Spine
SPINE Volume 39, Number 2, pp 177-184 ©2014, Lippincott Williams & Wilkins
SURGERY
Experience With Intrawound Vancomycin Powder for Spinal Deformity Surgery Joel R. Martin, MS,* Owoicho Adogwa, MD, MPH,* Christopher R. Brown, MD,t Carlos A. Bagley, MD,* William J. Richardson, MD,t Shivanand P. Lad, MD, PhD,* Maragatha Kuchibhatia, PhD,t and Oren N. Gottfried, MD*
Study Design. Retrospective cohort study. Objective. To evaluate the ability of local vancomycin powder to prevent deep wound infection after thoracolumbar and lumbar spinal fusion for open deformity cases. Summary of Background Data. Recent studies report that local delivery of vancomycin powder is associated with a decrease in spinal surgical site infection (SSI). This study compares deformity fusion cases before and after the routine application of spinal vancomycin powder. Metiiods. Posterior spinal deformity surgical procedures by a single institution were reviewed from January 2011 to April 2013. Routine application of vancomycin powder started in April 2012. Inclusion criteria included adult patients who underwent posterior fusion for deformity pathologies, including spondylolisthesis, kyphosis, sagittal imbalance, and scoliosis. Each cohort's baseline characteristics including infection risk factors, operative data, and rates of wound infection were compared. Associations between infection and vancomycin powder, with and without propensity score adjustment for risk factors were determined using logistic regression. Results. A total of 306 patients were included in the study All measured baseline and operative variables were statistically similar between untreated (n = 150) and those who received vancomycin powder (n = 156). No significant change in deep wound infection rate was seen between the control (5.3%) and intervention group (5.1%, P = 0.936). Logistic regression with and without propensity score adjusted for risk factors demonstrated that the use of vancomycin powder did not impact the development of SSI (odds ratio [95% confidence intervall: 1.01 [0.36-2.79], P = 0.9910)
From the *Division of Neurosurgery, Department of Surgery; tDivision of Orthopaedic Surgery, and ^Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC. Acknowledgment date: July 2, 2013. First revision date: September 27, 2013. Acceptance date: October 1, 2013. The device(s)/drug(s) is/are FDA-approved or approved by corresponding national agency for tbis indication. No funds were received in support of tbis work. Relevant financial activities outside the submitted work: consultancy payment for lecture, royalties. Address correspondence and reprint requests to Oren N. Gottfried, MD, Division of Neurosurgery, Department of Surgery, Duke University Medical Center, Box 3807, Durham, NC 27701; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000071 Spine
and (odds ratio [95% confidence interval]: 0.87 [0.31-2.42], P = 0.7876), respectively. Conclusion. The local application of powdered vancomycin was not associated with a significant difference in the rate of deep SSI after spinal deformity surgery, and other treatment modalities are necessary to limit infection for this high-risk group. This study is in contrary to prior studies, which have reported a decrease in SSI with vancomycin powder. Key words: vancomycin powder, local antibiotics, wound infection, spine surgery, posterior lumbar fusion, deformity Levei of Evidence: 2 Spine 2014;39:177-184
T
he incidence of surgical site infections (SSIs) after spine surgery ranges from 0.6% to 12%.'"'" Despite the use of prophylactic systemic antibiotics and improved operative technique, postoperative SSIs remain of serious concern. Deep wound infection can have a profound impact on patients because they often require prolonged hospital stay, additional surgical interventions, lengthy intravenous antibiotic administration, and delayed rehabilitation.^-'' Furthermore, numerous studies have demonstrated elevated mortality risk in patients who experience a postoperative SSI.-'^-^ Given the incidence of deep wound infections, several authors explored and subsequendy published their results on locally applied prophylactic antibiotics.'"'' Intrawound vancomycin powder is an attractive option for additional prophylaxis against postoperative SSI, and initial studies have shown favorable results. Recently, the addition of local vancomycin powder to standard systemic antibiotic prophylaxis was shown to reduce SSI rates from 2.6%-15% to 0%-2.5%."-'''i*.i"« O'Neil et al'' reported a 13% reduction in the incidence of postoperative deep SSIs in patients undergoing posterior spinal stabilization of traumatic injuries. Similarly, Sweet et al^* observed a 2.4% decrease in the incidence of postoperative SSIs in patients undergoing elective spinal fusions using vancomycin powder prophylaxis. Although the theoretical advantage of locally applied vancomycin powder to standard systemic antibiotic prophylaxis is to reduce the incidence of SSIs, the results of puhlished studies on this subject are not all consistent, and it has not been explored for all high-risk populations.'" Prior www.spinejournal.com
177
Spine
SURGERY
Studies have not assessed the impact of local vancomycin for spinal deformity that involves longer incisions, increased operative time and blood loss, patients with more significant comorbidities, and higher infection rates.^"'-'^"^' Accordingly, this study compares postoperative deep surgical infection rates in a large continuous cohort of patients undergoing spinal deformity fusion surgery involving thoracolumbar and lumbar vertebrae before and after the routine application of locally applied vancomycin powder.
MATERIALS AND METHODS
Experience With Intrawound Vancomycin • Martin et al
affirmative or negative vancomycin usage were excluded from analysis. The surgical approach involved an open midline posterior incision for all patients. Minimally invasive cases were excluded. The decision to use subfascial drains were based on surgeon preference. However, in most cases, subfascial drains were used. In all cases, the wounds were closed with absorbable suture in the fascia and subcutaneous layers, and with staple or suture closure of the skin. After skin closure, incisions were cleaned again with chlorhexidine and a sterile dressing was applied. Dressings and drains were kept in place until the second postoperative day.
Patient Selection The primary aim of this study was to determine whether the addition of locally apphed vancomycin powder to standard systemic antibiotic prophylaxis would impact the incidence of postoperative deep wound infections for patients who underwent thoracolumbar or lumbar spinal deformity surgery by one of nine spine surgeons at our institution. The institutional review board approved this study. A retrospective review of hospital records from January 2011 to April 2013 was performed of adult patients older than 18 years who had undergone open posterior spinal fusion using pedicle screws and rod instrumentation that included lumbar spine (thoracolumbar, thoracosacral, iliothoracic, lumbar, lumbosacral, and iliolumbar) fusion for deformity cases. Inclusion causes included all types of deformity pathologies, including spondylolisthesis, kyphosis, sagittal imbalance, and scoliosis. Exclusion criteria included traumatic and neoplastic causes. Patient demographics, clinical presentation, comorbidities, radiological studies, and all treatment variables were reviewed for each case.
Standard Pre- and Postoperative Systemic Prophylactic Antibiotic Regiment All patients received standard systemic antibiotic prophylaxis consisting of IV cefazolin within 1 hour of surgical incision followed by IV cefazolin every 8 hours for 1 day. If the patient was allergic to penicillin, clindamycin was used instead. Other intravenous antibiotics that were used infrequently included: ciprofioxacin, piperacillin/tazobactam, or vancomycin. All patients were prepared with chlorhexidine. Fusion levels were determined by preoperative imaging. Before skin closure, irrigation with 2 L of normal saline by gravity was performed.
