Extended Antibiotic Coverage in the Management of Type II Open Fractures

Abstract

Background:

Responsible antibiotic stewardship requires surgeons treating open fractures to use the narrowest appropriate antibiotic coverage possible to prevent infection. Because inter-observer agreement about the application of the Gustilo–Anderson open fracture classification is moderate at best, antibiotic selection can be overly aggressive. The purpose of this study was to evaluate the outcomes of Type II open fractures treated with gram-positive coverage only (GP) versus broad-spectrum antibiotic coverage (BS) with piperacillin–tazobactam (PT).

Methods:

A retrospective review of all Type II open fractures was performed at a single Level one trauma center over a 5-year period (2013–2017). All patients received prophylactic antibiotics on arrival on the basis of the best judgment of classification by the house officer on call. The final Gustilo–Anderson open fracture classification was assigned intra-operatively by the operating surgeon. Two groups were created, a GP antibiotic group (cefazolin and/or clindamycin) and a BS group (PT). A minimum of 3-month follow-up was required for inclusion. Patient demographics, cost of treatment, fracture-related infection (FRI) rates, and infecting bacteria were assessed.

Results:

The GP group contained 70 open fractures and the BS group contained 74 open fractures. Between the groups, there were no differences in age, sex, race, Body Mass Index, American Society of Anesthesiologists Class, or smoking status. There were no statistical differences in Injury Severity Score (ISS), fracture location, fixation method, or rates of staged management with external fixation. There was no difference in FRI rate between the GP and BS groups (8.6% versus 10.8%; p = 0.78). The bacteria responsible for FRI were similar in the GP and BS groups. The hospital charge for PT was 4.39 × the cost of cefazolin.

Conclusions:

The use of BS coverage in Type II open fractures does not result in a lower infection rate and adds significant cost to patient care. These data support the use of a GP-only antibiotic regimen for Type II open fractures.

Open fractures represent a significant source of morbidity in the adult trauma population [1]. Many of these injuries are complicated by gross contamination, soft tissue injury, and bone loss. Open fractures increase the risk of fracture-related infection (FRI) significantly, leading to multiple procedures, additional morbidity from infection treatment, increased length of stay, and higher medical costs [2,3].

Recent literature demonstrates that early administration of antibiotics is an important factor in reducing FRI risk [4 –6]; however, debate remains over the preferred antibiotic [7]. Gustilo and Anderson introduced an open fracture classification system based on fracture severity that has largely been used to direct antibiotic coverage over the last 40 years. A first-generation cephalosporin was recommended for Type I and Type II open fractures with the addition of an aminoglycoside for gram-negative coverage in Type III open fractures [8]. However, because of the concern about the harmful side effects of aminoglycosides, alternative antibiotics (e.g., piperacillin–tazobactam (PT), ceftriaxone) have been proposed as substitutes for broad-spectrum (BS) coverage with less potential for patient morbidity [9 –11].

Current East Guidelines for appropriate open fracture antibiotic prophylaxis rely on accurate classification of the fracture using the Gustilo–Anderson system, but studies have shown relatively low inter-observer reliability among physicians when using this classification system [12 –14]. Furthermore, final fracture classification cannot be made definitively until intra-operative assessment of the injury has occurred [15]. As a result, many open fractures may be classified inappropriately and treated with BS antibiotic coverage in an effort to err on the side of caution in borderline Type II/Type III fractures. This has led to poor physician compliance with the East Guidelines and inappropriate use of BS antibiotics in misclassified Types I and II open fractures [16].

Broad-spectrum antibiotics have a potential for more harmful systemic side effects, greater cost, and increased 3antibiotic resistance without providing any proved reduction in infection risk in patients with Type I or Type II open fractures [4,7]. Consequently, it is beneficial to the patient and healthcare system to provide the narrowest effective antibiotic coverage when possible.

The purpose of this study was to evaluate the outcomes of Type II open fractures treated with prophylactic GP versus BS antibiotic coverage. The primary outcome was rate of FRI and the secondary outcome was the cost of treatment. The authors hypothesized that BS antibiotics do not lead to a significant reduction in infection rate and are associated with higher costs.

