated with type I, II and IIIA open fractures were closed by the treating orthopaedic surgeon as soon as the wound was deemed sufficiently debrided. Wounds that were not amenable to closure and/or definitive fixation were treated with external fixation as needed and referred to the plastic surgery team as soon as the need for flap coverage was determined (typically referred within 3 days of injury) and flap coverage in these cases was performed as soon as possible and as possible in the same setting as definitive fixation of the fracture.

Patients were evaluated clinically during their index hospitalization, as well as in standard clinic appointments at 2, 6, 12 and 24 weeks postoperatively. Table 1 details the standardized VTE prophylaxis protocol depending on the procedure performed. Patient, injury, and surgical characteristics were recorded, and patients were followed for diagnosis of deep surgical site infection.

Table 1.

Treatment protocol for the prevention of VTE (venous thromboembolism) in adult orthopaedic trauma patients.

	Pharmacologic Prophylaxis	Mechanical Prophylaxis
Hip/Femur/Tibial Shaft, Plateau, Pilon Fracture	-Enoxaparin 40 mg daily x 14 days -Upon completion of Enoxaparin- Begin 162 mg aspirin x14 days	59
Pelvis/Acetabulum Fracture	-Coumadin x4 weeks (Target INR = 2.0–2.5)	21.8
Ankle/Foot Fractures	-Enoxaparin 40 mg daily while in house -ASA 162 mg daily upon discharge until discontinued by surgeon	Male

Features evaluated included age, smoking status, presence of diabetes mellitus, body mass index (BMI), American Society of Anesthesiologists physical status classification (ASA), MRSA nasal colonization on admission, and open vs closed fracture type (Gustilo Anderson classification). All patient characteristics were extracted from the electronic medical record, including operating room record. The primary outcome was deep surgical site infection within 90 days of injury, as defined by Centers for Disease Control criteria (Table 2), which was confirmed through examination of the medical record, including operative notes, clinic notes, and microbiology results. A power analysis required 388 patients to demonstrated our population had an infection rate less than 1% and that this was significantly lower than rates reported in the literature assuming an alpha of 0.05 and power of 85%. Data were analyzed using Fisher's exact testing with statistical significance set at p < 0.05.

Table 2.

Centers for disease control criteria for deep surgical site infection.

Surgical site infection occurring within 90 days, involving the deep soft tissue (muscle and fascial layers) as well as one of the following:
1. Purulent drainage from deep incision
2. Wound dehiscence or deliberate opening by the surgeon (culture positive or not cultured) in the setting of fever (>38 °C) and/or localized pain or tenderness (negative culture does not fulfill this criterion)
3. Abscess or other evidence of infection involving the deep incision found on direct examination, during invasive procedure, or by histopathologic examination or imaging
4. Diagnosis of a deep incisional SSI by a surgeon

3. Results

Deep surgical site infection occurred in 4 of 468 trauma procedures (0.85%). Of the four infections, one was culture positive for MRSA, one was culture positive for MSSA, one was culture positive for serratia marcescens, and one was culture positive for multiple non-staphylococcus species. The characteristics of the patients with deep surgical site infection within 90 days can be seen in Table 3. Although they were excluded from our final results, none of the 22 patients with unknown MRSA status developed an infection.

Table 3.

Comorbidities and fracture characteristics of patients with deep surgical site infection.

	Infection #1	Infection #2	Infection #3	Infection #4
Age	51	59	23	85
BMI	33.5	21.8	30.4	27.7
Gender	Male	Male	Male	Female
Smoker	No	No	Yes	No
Diabetes mellitus	No	No	No	No
MRSA Status	Negative	Negative	Negative	Positive
ASA	2	1	2	4
Surgery Length (min)	220	207	123	92
Length of Stay (days)	2	7	6	10
Fracture Type	Open Type IIIB	Open Type IIIA	Open Type IIIA	Closed
Bacteria	MSSA	Clostridium sporogenes Ctirogacter freundii Clostridium butyricum Enterococcus faecalis	S. marcenscens	MRSA

Procedures complicated by deep surgical site infection included 4 of 394 non-diabetic patients (1.03%) and 0 of 74 diabetic patients (0%), 1 of 115 active smokers (0.87%), 3 of 353 nonsmokers (0.85%), 2 of 285 obese patients (BMI ≥30) (0.70%), 2 of 159 non-obese patients (1.26%), 1 of 217 patients with an ASA score ≥3 (0.46%) and 3 of 251 patients with an ASA score <3 (1.20%), 1 of 135 patients age > 65 (0.74%), 3 of 333 patients age <65 (0.90%), 1 of 13 patients with MRSA nasal colonization (7.69%), 3 of 455 patients without MRSA nasal colonization (0.66%), 3 of 51 open fracture (5.88%), and 1 of 417 closed fractures (0.24%). There were 24 revision cases for nonunion and one for malunion, none of which were complicated by deep SSI. Of these factors, only open versus closed fracture demonstrated a statistically significant difference in the rate of deep surgical site (P < 0.05, detailed in Table 4).

Table 4.

Comparison of comorbidities and fracture type between infected and uninfected patients^a.

Patient/Fracture Characteristics	Uninfected (N = 464)	Infected (N = 4)	P
Age ≥65, n (%)	134	1	1.00
Smoking, n (%)	114	1	1.00
Diabetes mellitus, n (%)	74	0	1.00
Obesity (BMI ≥30), n (%)b	157	2	0.62
ASA ≥3, n (%)	216	1	0.62
MRSA +, n (%)	12	1	0.11
Open Fracture, n (%)	48	3	<0.01***

***Statistical significance was set at p < 0.05.

^aAll patient/fracture characteristics were compared by Fisher's exact testing.