Treatment and Control Cohorts Two patient cohorts were developed. The treatment group consisted of consecutive patients who received 2 g of intraoperative vancomycin powder (April 2012-April 2013), spread throughout the local wound. The powder was placed directly on the muscle, fascia, and subcutaneous tissues taking care not to expose bone graft or dura. The control group consisted of consecutive patients who received their operation prior to the institutional use of vancomycin powder (January 2011-March 2012). Patients who lacked documentation of 178
www.spinejournal.com
Diagnostic Evaluation The definition of a deep SSI developed by the Genters for Disease Gontrol and Prevention were used to make a SSI diagnosis.^^ According to the Genters for Disease Gontrol and Prevention definitions, deep space incisional SSIs all occur within 30 days after surgery. Additionally, they involve purulent drainage, isolation of organism, signs or symptoms of infection (such as pain or tenderness, locaHzed swelling, redness, or heat) combined with positive culture results, and diagnosis by a surgeon or attending physician. Thus, for this study, a SSI was defined as being diagnosed during the initial hospitalization or during a hospital readmission or postoperative clinic appointment within 30 days of the surgery. All patients in this study had standard laboratory tests on admission to the hospital including: erythrocyte sedimentation rates, peripheral white blood cell counts, G-reactive protein, complete urine analysis, and microbiology and blood cultures. Erythrocyte sedimentation rate was determined by the Westergren method and considered abnormal if greater than 15 mm/hr. G-reactive protein was considered abnormal if greater than 5 mg/dL. Bacterial identification and susceptibility testing were performed according to the guidelines of the Genters for Disease Gontrol and Prevention.
Clinical Parameters Preoperative and intraoperative data for each patient were collected with use of patient charts and computerized medical records. Surgical infection risk factors were documented for each patient, including body mass index, obesity, diabetes mellitus, insulin use, chronic steroid use, prior lumbar surgery, chronic obstructive pulmonary disease, tobacco use, neurological deficit, immunodeficiency, and osteoporosis. Pertinent operative details were also collected, including number of levels instrumented, estimated blood loss, duration of anesthesia, intraoperative steroid use, blood transfusion, and durai tear.
Statistical Analysis Each cohort's baseline characteristics, operative data, and rates of wound infection were compared, using x^ tests for categorical variables and Student t tests for continuous variables. Associations between SSI and vancomycin powder, with and without propensity score adjustment for risk factors were determined using logistic regression. lanuary 2014
Spine
SURGERY
Experience With Intrawound Vancomycin • Martin et al
RESULTS A total of 306 adult patients undergoing deformity fusions were included in the study, with 150 patients (49 male, 101 female) in the untreated group and 156 patients (49 male, 107 female) in the vancomycin-treated group. Three patients were excluded because of incomplete records. Ninety-nine patients who underwent nondeformity instrumentations were not included. The untreated and treated groups were statistically similar (P > 0.05) with regard to baseline and operative
variables (Table 1). Baseline patient age was similar between both cohorts (untreated, 62.7; treated, 63.4; P = 0.618). Mean body mass index and prevalence of diabetes were similar between both cohorts (untreated, 29.6; treated, 30.0; P = 0.541) and (untreated, 18.0%; treated, 23.1%; P = 0.272). Similarly, tobacco use was not statistically different (untreated, 18.0%; treated, 12.8%; P = 0.209). The mean duration of anesthesia was similar between both cohorts (untreated, 322.4 min; treated, 328.1 min;
^^^^^^PmpârîsorTôîfetïëîrtnvîïïrDëfonTif^^ I ^ ^ ^ H B i l l ^ ^"d ^^^i" theijÉiÉHilHÉBn of VajHÍÉHin Powder j ^ ^ i ^ ^ ^ ^ ^ H ^ ^ H ^ ^ H ^ I • — Treated: 156 pts
P
62.7
63.4
0.618
27(18.0%)
36(23.1%)
0.272
7 (4.7%)
10 (6.4%)
0.506
Untreated: 150 pts
1
Baseline features Age (mean), yr Diabetes Diabetes, insulin BMI (mean)
29.6
30.0
0.541
Obesity
64 (42.7%)
74 (47.4%)
0.402
Previous lumbar fusion
47(31.3%)
50(32.1%)
0.893
Previous lumbar surgery
75 (50%)
72 (46.2%)
0.501
Current EtOH abuse
3 (2.0%)
6 (3.8%)
0.339
Current smoker
27(18.0%)
20(12.8%)
0.209
illicit drug user
1 (0.7%)
1 (0.6%)
0.978
CAD
33 (22.0%)
30(19.2%)
0.549
Osteoporosis
25(16.7%)
24(15.4%)
0.760
COPD
10(6.7%)
15(9.6%)
0.346
Current steroid use
10(6.7%)
10 (6.4%)
0.928
Hyperlipidemia
91 (60.7%)
80(51.3%)
0.098
Hypertension
105 (70.0%)
114(73.1%)
0.551
Neurological deficit
6 (4.0%)
5 (3.2%)
0 709
Immunodeficiency
1 (0.7%)
1 (0.6%)
0.978
4.5
5.1
0.059
EBL (mean), mL
1050.5
1314.7
0.101
Duration of anesthesia (mean), min
322.4
328.1
0.731
Use of postoperative drain
137(91.3%)
145(92.9%)
0.599
Intraoperative steroid use
38 (25.3%)
43 (27.6%)
0.658
Blood transfusion
75 (50.0%)
87 (55.8%)
0.312
14 (9.3%)
9 (5.8%)
0.237
8 (5.3%)
8(5.1%)
0.936
Operative details No. of levels instrumented (mean)
Durai tear Outcomes Deep SSI
pts indicates patients; BMI, body mass index; EtOH, ethanol; CAD, coronary artery disease; COPD, chronic obstructive pulmonarv disease- SSI surgical site infection. ' ' - s
Spine
wwvi/.spinejournai.com
179
Experience With Intrawound Vancomycin • JVIartin et al
SURGERY
P = 0.731). The mean estimated blood loss was smaller in the untreated cohort (untreated, 1050.5 mL; treated, 1314.7 mL; P = 0.101). Postoperative drains were used in a similar number of patients (untreated, 91.3%; treated, 92.9%; P = 0.599). Vancomycin-treated patients trended toward an increased number of surgical levels (untreated, 4.5; treated, 5.1; P = 0.059). Of interest, a similar number of patients from both cohorts experienced a durotomy (untreated, 9.3%; treated, 5.8%; P = 0.237). Both the untreated and treated groups had 8 deep SSIs (untreated, 5.3%; treated, 5.1%; P = 0.936) (Table 1; Figure 1). Small, nonsignificant differences were seen in patients with a previous lumbar surgery (n = 147; untreated, 6.7%; treated, 5.6%; P = 0.779), previous lumbar fusion (n = 97; untreated, 10.6%; treated, 8.0%; P = 0.654), and those with no history of lumbar surgery (n = 159; untreated, 4.0%; treated, 4.8%; P = 0.815). Lastly, when evaluating patients with diabetes, tobacco use, immunodeficiency, neurological deficits, myelomeningocele, or paralysis, there was not a significant difference in infection rate (n = 107; untreated, 5.5%; treated, 5.8%; P = 0.944). The mean number of postoperative days to infection was 15.1 days for the treated cohort and 15.6 days for the untreated cohort (P = 0.90) (Table 2). A comparison of other baseline or operative variables between untreated and treated patients with a deep SSI revealed no significant differences (Table 2). Likewise, there was no significant difference between choice of pre- and perioperative FV antibiotic in both groups: cefazolin (untreated, 74.7%; treated, 78.8%; P = 0.387), clindamycin (untreated, 16.7%; treated, 15.4%; P = 0.760), vancomycin (untreated, 8.0%; treated, 5.1%; P = 0.310), and single cases of ciprofioxacin and piperacillin/tazobactam. A detailed description of each SSI is shown in Tables 3 and 4 for vancomycin-treated and untreated patients, respectively. There was no significant difference in culture results for SSIs. The majority of documented infections in both groups were
CD ••a
o SP oS
Treated
Untreated
Figure 1. Incidence of postoperative surgical site infections in patients receiving intraoperative vancomycin powder in addition to standard systemic antibiotic prophylaxis. An equal number of patients in botb cohorts experienced a deep SSI. SSI indicates surgical site infection. 180
www.spinejournal.com
due to either gram-positive cocci (other than methicillin-resistant Staphylococcus aureus [MRSA]) (untreated, 4/8; treated, 3/8) or gram-negative rods (untreated, 4/8; treated, 4/8). There were 2 MRSA infections in the untreated group and 1 MRSA infection in the treated group. Logistic regression was used to determine the effect of use of vancomycin powder to predict SSI. The use of vancomycin powder was found to have no impact on the development of SSI (odds ratio [95% confidence interval]: 1.01 [0.36-2.79], P = 0.9910). Propensity to predict the usage of vancomycin powder was obtained using the preoperative baseline variables in Table 1. Propensity score adjusted for effect of use of vancomycin powder was not found to be a significant predictor for the development of SSI, as well (odds ratio [95% confidence interval]: 0.87 [0.31, 2.42], P = 0.7876).