Patients and Methods

After Institutional Review Board approval, a retrospective review of all open fractures was performed at a single Level one trauma center over a 5-year period (2013–2017). Open fractures were identified using CPT code 11012. Hand and forefoot open fractures and patients without adequate follow-up were excluded from the study.

On arrival in the emergency department, all patients received prophylactic antibiotics based on the judgment of the Gustilo–Anderson classification by the house officer on call. Final open fracture classification was assigned intra-operatively by the operating surgeon. For the purposes of this study, all Type I and Type III fractures were excluded. At our institution, patients with Type I and II open fractures receive cefazolin (or clindamycin when allergies to cephalosporins are present) and those with Type III open fractures receive PT. Antibiotics were given on arrival to the emergency room and continued for a minimum of 24 hours after surgical debridement. In cases where a secondary debridement was undertaken at the time of definitive fixation (e.g., because of soil contamination, peri-articular fracture treated in a staged fashion), pre-operative antibiotics and 24 hours of post-operative antibiotics were administered.

Two groups were created: A GP coverage group (cefazolin or clindamycin) and a BS coverage group. Only antibiotics given within 48 hours of arrival were used to delineate the groups in the study. A minimum of 3-month follow-up was required for inclusion.

Patient demographics assessed included age, sex, race, Body Mass Index (BMI), smoking status, and American Society of Anesthesiologists Physical Status Classification (ASA Class). Injury characteristics included Injury Severity Score (ISS), fracture location, mechanism of injury, and gross soil contamination of the fracture. Treatment variables evaluated included time from presentation to first dose of antibiotic, fixation method, staging with external fixation, primary versus delayed closure, number of surgical debridements before closure, and patient financial charges for treatment. Outcome variables assessed were length of stay (LOS), subsequent FRI requiring return to the operating room, time to FRI, and intra-operative cultures from the site of infection. An FRI was limited to patients requiring a return to the operating room because retrospective chart review was unreliable for eliciting any potential infections treated without surgery.

All data were stored in a Microsoft Excel file, and all data analysis was performed using SAS/STAT^® software v. 9.4. Categorical variables were assessed with the χ² or Fisher exact test; continuous variables were assessed with the Student t-test or analysis of variance. A p value <0.05 was considered significant.

Results

One hundred and forty-four Type II open fractures were identified and met the inclusion criteria, with 74 receiving only GP coverage and 70 receiving BS coverage. There were no differences between the GP and BS groups in terms of age, sex, race, BMI, smoking status, or ASA Class, as shown in Table 1.

Table 1.

Patient Demographics

	Gram-positive coverage only (n = 70)	Broad-spectrum coverage (n = 74)	p
Age (SD; min–max)	43.1 (17.4, 17.0–97.0)	38.5 (15.6, 17.0–86.0)	0.095
Male/female (%)	43 (61.4)/27 (38.6)	43 (58.1)/31 (41.9)	0.735
Ethnicity (%)			0.144
Caucasian	42 (60.0)	36 (48.7)
African American	18 (25.7)	32 (43.2)
Hispanic	4 (5.7)	3 (4.1)
Other	6 (8.6)	3 (4.1)
BMI (SD; min–max)	28.5 (7.4, 18.0–62.2)	29.8 (9.0, 18.7–73.9)	0.339
Smoking status (%)			0.241
Never	44 (62.9)	39 (52.7)
Current	26 (37.1)	35 (47.3)
ASA class (%)	N = 64	N = 71	0.238
1	3 (4.7)	3 (4.2)
2	23 (35.9)	26 (36.6)
3	38 (59.4)	29 (40.8)
4	4 (6.3)	11 (15.5)
5	0	1 (1.4)

ASA = American Society of Anesthesiologists; BMI = Body Mass Index; SD = standard deviation.