^b24 patients did not have enough information recorded during their hospital admission to determine BMI.

The 51 open fractures consisted of six type I fractures, eight type II fractures, 22 Type IIIA fractures, 13 Type IIIB fractures, and two type IIIC fractures. All three of the infections among open fractures occurred in type III injuries (two of 22 type IIIA fractures and one of 13 type IIIB fractures). Of the 13 MRSA positive patients, nine did not receive antibiotic prophylaxis adequately covering MRSA, one of which developed a deep surgical site infection within 90 days of the procedure. The infection in the MRSA positive patient was culture positive for MRSA. None of the 468 trauma procedures had any complications (i.e. skin reaction or allergy) related to the regional decolonization protocol.

4. Discussion

The purpose of the study was to determine the rate of deep SSI for open orthopaedic trauma procedures performed in a standardized practice setting. Further, we sought to contrast our results with previously reported rates of infection for open orthopaedic trauma surgery and elective joint arthroplasty procedures. We found a rate of deep SSI among orthopaedic trauma patients (0.85%) more consistent with previous reports for elective joint arthroplasty than to previous reports for orthopaedic trauma. Further analysis demonstrated only open versus closed fracture demonstrated a statistically significant difference in the rate of deep surgical site infection within 90 days. Taking this into account, the infection rate among 417 closed injuries was (0.24%).

Importantly, this population included a subset of MRSA positive patients, and the SSI remained low despite limitations in providing antibiotic prophylactic coverage for MRSA in a majority of MRSA positive patients. Based on nasal swab collection time, test result time, and preoperative antibiotic time, the primary barrier to providing prophylactic antibiotic coverage against MRSA, was due to emergent and urgent surgeries being performed prior to the MRSA nasal carriage screening test result being available. The transport time involved in getting the swab to the lab for testing was a major barrier to obtaining a preoperative MRSA screen result. Additionally, our results err toward reporting a higher infection rate by excluding patients who had an unknown MRSA status (none of whom developed an infection).

Our overall rate of infection compares favorably to a previously reported range of 2.5%–4.2% in orthopaedic trauma, with MRSA nasal carriage associated with even higher rates of infection in previous literature.^{1,11,26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37} Notably, we report an infection rate for orthopaedic trauma patients (including a subset of open fractures) that aligns more closely with rates reported for elective joint arthroplasty surgery (Fig. 1).^{18, 19, 20} These results indicate the potential for significant benefit in standardizing practice among orthopaedic trauma partners. To our knowledge, it is also the first study to evaluate the preoperative scrub technique prior to skin preparation. This simple, low-cost, low-risk practice deserves further attention.

Fig. 1.

The infection rate demonstrated in this study among open orthopaedic trauma procedures in the setting of regional decolonization is more consistent with elective total joint arthroplasty than what has been previously reported for orthopaedic trauma.

Surgical site infection is an ongoing concern in orthopaedic surgery. It is associated with worsened patient outcomes, a significant increase in length of hospital stay, re-hospitalization rate, and overall healthcare costs.^23,24 A known independent risk factor for SSI is Staphylococcus colonization, particularly MRSA colonization, which has generally been increasing in prevalence among orthopaedic patients.^{1, 2, 3, 4, 5,11,25} Previous treatment strategies to address bacterial colonization have involved pre-admission screening and treatment regimens in the elective orthopaedic patient, which is less feasible in the orthopaedic trauma patient due to the often emergent or urgent nature of trauma surgery. We document the safe use of a preoperative scrub technique in orthopaedic trauma.

Patient characteristics such as smoking, obesity, elevated ASA score, increased age, and MRSA colonization are well established risk factors for SSI in the orthopaedic patient.^{2,11,23,32,33} However, the relatively low rate of SSI observed in this study was found despite a significant prevalence of smoking, obesity and other described risk factors. These results support that the risk associated with these variables may be at least partially mitigated through standardization of care.

Three of the four SSIs occurred among type III open fractures. These findings agree with previous research demonstrating SSI to be strongly associated with open fracture classification.^1,34,35 Additionally, it should be noted that our overall rate of infection among a population of both open and closed fractures (0.85%) compared favorably to rates previously reported for closed fractures alone.^36,37

We recognize the limitations related to this study. Most notably, causality cannot be proven with this observational cohort design. The causes of SSI are multifactorial and controlling for all contributing variables is difficult. Future research into the described preoperative regional decolonization protocol or the other standardized practices employed would benefit from comparative studies to more conclusively determine the efficacy of these protocols in minimizing the risk of infection. Within our institution, we do not feel that equipoise exists to allow for a comparative study given the low rate of infection we are achieving with our current practice.

Our findings demonstrate standardized orthopaedic trauma care to be associated with a very low rate of postoperative infection. Specifically, we found a chlorhexidine and alcohol mechanical scrub of the operative extremity prior to final skin preparation to be a low-risk, low-cost, efficient process that can be successfully incorporated into multi-modal efforts to minimize surgical site infection in orthopaedic trauma patients. Further research is warranted on this topic as standardization of care has the potential to minimize other forms of complications and to reduce medical error through decreased variability.

Funding

No authors have received grant support or research funding for this study. Additionally, no authors have any proprietary interests as it relates to any materials described in the study.