DISCUSSION In addition to systemic antibiotics, spinal intrawound vancomycin is an attractive option. Local powdered vancomycin is inexpensive, has broad coverage against typical organisms responsible for deep spinal wounds, and seems to carry a lower risk of systemic complications than intravenous form. Most published data have supported its use, reporting decreased rates of postoperative wound infection and no adverse side effects.'-"-'* In cardiac surgery literature, Vander Salm et aP^ showed that topical vancomycin reduced the incidence of sternal wound infections from 3.6% to 0.45% (P = 0.02) after cardiac surgery. Lazar et aP addressed the concern of persistent serum levels of vancomycin after topical sternal application, that may contribute to the emergence of drug-resistant infections, by reporting a significant decrease in serum vancomycin levels from the night of surgery to the sixth postoperative day (P < 0.0001). In a rodent study, Ozcan et aP- used topical and intravenous vancomycin on rats with deep sternal wounds inoculated with MRSA. The group reported a significant (P < 0.05) reduction in MRSA wound organisms in 3 treated groups (topical, IV, and topical plus IV), with topical plus rV having the largest reduction. In neurosurgical and orthopedic literature, several studies have looked at the addition of local vancomycin to standard systemic antibiotic prophylaxis use in posterior spinal fusion surgery (Table 5). Cohort studies have reported that local vancomycin reduced SSI rates in adult thoracolumbar fusion surgery (n = 165-1732, 2.6%-12% vs. 0%-0.2%, P < O.OOl),'''-'** posterior fusion after traumatic causes (n = 110, 13% vs. 0%, P = 0.02),'^ and posterior cervical decompression and fusion surgical procedures (n = 112-171, 10.9%15% vs. 0%-2.5%, P < 0.05).'^''* Similarly, after introduction of vancomycin powder, Molinari ei fl/" reported a low 1.20% SSI rate (n = 663) in spine fusion surgery. In all studies, no adverse events were noted in any of the intervention groups associated with the use of vancomycin powder. Conversely, one group reported no significant change in postoperative deep wound infection. Mohammed et aP° compared open vascular procedures with and without intrawound vancomycin powder. In the 454 patient study, there was a January 2014
Spine
SURGERY
TABLE 2.
Experience With Intrawound Vancomycin • Martin et al
pomparison of Patients With Deförrffity UnderpingThoracolumbar anci L u m b a i ^ ^ H | Fusion With Postoperative Deep Wound Infection Before and After the Introductioii^ Untreated: 8 pts
Treated: 8 pts
P
15.6
15.1
0.90
67.0
57.3
0.27
Diabetes
2 (25.0%)
2 (25.0%)
1
Diabetes, insulin
1 (12.5%)
0 (0%)
0.30
32.0
32.8
0.86
Obesity
4 (50.0%)
5 (62.5%)
0.61
Previous lumbar fusion
5 (62.5%)
4 (50.0%)
0.61
Previous lumbar surgery
5 (62.5%)
4 (50.0%)
0.61
Current FtOH abuse
1 (12.5%)
0 (0%)
0.30
Current smoker
2 (25.0%)
2 (25.0%)
1
Illicit drug user
0 (0%)
0 (0%)
CAD
1 (12.5%)
2 (25.0%)
0.52
Osteoporosis
2 (25.0%)
1 (12.5%)
0.52
COPD
1 (12.5%)
1 (12.5%)
1
0 (0%)
1 (12.5%)
0.30
Hyperlipidemia
5 (62.5%)
5 (62.5%)
1
Hypertension
8(100%)
5 (62.5%)
0.055
Neurological deficit
0 (0%)
0 (0%)
Immunodeficiency
0 (0%)
0 (0%)
6.9
6.1
0.54
EBL (mean), mL
2831.3
2072.5
0.42
Duration of anesthesia (mean), min
420.5
481.3
0.48
Use of postoperative drain
8(100%)
8 (100%)
Intraoperative steroid use
1 (12.5%)
2 (25.0%)
0.52
Blood transfusion
7 (87.5%)
5 (62.5%)
0.25
Durai tear
1 (12.5%)
1 (12.5%)
1
Postoperative details Time to infection (mean), d Baseline features Age (mean), yr
BMI (mean)
Current steroid use
Operative details No. levels instrumented (mean)
pts indicates patients; BMI, body mass index; EtOH, ethanol; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease- FBL estimated blood loss. ' '
small but statistically significant decrease in the 30-day incidence of superficial infections (18.9% vs. 11.5%, P = 0.033), but no significant difference in the incidence of deep wound infections (6.1% vs. 5.7%, P = 0.692) or overall dehiscence rates (22.2% vs. 17.7%, P = 0.239). The group postulated that the blunted effect of local vancomycin on deep wounds could be due to the local vascular environment resulting in Spine
poor systemic absorption of the local antibiotic lessening its systemic effect on deep wound infections. Similarly, our study did not show a significant reduction in deep SSIs in posterior spinal deformity fusion cases (n = 306, 5.3% vs. 5.1%, P = 0.936). Our study involved a population of patients with spinal deformity that are at higher risk for infection due to very large incisions, greater time of surgery, www.spinejournal.com
181
Spine
Experience With Intrawound Vancomycin • Martin et al
SURGERY
"•
TAB LE 3.
•m
i
•nts From Treatmeriï (Vancomycin) Group With Deep Rostoperative Surgical
1
Infection Onset (d)
Fusion Levels
Tobacco Use
DM
BMl
Other Risk Factors
Culture
pt
Age Sex
1
42 F
Pseudarthrosis, spondylolisthesis
28
L3-S1
No
No
37.2
Prior fusion
Coag-neg Staph
2
63 F
Lumbar scoliosis, spondylolisthesis
30
L3-S1
Yes
No
26.2
COPD
E. cloacae
3
39 F
Pseudarthrosis, spondylolisthesis
13
L4-S1
No
Yes
41.3
Prior fusion
MSSA
4
67 F
Charcot spine with instability
14
L1-S1
Yes
Yes
52.9
5. marcescens
5
18F
Thoracolumbar scoliosis
4
T1-L2
No
No
20.1
S. marcescens
6
74 F
Spondylolisthesis, junctional kyphosis
12
TU-SI
No
No
30.8
Prior fusion, CAD
Coag-neg Staph
7
82 M
Thoracolumbar scoliosis
7
TiO-ilium
No
No
22.4
CAD
MRSA
8
73 F
Sagittal imbalance, pseudarthrosis
13
TU-SI
No
No
31.8
Prior fusion, steroid use, osteoporosis
K. pneumoniae, C. freundii
Diagnosis
p i indicates patients; DM, diabetes mellitus; BMl, body mass index; COPD, chronic obstructive pulmonary disease; CAD, coronary artery disease; MSSA, methiciilin-sensitive Staphyloeoccus aureus; Coag-neg Staph, Coaguiase-negative staphyiococci; F. cloacae, Fnterobacter cloacae; S. marcescens. Serrada marcescens; K. pneumoniae, Klebsiella pneumoniae; C freundii, Citrobacter freundii.