Table 2 summarizes the injury and treatment variables in the coverage groups. There was no difference in fracture location. Tibia fracture was most common in both groups (41.4% GP and 62.2% BS). The groups did differ in mechanism of injury, with the GP group containing a higher percentage of patients with falls (p = 0.039) and the BS group containing a higher percentage of victims of motor vehicle/motorcycle accidents (p = 0.039). However, there was no significant difference in ISS scores between the two groups, nor was there a significant difference in the percentage with soil contamination of the fracture.

Table 2.

Injury Characteristics and Treatment

	Gram-positive coverage (GP) (n = 70)	Broad-spectrum coverage (BS) (n = 74)	p
Injury Severity Score (SD; min–max)	13.9 (9.2, 1.0–50.0)	16.9 (9.1, 4.0–43.0)	0.058
Fracture location (%)			0.123
Humerus	5 (7.1)	5 (6.8)
Radius	9 (12.9)	5 (6.8)
Ulna	5 (7.1)	0
“Both Bones”	1 (1.4)	2 (2.7)
Femur	13 (18.6)	8 (10.8)
Tibia	29 (41.4)	46 (62.2)
Bimalleolar	2 (2.9)	1 (1.4)
Trimalleolar	0	2 (2.7)
Talus	2 (2.9)	4 (5.4)
Calcaneus	0	1 (1.4)
Other	4 (5.7)	0
Mechanism of Injury (%)			0.039
Motor vehicle accident	29 (41.4)	48 (64.9)
Motorcycle accident	13 (18.6)	12 (16.2)
Pedestrian vs. automobile	3 (4.3)	3 (4.1)
Fall	18 (25.7)	8 (10.8)
Other	7 (10.0)	3 (4.1)
Soil contamination (%)	13 (18.6)	14 (18.9)	1.000
Time to antibiotics (h) (SD; min–max)	2.7 (2.9, 0.0–11.9)	1.9 (2.8, 0.0–12.1)	0.129
Fixation method (%)			0.206
None	6 (8.6)	2 (2.7)
External only	1 (1.4)	3 (4.1)
Internal	63 (90.0)	69 (93.2)
Staged with external (%)	4 (5.7)	12 (16.2)	0.063
Wound closed primarily (%)	67 (95.7)	67 (90.5)	0.328
Debridements before final closure (SD; min–max)	1.3 (0.6, 1.0–3.0)	1.7 (1.0, 1.0–5.0)	0.007

Statistically significant values are in boldface type.

SD = standard deviation.

There were no differences between the groups in time to antibiotic initiation or fixation method, with more than 90% of patients in both groups receiving open reduction and internal fixation. Only 16.2% of the patients in the BS group and 5.7% in the GP group received staged treatment with external fixation (p = 0.063). More than 90% of open fracture wounds in both groups were closed primarily, with the remainder receiving delayed primary closure at the time of definitive closure. The BS group averaged more debridements (including debridement at the time of fixation) (1.7) before closure than the GP group (1.3) (p = 0.007).

Table 3 shows the outcomes for the GP and BS coverage groups. Patients receiving BS coverage had hospital stays of 11.0 days on average, whereas patients receiving GP coverage stayed for an average of 7.7 days (p = 0.074). There was no difference in FRI rate between the GP and BS groups (8.6% versus 10.8%; p = 0.78). In a subset analysis of tibia fractures, there was no difference in FRI between groups (10.3% GP versus 15.2% BS; p = 0.73). The average time to onset of infection was 130 days in the GP group and 126 days in the BS group. Finally, the hospital charges for each antibiotic agent are shown in Table 4. The charge for four doses of 2 g of cefazolin was $73.92, whereas the hospital charge for four doses of 3.375 g of PT was $324.24, 4.39 × the cost of cefazolin.

Table 3.

Mean Length of Stay and Fracture-Related Infection (FRI) Rates

	Gram-positive coverage (n = 70)	Broad-spectrum c^{overage (n = 74)}	p
Length of stay (d) (SD; min–max)	7.7 (11.8, 1.5–88.7)	11.0 (10.7, 1.0–51.9)	0.074
FRI requiring return to operating room (%)	6 (8.6)	8 (10.8)	0.781
FRI in tibia fractures (%)	3 (10.3)	7 (15.2)	0.732
Time from fixation to FRI (SD [d]; min–max)	130.3 (157.0, 8–424)	125.9 (107.1, 25–319)	0.949

SD = standard deviation.