duration of anesthesia, blood loss,^ use of instrumentation,'' increased levels of instrumentation,^"-^' and other complications.' Because of these risk factors, patients with spinal deformity pose a greater risk of postoperative SSI. A possible theory for our contrary results as compared with other vancomycin studies is that local vancomycin antibiotic may not have enough capacity to overcome the increased infection risk factors seen in patients with deformity. Alternatively, most prior studies
used 1 g of vancomycin powder'^''^'^ and we averaged 1.75 g, which may not be sufficient to prevent infection for these larger incisions, and further studies are needed to evaluate this point. Finally, the addition of other treatment modalities may be needed to prevent SSIs in this high-risk population, such as improved nutrition or more stringent surgical selection criteria. This study was larger than several of the prior studies,''-"^""* and attempted to control for many spinal SSI risk factors
mBlHP
TABLE 4 . " Infection Onset (d)
Fusion Levels
Tobacco Use
DM
BMl
other Risk Factors
Culture
pt
Age, Sex
9
51 M
Spondylolisthesis, stenosis
14
L2-L5
No
No
41.9
Prior fusion
MSSA
10
63 F
Spondylolisthesis, sagittal imbalance
12
L3-L5
Yes
No
19.2
Osteoporosis
MSSA
11
62 F
Spondylolisthesis, sagittal imbalance
22
T12-L5
No
No
34.4
Prior fusion, COPD
Coag-neg Staph
12
67 F
Lumbar kyphosis, sagittal imbalance
7
T10-S1
Yes
Yes
28.7
Prior fusion
MRSA, E. aerogenes
13
74 M
Lumbar kyphosis, sagittal imbalance
8
T12-L5
No
No
32.6
Prior fusion, CAD
MRSA
14
79 M
Thoracolumbar scoliosis
26
T6-ilium
No
Yes
28.8
15
76 F
Thoracolumbar scoliosis
21
T6-ilium
No
No
28
Osteoporosis
Enterococcus, P. aeruginosa
16
64 F
Lumbar kyphosis, sagittal imbalance
15
T11-L5
No
No
42.1
Prior fusion
P. aeruginosa
Diagnosis
P. mirabilis
pt indicates patient; DM, diabetes mellitus; BMl, body mass index; CORD, chronic obstructive pulmonary disease; CAD, coronary artery disease; MSSA, methiciilin-sensitive Staphytococcus aureus; Coag-neg Staph, Coaguiase-negative staphylococci; E. aerogenes, Enterobacter aerogenes; P. aeruginosa, Pseudomonas aeruginosa; P. mirabilis, Proteus mirabilis.
182
www.spinejournal.com
January 2014
SURGERY
Experience With Intrawound Vancomycin • Martin et al
TABLE 5.
^^SMÏBlf«'#
Author and Year Sweet ef a/,''2011
Location Thoracic, lumbar
Patients With Fusion
Cause
Infection Rate (Cases With Fusion) Control: 2.6%
All posterior fusion cases
1732 Treated: 0.2% ( P < 0.0001)
O'Neill eta/,''2011
Cervical, thoracic, lumbar
Posterior trauma fusion cases
110
Molinari ef a/," 2012
Cervical, thoracic, lumbar
All spine cases
663
Strom ei a/,'»2013
Lumbar
All posterior cases
165
Control: 13% Treated: 0% (P = 0.02) Treated: 1.20% Control: 12% Treated: 0% (P = 0.0008)
Strom eia/,'*2013
Cervical
All posterior fusion cases
171
Control: 10.9% Treated: 2.5% (P = 0.0384)
Caroomefa/,"'2013
Cervical
Posterior multilevel spondylotic myelopathy fusion cases
including diabetes, steroid use, previous spinal surgery, obesity, smoking, duration of anesthesia, estimated blood loss, drug use,' blood transfusion, coronary artery disease, osteoporosis, durai tear, and chronic obstructive pulmonary disease.' Vancomycin-treated patients trended toward higher blood loss and increased number of levels treated. These factors would be expected to increase the risk of deep infection in the treatment group. However, both groups of 8 patients with postoperative infections had a balanced distribution of chnical and operative variables (Table 2), and our logistic regression analysis controlling for risk factors and operative details did not show an impact for vancomycin in affecting incidence of infection. Furthermore, we reviewed clinical records, operative and nursing notes, and surveyed all surgeons and did not identify any changes in practice patterns or external factors that changed during these two time periods before and after the routine use of vancomycin powder. Simple linear regression analysis by month showed no significant relationship between infection rate and date of operation (P = 0.714). The limitations inherent in our study have implications for its interpretation. A power analysis demonstrated that there was sufficient sample size to detect a change in the rate of infection seen in prior studies using vancomycin powder. However, given that the event rate for postoperative infection is small, this study is not powered to detect subtle differences. This study was limited by its retrospective design, focus on deformity diseases, and resulting confounding factors that may have been undetected. We also focused on deep wound complications within the 30-day postoperative period, not accounting for superficial or possible late infections. Also, the study involved 9 surgeons that have similar but nonexact techniques. Note the distribution of infections showed that no one surgeon had a significantly greater infection rate. Also, both cohorts centered on the introduction of vancomycin powder, which initially began because of surgeon's discretion and may have introduced selection bias. Additionally, serum vancomycin levels were not measured, which could have Spine
112
Control: 15% Treated: 0% (P = 0.007)
aided in proper dosing and assessing its duration of impact. However, in our study, we did not detect toxicity or complications from the routine use of vancomycin powder. Finally, drains were common in this series, and could have drained or limited the vancomycin concentration.''' Despite these limitations, our study demonstrated that the use of locally applied vancomycin powder did not reduce the incidence of 30-day postoperative deep SSI rate in spinal deformity fusion. This study is in contrary to prior spinal studies, which have reported a decrease in SSI with the addition of perioperative intrawound vancomycin powder.
CONCLUSION Despite our conclusions, our institution is continuing to use and investigate the use of intrawound spinal vancomycin powder because our sample size is too small to detect any subtle but beneficial effects of this intervention. We continue to use 2 g for larger surgical procedures, but also question whether a larger dose may be beneficial. A randomized prospective study will be performed to corroborate our findings.
Key Points •
In this retrospective review of 306 patients undergoing posterior thoracolumbar or lumbar deformity fusion, the infection rate did not significantly change (untreated 5.3% vs. treated 5.1%, P = 0.936) after the application of local vancomycin powder.
•
The lack of efficacy of intrawound vancomycin to reduce surgery site infections for spinal deformity is contrary to results for other spine surgical procedures demonstrating that it may not be able to overcome the increased infection risk factors seen in patients with deformity or that increased dosages may be needed. www.spinejournal.com
183
Spine •
SURGERY
Further prospective studies are needed to evaluate if intrawound vancomycin powder combined with other treatnnent modalities may benefit patients undergoing spinal deformity surgery.
References 1. Pull ter Gunne AF, van Laarhoven CJ, Cohen DB. Surgical site infection after osteotomy of the adult spine: does type of osteotomy matter? Spine} 20W;W:410-6. 2. Kirkland KB, Briggs JP, Trivette SL, et al. The impact of surgicalsite infections in the 1990s: attributable mortality, excess length of hospitalization, and extra costs. Infect Control Hosp Epidemiol 1999;20:725-30. 3. Koutsoumbelis S, Hughes AP, Girardi FP, et al. Risk factors for postoperative infection following posterior lumbar instrumented arthrodesis. / Bone Joint Surg Am 2011;93:1627-33. 4. Massie JB, Heller JG, Abitbol JJ, et al. Postoperative posterior spinal wound infections. Clin Orthop Relat Res 1992;99-108. 5. Wimmer C, Gluch H, Franzreb M, et al. Predisposing factors for infection in spine surgery: a survey of 850 spinal procedures. / S/?¿nal Disord I998;n:124-S. 6. Whitehouse JD, Friedman ND, Kirkland KB, et al. The impact of surgical-site infections following orthopedic surgery at a community hospital and a university hospital: adverse quality of life, excess length of stay, and extra cost. Infect Control Hosp Epidemiol 2002;23:\83-9. 7. Klevens RM, Edwards JR, Richards GL, Jr, et al. Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health Rep 2007;122:160-6. 8. Anderson DJ, Kaye KS, Glassen D, et al. Strategies to prevent surgical site infections in acute care hospitals. Infect Control Hosp £p;ciem;o/2008;29(suppl 1):S51-61. 9. Lazar HL, Barlam T, Gabral H. The effect of topical vancomycin applied to sternotomy incisions on postoperative serum vancomycin levels. / Card Surg 2011;26:461-5. 10. Mohammed S, Pisimisis GT, Daram SP, et al. Impact of intraoperative administration of local vancomycin on inguinal wound complications. ] Vase Surg 2013;57:1079-83.
184
www.spinejournal.com
Experience With Intrawound Vancomycin • Martin et al
11. Molinari RW, Khera OA, Molinari WJ. Prophylactic intraoperative powdered vancomycin and postoperative deep spinal wound infection: 1,512 consecutive surgical cases over a 6-year period. Eur Spi«e/ 2012;21(suppl 4):S476-82. 12. Ozcan AV, Demir M, Onem G, et al. Topical versus systemic vancomycin for deep sternal wound infection caused by methicillinresistant Staphylococcus aureus in a rodent experimental model. Tex Heart Inst] 2006;33:107-10. 13. Strom RG, Pacione D, Kalhorn SP, et al. Decreased risk of wound infection after posterior cervical fusion with routine local application of vancomycin powder. Spine (Phila Pa 1976) 2013. 14. Sweet FA, Roh M, Sliva C. Intrawound application of vancomycin for prophylaxis in instrumented thoracolumbar fusions: efficacy, drug levels, and patient outcomes. Spine (Phila Pa 1976) 2011;36:2084-8. 15. Vander Salm TJ, Okike ON, Pasque MK, et al. Reduction of sternal infection by application of topical vancomycin. / Thorac Cardiovasc Surg 19S9;9S:618-22. 16. Garoom G, Tullar JM, Benton FG, Jr, et al. Intrawound vancomycin powder reduces surgical site infections in posterior cervical fusion. Spine (Phila Pa 1976) 2013. 17. O'Neill KR, Smith JG, Abtahi AM, et al. Reduced surgical site infections in patients undergoing posterior spinal stabilization of traumatic injuries using vancomycin powder. Spine ] 2011;ll: 641-6. 18. Strom RG, Pacione D, Kalhorn SP, et al. Lumbar laminectomy and fusion with routine local application of vancomycin powder: decreased infection rate in instrumented and non-instrumented cases. Clin Neurol Neurosurg 2013. 19. Klein JD, Hey LA, Yu GS, et al. Perioperative nutrition and postoperative complications in patients undergoing spinal surgery. Spine (Phila Pa 1976) 1996;21:2676-82. 20. Olsen MA, Nepple JJ, Riew KD, et al. Risk factors for surgical site infection following orthopaedic spinal operations. / Bone Joint Surg Am 2008;90:62-9. 21. Schimmel JJ, Horsting PP, de Kleuver M, et al. Risk factors for deep surgical site infections after spinal fusion. Eur Spine J 2010;19:1711-9. 22. Mangram AJ, Horan TG, Pearson ML, et al. Guideline for prevention of surgical site infection, 1999. Hospital Infection Gontrol Practices Advisory Gommittee. Infect Control Hosp Epidemiol 1999;20:250-78; quiz 279-80.
January 2014
Copyright of Spine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
SPINE Volume 39, Number 2, pp 177-184 ©2014, Lippincott Williams & Wilkins
SURGERY
Experience With Intrawound Vancomycin Powder for Spinal Deformity Surgery Joel R. Martin, MS,* Owoicho Adogwa, MD, MPH,* Christopher R. Brown, MD,t Carlos A. Bagley, MD,* William J. Richardson, MD,t Shivanand P. Lad, MD, PhD,* Maragatha Kuchibhatia, PhD,t and Oren N. Gottfried, MD*
Study Design. Retrospective cohort study. Objective. To evaluate the ability of local vancomycin powder to prevent deep wound infection after thoracolumbar and lumbar spinal fusion for open deformity cases. Summary of Background Data. Recent studies report that local delivery of vancomycin powder is associated with a decrease in spinal surgical site infection (SSI). This study compares deformity fusion cases before and after the routine application of spinal vancomycin powder. Metiiods. Posterior spinal deformity surgical procedures by a single institution were reviewed from January 2011 to April 2013. Routine application of vancomycin powder started in April 2012. Inclusion criteria included adult patients who underwent posterior fusion for deformity pathologies, including spondylolisthesis, kyphosis, sagittal imbalance, and scoliosis. Each cohort's baseline characteristics including infection risk factors, operative data, and rates of wound infection were compared. Associations between infection and vancomycin powder, with and without propensity score adjustment for risk factors were determined using logistic regression. Results. A total of 306 patients were included in the study All measured baseline and operative variables were statistically similar between untreated (n = 150) and those who received vancomycin powder (n = 156). No significant change in deep wound infection rate was seen between the control (5.3%) and intervention group (5.1%, P = 0.936). Logistic regression with and without propensity score adjusted for risk factors demonstrated that the use of vancomycin powder did not impact the development of SSI (odds ratio [95% confidence intervall: 1.01 [0.36-2.79], P = 0.9910)
From the *Division of Neurosurgery, Department of Surgery; tDivision of Orthopaedic Surgery, and ^Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC. Acknowledgment date: July 2, 2013. First revision date: September 27, 2013. Acceptance date: October 1, 2013. The device(s)/drug(s) is/are FDA-approved or approved by corresponding national agency for tbis indication. No funds were received in support of tbis work. Relevant financial activities outside the submitted work: consultancy payment for lecture, royalties. Address correspondence and reprint requests to Oren N. Gottfried, MD, Division of Neurosurgery, Department of Surgery, Duke University Medical Center, Box 3807, Durham, NC 27701; E-mail: [email protected] DOI: 10.1097/BRS.0000000000000071 Spine
and (odds ratio [95% confidence interval]: 0.87 [0.31-2.42], P = 0.7876), respectively. Conclusion. The local application of powdered vancomycin was not associated with a significant difference in the rate of deep SSI after spinal deformity surgery, and other treatment modalities are necessary to limit infection for this high-risk group. This study is in contrary to prior studies, which have reported a decrease in SSI with vancomycin powder. Key words: vancomycin powder, local antibiotics, wound infection, spine surgery, posterior lumbar fusion, deformity Levei of Evidence: 2 Spine 2014;39:177-184
T
he incidence of surgical site infections (SSIs) after spine surgery ranges from 0.6% to 12%.'"'" Despite the use of prophylactic systemic antibiotics and improved operative technique, postoperative SSIs remain of serious concern. Deep wound infection can have a profound impact on patients because they often require prolonged hospital stay, additional surgical interventions, lengthy intravenous antibiotic administration, and delayed rehabilitation.^-'' Furthermore, numerous studies have demonstrated elevated mortality risk in patients who experience a postoperative SSI.-'^-^ Given the incidence of deep wound infections, several authors explored and subsequendy published their results on locally applied prophylactic antibiotics.'"'' Intrawound vancomycin powder is an attractive option for additional prophylaxis against postoperative SSI, and initial studies have shown favorable results. Recently, the addition of local vancomycin powder to standard systemic antibiotic prophylaxis was shown to reduce SSI rates from 2.6%-15% to 0%-2.5%."-'''i*.i"« O'Neil et al'' reported a 13% reduction in the incidence of postoperative deep SSIs in patients undergoing posterior spinal stabilization of traumatic injuries. Similarly, Sweet et al^* observed a 2.4% decrease in the incidence of postoperative SSIs in patients undergoing elective spinal fusions using vancomycin powder prophylaxis. Although the theoretical advantage of locally applied vancomycin powder to standard systemic antibiotic prophylaxis is to reduce the incidence of SSIs, the results of puhlished studies on this subject are not all consistent, and it has not been explored for all high-risk populations.'" Prior www.spinejournal.com
177
Spine
SURGERY
Studies have not assessed the impact of local vancomycin for spinal deformity that involves longer incisions, increased operative time and blood loss, patients with more significant comorbidities, and higher infection rates.^"'-'^"^' Accordingly, this study compares postoperative deep surgical infection rates in a large continuous cohort of patients undergoing spinal deformity fusion surgery involving thoracolumbar and lumbar vertebrae before and after the routine application of locally applied vancomycin powder.