Table 4.

Hospital Charge for Antibiotic Agents

	Prophylactic dose	Price of 4 doses
Cefazolin	2 g × 4 doses	$73.92
Piperacillin–tazobactam	3.375 g × 4 doses	$324.24

With respect to intra-operative culture results, there were no differences in rates of polymicrobial infection (p = 0.608), gram-positive bacteria (p = 0.385), gram-negative bacteria (p = 0.608), MRSA (p = 0.487), or multidrug resistant (MDR) organisms (p = 1.000) (Table 5). The GP FRI group had 40.0% MDR bacteria and the BS FRI group had 37.5%.

Table 5.

Comparison of Positive Intra-Operative Cultures of Fracture-Related Infection

	Gram-positive coverage	Broad spectrum coverage	p
No. of positive cultures	5	8
Polymicrobial (%)	2 (40.0)	2 (25.0)	0.608
Gram positive (%)	4 (80.0)	8 (100)	0.385
Gram negative (%)	2 (40.0)	2 (25.0)	0.608
Methicillin-resistant Staphylococcus aureus (%)	0	2 (25.0)	0.487
Multi–drug-resistant organism (%)	2 (40.0)	3 (37.5)	1.000

Discussion

These data showed the use of BS antibiotics compared with GP antibiotics in patients with Type II open fractures provided no advantage in decreasing the infection rate. Additionally, among patients who developed an FRI, there was no correlation between the cultured surgical site bacteria and the prophylactic antimicrobial coverage the patient received.

In 1984, Gustilo and Anderson reported 77% of cultures from open fractures grew gram-negative bacteria, and from these data concluded that coverage of these organisms would be beneficial in high-grade open fractures. Additional studies by Robinson and Patzakis supported recommendations for BS coverage in high-energy open fractures [6,17]. However, subsequent review of these studies has led to debate over their reliability and the need for gram-negative coverage in these fractures [7,18].

To our knowledge, no study has evaluated the effectiveness of BS antibiotic coverage exclusively in Type II open fractures, although several studies have assessed this coverage in Type III fractures and report mixed results. Bankhead-Kendal et al. reviewed 126 Type III open fractures and showed no difference in the rate of infection between their groups who received either a cephalosporin or a cephalosporin and an aminoglycoside [19]. Redfern et al. reviewed 72 patients with Type III open fractures who received either cefazolin and gentamicin or PT and found similar rates of infection at one year [10]. Shawar et al. reviewed 85 patients with Type III open fractures who received either clindamycin/cephalosporin with tobramycin versus only PT and found a higher incidence of FRI in the tobramycin-based therapy group [17].

Although each of these institutions had differing antibiotic protocols, it is clear surgeons have some interest in shifting antibiotic coverage in open fractures away from aminoglycosides. In 2006, the Surgical Infection Society performed an extensive review of the conflicting data and reported insufficient evidence to support use of BS antimicrobial coverage in open fractures [7] and highlighted the importance of antibiotic stewardship in preventing bacterial resistance. We hope that these data continue to help drive our specialty to formulate high-quality prospective studies to understand better the need for and use of BS antibiotic coverage in open fractures.

Most gram-negative infections after open fractures are thought to be caused by hospital-acquired bacteria, which can be MDR [20]. However, the over-utilization of broad antibiotic coverage leads to a higher rate of antimicrobial resistance [21]. In this study, we found no difference between the groups in the rates of gram-negative or polymicrobial infections, with nearly equal percentages of MDR bacteria in each antibiotic group, suggesting that BS coverage did not provide any additional benefit in preventing gram-negative, polymicrobial, or MDR infections in Type II open fractures. As we move forward as a specialty within the global healthcare economy, it is incumbent on us as good antibiotic stewards to promote appropriately narrow but effective open fracture antibiotic coverage. This is one area where we can play a role in decreasing rates of MDR across all of medicine.