MATERIALS AND METHODS
Experience With Intrawound Vancomycin • Martin et al
affirmative or negative vancomycin usage were excluded from analysis. The surgical approach involved an open midline posterior incision for all patients. Minimally invasive cases were excluded. The decision to use subfascial drains were based on surgeon preference. However, in most cases, subfascial drains were used. In all cases, the wounds were closed with absorbable suture in the fascia and subcutaneous layers, and with staple or suture closure of the skin. After skin closure, incisions were cleaned again with chlorhexidine and a sterile dressing was applied. Dressings and drains were kept in place until the second postoperative day.
Patient Selection The primary aim of this study was to determine whether the addition of locally apphed vancomycin powder to standard systemic antibiotic prophylaxis would impact the incidence of postoperative deep wound infections for patients who underwent thoracolumbar or lumbar spinal deformity surgery by one of nine spine surgeons at our institution. The institutional review board approved this study. A retrospective review of hospital records from January 2011 to April 2013 was performed of adult patients older than 18 years who had undergone open posterior spinal fusion using pedicle screws and rod instrumentation that included lumbar spine (thoracolumbar, thoracosacral, iliothoracic, lumbar, lumbosacral, and iliolumbar) fusion for deformity cases. Inclusion causes included all types of deformity pathologies, including spondylolisthesis, kyphosis, sagittal imbalance, and scoliosis. Exclusion criteria included traumatic and neoplastic causes. Patient demographics, clinical presentation, comorbidities, radiological studies, and all treatment variables were reviewed for each case.
Standard Pre- and Postoperative Systemic Prophylactic Antibiotic Regiment All patients received standard systemic antibiotic prophylaxis consisting of IV cefazolin within 1 hour of surgical incision followed by IV cefazolin every 8 hours for 1 day. If the patient was allergic to penicillin, clindamycin was used instead. Other intravenous antibiotics that were used infrequently included: ciprofioxacin, piperacillin/tazobactam, or vancomycin. All patients were prepared with chlorhexidine. Fusion levels were determined by preoperative imaging. Before skin closure, irrigation with 2 L of normal saline by gravity was performed.
Treatment and Control Cohorts Two patient cohorts were developed. The treatment group consisted of consecutive patients who received 2 g of intraoperative vancomycin powder (April 2012-April 2013), spread throughout the local wound. The powder was placed directly on the muscle, fascia, and subcutaneous tissues taking care not to expose bone graft or dura. The control group consisted of consecutive patients who received their operation prior to the institutional use of vancomycin powder (January 2011-March 2012). Patients who lacked documentation of 178
www.spinejournal.com
Diagnostic Evaluation The definition of a deep SSI developed by the Genters for Disease Gontrol and Prevention were used to make a SSI diagnosis.^^ According to the Genters for Disease Gontrol and Prevention definitions, deep space incisional SSIs all occur within 30 days after surgery. Additionally, they involve purulent drainage, isolation of organism, signs or symptoms of infection (such as pain or tenderness, locaHzed swelling, redness, or heat) combined with positive culture results, and diagnosis by a surgeon or attending physician. Thus, for this study, a SSI was defined as being diagnosed during the initial hospitalization or during a hospital readmission or postoperative clinic appointment within 30 days of the surgery. All patients in this study had standard laboratory tests on admission to the hospital including: erythrocyte sedimentation rates, peripheral white blood cell counts, G-reactive protein, complete urine analysis, and microbiology and blood cultures. Erythrocyte sedimentation rate was determined by the Westergren method and considered abnormal if greater than 15 mm/hr. G-reactive protein was considered abnormal if greater than 5 mg/dL. Bacterial identification and susceptibility testing were performed according to the guidelines of the Genters for Disease Gontrol and Prevention.
Clinical Parameters Preoperative and intraoperative data for each patient were collected with use of patient charts and computerized medical records. Surgical infection risk factors were documented for each patient, including body mass index, obesity, diabetes mellitus, insulin use, chronic steroid use, prior lumbar surgery, chronic obstructive pulmonary disease, tobacco use, neurological deficit, immunodeficiency, and osteoporosis. Pertinent operative details were also collected, including number of levels instrumented, estimated blood loss, duration of anesthesia, intraoperative steroid use, blood transfusion, and durai tear.
Statistical Analysis Each cohort's baseline characteristics, operative data, and rates of wound infection were compared, using x^ tests for categorical variables and Student t tests for continuous variables. Associations between SSI and vancomycin powder, with and without propensity score adjustment for risk factors were determined using logistic regression. lanuary 2014
Spine
SURGERY
Experience With Intrawound Vancomycin • Martin et al
RESULTS A total of 306 adult patients undergoing deformity fusions were included in the study, with 150 patients (49 male, 101 female) in the untreated group and 156 patients (49 male, 107 female) in the vancomycin-treated group. Three patients were excluded because of incomplete records. Ninety-nine patients who underwent nondeformity instrumentations were not included. The untreated and treated groups were statistically similar (P > 0.05) with regard to baseline and operative
variables (Table 1). Baseline patient age was similar between both cohorts (untreated, 62.7; treated, 63.4; P = 0.618). Mean body mass index and prevalence of diabetes were similar between both cohorts (untreated, 29.6; treated, 30.0; P = 0.541) and (untreated, 18.0%; treated, 23.1%; P = 0.272). Similarly, tobacco use was not statistically different (untreated, 18.0%; treated, 12.8%; P = 0.209). The mean duration of anesthesia was similar between both cohorts (untreated, 322.4 min; treated, 328.1 min;
^^^^^^PmpârîsorTôîfetïëîrtnvîïïrDëfonTif^^ I ^ ^ ^ H B i l l ^ ^"d ^^^i" theijÉiÉHilHÉBn of VajHÍÉHin Powder j ^ ^ i ^ ^ ^ ^ ^ H ^ ^ H ^ ^ H ^ I • — Treated: 156 pts
P
62.7
63.4
0.618
27(18.0%)
36(23.1%)
0.272
7 (4.7%)
10 (6.4%)
0.506
Untreated: 150 pts
1
Baseline features Age (mean), yr Diabetes Diabetes, insulin BMI (mean)
29.6
30.0
0.541
Obesity
64 (42.7%)
74 (47.4%)
0.402
Previous lumbar fusion
47(31.3%)
50(32.1%)
0.893
Previous lumbar surgery
75 (50%)
72 (46.2%)
0.501
Current EtOH abuse
3 (2.0%)
6 (3.8%)
0.339
Current smoker
27(18.0%)
20(12.8%)
0.209
illicit drug user
1 (0.7%)
1 (0.6%)
0.978
CAD
33 (22.0%)
30(19.2%)
0.549
Osteoporosis
25(16.7%)
24(15.4%)
0.760
COPD
10(6.7%)
15(9.6%)
0.346
Current steroid use
10(6.7%)
10 (6.4%)
0.928
Hyperlipidemia
91 (60.7%)
80(51.3%)
0.098
Hypertension
105 (70.0%)
114(73.1%)
0.551
Neurological deficit
6 (4.0%)
5 (3.2%)
0 709
Immunodeficiency
1 (0.7%)
1 (0.6%)
0.978
4.5
5.1
0.059
EBL (mean), mL
1050.5
1314.7
0.101
Duration of anesthesia (mean), min
322.4
328.1
0.731
Use of postoperative drain
137(91.3%)
145(92.9%)
0.599
Intraoperative steroid use
38 (25.3%)
43 (27.6%)
0.658
Blood transfusion
75 (50.0%)
87 (55.8%)
0.312
14 (9.3%)
9 (5.8%)
0.237
8 (5.3%)
8(5.1%)
0.936
Operative details No. of levels instrumented (mean)
Durai tear Outcomes Deep SSI
pts indicates patients; BMI, body mass index; EtOH, ethanol; CAD, coronary artery disease; COPD, chronic obstructive pulmonarv disease- SSI surgical site infection. ' ' - s
Spine
wwvi/.spinejournai.com
179
Experience With Intrawound Vancomycin • JVIartin et al
SURGERY
P = 0.731). The mean estimated blood loss was smaller in the untreated cohort (untreated, 1050.5 mL; treated, 1314.7 mL; P = 0.101). Postoperative drains were used in a similar number of patients (untreated, 91.3%; treated, 92.9%; P = 0.599). Vancomycin-treated patients trended toward an increased number of surgical levels (untreated, 4.5; treated, 5.1; P = 0.059). Of interest, a similar number of patients from both cohorts experienced a durotomy (untreated, 9.3%; treated, 5.8%; P = 0.237). Both the untreated and treated groups had 8 deep SSIs (untreated, 5.3%; treated, 5.1%; P = 0.936) (Table 1; Figure 1). Small, nonsignificant differences were seen in patients with a previous lumbar surgery (n = 147; untreated, 6.7%; treated, 5.6%; P = 0.779), previous lumbar fusion (n = 97; untreated, 10.6%; treated, 8.0%; P = 0.654), and those with no history of lumbar surgery (n = 159; untreated, 4.0%; treated, 4.8%; P = 0.815). Lastly, when evaluating patients with diabetes, tobacco use, immunodeficiency, neurological deficits, myelomeningocele, or paralysis, there was not a significant difference in infection rate (n = 107; untreated, 5.5%; treated, 5.8%; P = 0.944). The mean number of postoperative days to infection was 15.1 days for the treated cohort and 15.6 days for the untreated cohort (P = 0.90) (Table 2). A comparison of other baseline or operative variables between untreated and treated patients with a deep SSI revealed no significant differences (Table 2). Likewise, there was no significant difference between choice of pre- and perioperative FV antibiotic in both groups: cefazolin (untreated, 74.7%; treated, 78.8%; P = 0.387), clindamycin (untreated, 16.7%; treated, 15.4%; P = 0.760), vancomycin (untreated, 8.0%; treated, 5.1%; P = 0.310), and single cases of ciprofioxacin and piperacillin/tazobactam. A detailed description of each SSI is shown in Tables 3 and 4 for vancomycin-treated and untreated patients, respectively. There was no significant difference in culture results for SSIs. The majority of documented infections in both groups were
CD ••a
o SP oS
Treated
Untreated
Figure 1. Incidence of postoperative surgical site infections in patients receiving intraoperative vancomycin powder in addition to standard systemic antibiotic prophylaxis. An equal number of patients in botb cohorts experienced a deep SSI. SSI indicates surgical site infection. 180
www.spinejournal.com
due to either gram-positive cocci (other than methicillin-resistant Staphylococcus aureus [MRSA]) (untreated, 4/8; treated, 3/8) or gram-negative rods (untreated, 4/8; treated, 4/8). There were 2 MRSA infections in the untreated group and 1 MRSA infection in the treated group. Logistic regression was used to determine the effect of use of vancomycin powder to predict SSI. The use of vancomycin powder was found to have no impact on the development of SSI (odds ratio [95% confidence interval]: 1.01 [0.36-2.79], P = 0.9910). Propensity to predict the usage of vancomycin powder was obtained using the preoperative baseline variables in Table 1. Propensity score adjusted for effect of use of vancomycin powder was not found to be a significant predictor for the development of SSI, as well (odds ratio [95% confidence interval]: 0.87 [0.31, 2.42], P = 0.7876).
DISCUSSION In addition to systemic antibiotics, spinal intrawound vancomycin is an attractive option. Local powdered vancomycin is inexpensive, has broad coverage against typical organisms responsible for deep spinal wounds, and seems to carry a lower risk of systemic complications than intravenous form. Most published data have supported its use, reporting decreased rates of postoperative wound infection and no adverse side effects.'-"-'* In cardiac surgery literature, Vander Salm et aP^ showed that topical vancomycin reduced the incidence of sternal wound infections from 3.6% to 0.45% (P = 0.02) after cardiac surgery. Lazar et aP addressed the concern of persistent serum levels of vancomycin after topical sternal application, that may contribute to the emergence of drug-resistant infections, by reporting a significant decrease in serum vancomycin levels from the night of surgery to the sixth postoperative day (P < 0.0001). In a rodent study, Ozcan et aP- used topical and intravenous vancomycin on rats with deep sternal wounds inoculated with MRSA. The group reported a significant (P < 0.05) reduction in MRSA wound organisms in 3 treated groups (topical, IV, and topical plus IV), with topical plus rV having the largest reduction. In neurosurgical and orthopedic literature, several studies have looked at the addition of local vancomycin to standard systemic antibiotic prophylaxis use in posterior spinal fusion surgery (Table 5). Cohort studies have reported that local vancomycin reduced SSI rates in adult thoracolumbar fusion surgery (n = 165-1732, 2.6%-12% vs. 0%-0.2%, P < O.OOl),'''-'** posterior fusion after traumatic causes (n = 110, 13% vs. 0%, P = 0.02),'^ and posterior cervical decompression and fusion surgical procedures (n = 112-171, 10.9%15% vs. 0%-2.5%, P < 0.05).'^''* Similarly, after introduction of vancomycin powder, Molinari ei fl/" reported a low 1.20% SSI rate (n = 663) in spine fusion surgery. In all studies, no adverse events were noted in any of the intervention groups associated with the use of vancomycin powder. Conversely, one group reported no significant change in postoperative deep wound infection. Mohammed et aP° compared open vascular procedures with and without intrawound vancomycin powder. In the 454 patient study, there was a January 2014
Spine
SURGERY
TABLE 2.