Our study also found a large difference in the cost of administrating an antibiotic (cefazolin) for gram-positive infections versus a BS antibiotic (PT). The latter was 4.39 × the cost of cefazolin. In addition to antimicrobial stewardship to help prevent resistance, we also have a fiscal responsibility as physicians to be to be financial stewards, especially in an era of increasing financial constraints on healthcare systems.

There are several limitations to our study. First, it was retrospective and therefore subject to potential bias. The BS group had more debridements prior to final coverage than the GP group; however, there was no difference in the observed gross contamination between the groups. This difference is likely attributable to a higher number of the BS fractures being treated in a staged fashion with external fixation and subsequent return to the operating room for repeat debridement and definitive fixation. We acknowledge that FRI can occur more than 3 months after surgery, but on the basis of previous studies, 80%–91% of FRI occur within the first 90 days after surgery, with several large retrospective studies demonstrating a mean time of <30 days [22 –24]. Additionally, this exclusion criterion was applied equally and should not introduce bias into the results. We also believe that there is no current standard of care for antibiotic coverage in Type III open fractures. Gustilo and Anderson made the recommendation of a first-generation cephalosporin and an aminoglycoside, but increasingly, surgeons are shifting away from the use of aminoglycosides because of their side effect profile [10,25]. Additionally, we recognize the current recommendation is only for gram-positive coverage of Type II fractures. We assume the BS group of Type II fractures in our study is secondary to the house officer on call erring on the side of “over-calling” injury severity while evaluating the patient in the emergency department and trauma bay. It is the responsibility of our house officers to recommend antibiotics in a timely fashion based on their assessment in the trauma bay [5,6]. In cases where there is uncertainty, house officers may misclassify Type II fractures as Type III and administer BS antibiotic coverage based on the assumption that broader coverage is a more effective way to prevent infection. This discrepancy in misclassifying a Type II as a Type III fracture is indicative of the challenges of the Gustilo–Anderson classification's inter-observer reliability in classifying open fractures [12,13]. This further highlights the need to understand the role of BS antibiotics in open fractures.

Conclusion

In conclusion, the use of BS antibiotic coverage in Type II open fractures adds significant cost to patient care without a clear benefit and does not decrease the infection rate. We support responsible financial and antimicrobial stewardship and the current recommendation for the use of GP rather than BS antibiotics in the prophylactic treatment of Type II open fractures.

Funding Information

Clay Spitler is a consultant for KCI, and Depuy Synthes has received research support from Stryker Corp, is a paid employee of JBJS/Lippincott Wolters Kluwer, and serves as a committee member in the American Academy of Orthopaedic Surgeons and the Orthopaedic Trauma Association. All other authors have no financial disclosures.

Author Disclosure Statement

No benefits in any form have been or will be received from a commercial party related directly or indirectly to the subject of this article.

References

Rodriguez

, Jung

, Goulet

, et al. Evidence-based protocol for prophylactic antibiotics in open fractures: Improved antibiotic stewardship with no increase in infection rates. J Trauma Acute Care Surg, 2014; 77:400–408.

Metsemakers

, Smeets

, Nijs

, et al. Infection after fracture fixation of the tibia: Analysis of healthcare utilization and related costs. Injury, 2017; 48:1204–1210.

Wijeratna

, McRoberts

, Porteous

. Cost of infection after surgery for intracapsular fracture of the femoral neck. Ann R Coll Surg Engl, 2015; 97:283–286.

Obremskey

, Metsemakers

, Schlatterer

, et al. Musculoskeletal infection in orthopaedic trauma: Assessment of the 2018 International Consensus Meeting on Musculoskeletal Infection. J Bone Joint Surg Am, 2020; 00-A(00):1–9.

Lack

, Karunakar

, Angerame

, et al. Type III open tibia fractures: Immediate antibiotic prophylaxis minimizes infection. J Orthop Trauma, 2015; 29:1–6.

Patzakis

, Wilkins

. Factors influencing infection rate in open fracture wounds. Clin Orthop Rel Res, 1989; 6:36–40.

Hauser

, Adams

, Eachempati S; Council of the Surgical Infection

Society

. Surgical Infection Society Guideline: Prophylactic antibiotic use in open fractures: An evidence-based guideline. Surg Infect, 2006; 7:379–405.

Gustilo

, Anderson

. Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones. J Bone Joint Surg Am, 1976; 58-A:453–458.

Folse

, Hill

, Graves

, et al. Risk factors for kidney dysfunction with the use of gentamicin in open fracture antibiotic treatment. J Orthop Trauma, 2018; 32:573–578.

10.

Redfern

, Pharm

, Wasilko

, et al. Surgical site infections in patients with Type 3 open fractures: Comparing antibiotic prophylaxis with cefazolin plus gentamicin versus piperacillin/tazobactam. J Orthop Trauma, 2016; 30:415–419.</BIBTX

11.

Patzakis

, Bains

, Lee

, et al. Prospective, randomized, double-blind study comparing single-agent antibiotic therapy, ciprofloxacin, to combination antibiotic therapy in open fracture wounds. J Orthop Trauma, 2000; 14:529–533.

12.

Ghoshal

, Enninghorst

, Sisak

, et al. An interobserver reliability comparison between the Orthopaedic Trauma Association's open fracture classification and the Gustilo and Anderson classification. Bone Joint J, 2018; 100-B(2):242–246.

13.

Brumback

, Jones

. Interobserver agreement in the classification of open fractures of the tibia. J Bone Joint Surg, 1994; 76:1162–1166.

14.

Hoff

, Bonadies

, Cachecho

, et al.; East Practice Management Guidelines Work Group: Update to practice management guidelines for prophylactic antibiotic use in open fractures. J Trauma Injury Infect Crit Care, 2011; 70:751–754.

15.

Gustilo

. To the Editor: On “Interobserver agreement in the classification of open fractures of the tibia: The results of a survey of two hundred and forty-five orthopaedic surgeons.”. J Bone Joint Surg Am, 1995; 77-A(8):1291–1292.

16.

Barton

, Mcmillian

, Crookes

, et al. Compliance with the Eastern Association for the Surgery of Trauma Guidelines for prophylactic antibiotics after open extremity fracture. Int J Crit Illness Injury Sci, 2012; 2:57–62.

17.

Robinson

, On

, Hadas

, et al. Microbiologic flora contaminating open fractures: Its significance in the choice of primary antibiotic agents and the likelihood of deep wound infection. J Orthop Trauma, 1989; 3:283–286.

18.

Murray

, Obremskey

, Hsu

, et al. Prevention of infections associated with combat-related extremity injuries. J Trauma Injury Infect Crit Care, 2011; 71:S235–S257.

19.

Kendall

, Murry

TGJ

, Agrawal DH

, et al. Antibiotics and open fractures of the lower extremity: Less is more. Eur J Trauma Emerg Surg, 2019; 45:125–129.

20.

Roth

, Fry

, Polk

. Infectious morbidity in extremity fractures. J Trauma, 1986; 26:757–761.

21.

Pogue

, Kaye

, Cohen

, et al. Appropriate antimicrobial therapy in the era of multidrug-resistant human pathogens. Clin Microbiol Infect, 2015; 21:302–312.

22.

Dicks

, Lewis

, Durkin

, et al. Surveying the surveillance: Surgical site infections excluded by the January 2013 updated surveillance definitions. Infect Control Hosp Epidemiol, 2014; 35:570–573.

23.

Sun

, Wang

, Tang

, et al. Incidence and risk factors for surgical site infection after open reduction and internal fixation of ankle fracture: A retrospective multicenter study. Medicine, 2018; 97:e9901.

24.

Wang

, Pei

, Chen

, Wang

. Incidence and predictors of surgical site infection after ORIF in calcaneus fractures: A retrospective cohort study. J Orthop Surg Res, 2018; 13:293.

25.

Shawar

, Ly

, Li

, et al. Piperacillin/tazobactam versus tobramycin-based antibiotic prophylaxis for Type III open fractures. Surg Infect, 2020; 21:23–28