Experience With Intrawound Vancomycin • Martin et al
pomparison of Patients With Deförrffity UnderpingThoracolumbar anci L u m b a i ^ ^ H | Fusion With Postoperative Deep Wound Infection Before and After the Introductioii^ Untreated: 8 pts
Treated: 8 pts
P
15.6
15.1
0.90
67.0
57.3
0.27
Diabetes
2 (25.0%)
2 (25.0%)
1
Diabetes, insulin
1 (12.5%)
0 (0%)
0.30
32.0
32.8
0.86
Obesity
4 (50.0%)
5 (62.5%)
0.61
Previous lumbar fusion
5 (62.5%)
4 (50.0%)
0.61
Previous lumbar surgery
5 (62.5%)
4 (50.0%)
0.61
Current FtOH abuse
1 (12.5%)
0 (0%)
0.30
Current smoker
2 (25.0%)
2 (25.0%)
1
Illicit drug user
0 (0%)
0 (0%)
CAD
1 (12.5%)
2 (25.0%)
0.52
Osteoporosis
2 (25.0%)
1 (12.5%)
0.52
COPD
1 (12.5%)
1 (12.5%)
1
0 (0%)
1 (12.5%)
0.30
Hyperlipidemia
5 (62.5%)
5 (62.5%)
1
Hypertension
8(100%)
5 (62.5%)
0.055
Neurological deficit
0 (0%)
0 (0%)
Immunodeficiency
0 (0%)
0 (0%)
6.9
6.1
0.54
EBL (mean), mL
2831.3
2072.5
0.42
Duration of anesthesia (mean), min
420.5
481.3
0.48
Use of postoperative drain
8(100%)
8 (100%)
Intraoperative steroid use
1 (12.5%)
2 (25.0%)
0.52
Blood transfusion
7 (87.5%)
5 (62.5%)
0.25
Durai tear
1 (12.5%)
1 (12.5%)
1
Postoperative details Time to infection (mean), d Baseline features Age (mean), yr
BMI (mean)
Current steroid use
Operative details No. levels instrumented (mean)
pts indicates patients; BMI, body mass index; EtOH, ethanol; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease- FBL estimated blood loss. ' '
small but statistically significant decrease in the 30-day incidence of superficial infections (18.9% vs. 11.5%, P = 0.033), but no significant difference in the incidence of deep wound infections (6.1% vs. 5.7%, P = 0.692) or overall dehiscence rates (22.2% vs. 17.7%, P = 0.239). The group postulated that the blunted effect of local vancomycin on deep wounds could be due to the local vascular environment resulting in Spine
poor systemic absorption of the local antibiotic lessening its systemic effect on deep wound infections. Similarly, our study did not show a significant reduction in deep SSIs in posterior spinal deformity fusion cases (n = 306, 5.3% vs. 5.1%, P = 0.936). Our study involved a population of patients with spinal deformity that are at higher risk for infection due to very large incisions, greater time of surgery, www.spinejournal.com
181
Spine
Experience With Intrawound Vancomycin • Martin et al
SURGERY
"•
TAB LE 3.
•m
i
•nts From Treatmeriï (Vancomycin) Group With Deep Rostoperative Surgical
1
Infection Onset (d)
Fusion Levels
Tobacco Use
DM
BMl
Other Risk Factors
Culture
pt
Age Sex
1
42 F
Pseudarthrosis, spondylolisthesis
28
L3-S1
No
No
37.2
Prior fusion
Coag-neg Staph
2
63 F
Lumbar scoliosis, spondylolisthesis
30
L3-S1
Yes
No
26.2
COPD
E. cloacae
3
39 F
Pseudarthrosis, spondylolisthesis
13
L4-S1
No
Yes
41.3
Prior fusion
MSSA
4
67 F
Charcot spine with instability
14
L1-S1
Yes
Yes
52.9
5. marcescens
5
18F
Thoracolumbar scoliosis
4
T1-L2
No
No
20.1
S. marcescens
6
74 F
Spondylolisthesis, junctional kyphosis
12
TU-SI
No
No
30.8
Prior fusion, CAD
Coag-neg Staph
7
82 M
Thoracolumbar scoliosis
7
TiO-ilium
No
No
22.4
CAD
MRSA
8
73 F
Sagittal imbalance, pseudarthrosis
13
TU-SI
No
No
31.8
Prior fusion, steroid use, osteoporosis
K. pneumoniae, C. freundii
Diagnosis
p i indicates patients; DM, diabetes mellitus; BMl, body mass index; COPD, chronic obstructive pulmonary disease; CAD, coronary artery disease; MSSA, methiciilin-sensitive Staphyloeoccus aureus; Coag-neg Staph, Coaguiase-negative staphyiococci; F. cloacae, Fnterobacter cloacae; S. marcescens. Serrada marcescens; K. pneumoniae, Klebsiella pneumoniae; C freundii, Citrobacter freundii.
duration of anesthesia, blood loss,^ use of instrumentation,'' increased levels of instrumentation,^"-^' and other complications.' Because of these risk factors, patients with spinal deformity pose a greater risk of postoperative SSI. A possible theory for our contrary results as compared with other vancomycin studies is that local vancomycin antibiotic may not have enough capacity to overcome the increased infection risk factors seen in patients with deformity. Alternatively, most prior studies
used 1 g of vancomycin powder'^''^'^ and we averaged 1.75 g, which may not be sufficient to prevent infection for these larger incisions, and further studies are needed to evaluate this point. Finally, the addition of other treatment modalities may be needed to prevent SSIs in this high-risk population, such as improved nutrition or more stringent surgical selection criteria. This study was larger than several of the prior studies,''-"^""* and attempted to control for many spinal SSI risk factors
mBlHP
TABLE 4 . " Infection Onset (d)
Fusion Levels
Tobacco Use
DM
BMl
other Risk Factors
Culture
pt
Age, Sex
9
51 M
Spondylolisthesis, stenosis
14
L2-L5
No
No
41.9
Prior fusion
MSSA
10
63 F
Spondylolisthesis, sagittal imbalance
12
L3-L5
Yes
No
19.2
Osteoporosis
MSSA
11
62 F
Spondylolisthesis, sagittal imbalance
22
T12-L5
No
No
34.4
Prior fusion, COPD
Coag-neg Staph
12
67 F
Lumbar kyphosis, sagittal imbalance
7
T10-S1
Yes
Yes
28.7
Prior fusion
MRSA, E. aerogenes
13
74 M
Lumbar kyphosis, sagittal imbalance
8
T12-L5
No
No
32.6
Prior fusion, CAD
MRSA
14
79 M
Thoracolumbar scoliosis
26
T6-ilium
No
Yes
28.8
15
76 F
Thoracolumbar scoliosis
21
T6-ilium
No
No
28
Osteoporosis
Enterococcus, P. aeruginosa
16
64 F
Lumbar kyphosis, sagittal imbalance
15
T11-L5
No
No
42.1
Prior fusion
P. aeruginosa
Diagnosis
P. mirabilis
pt indicates patient; DM, diabetes mellitus; BMl, body mass index; CORD, chronic obstructive pulmonary disease; CAD, coronary artery disease; MSSA, methiciilin-sensitive Staphytococcus aureus; Coag-neg Staph, Coaguiase-negative staphylococci; E. aerogenes, Enterobacter aerogenes; P. aeruginosa, Pseudomonas aeruginosa; P. mirabilis, Proteus mirabilis.
182
www.spinejournal.com
January 2014
SURGERY
Experience With Intrawound Vancomycin • Martin et al
TABLE 5.
^^SMÏBlf«'#
Author and Year Sweet ef a/,''2011
Location Thoracic, lumbar
Patients With Fusion
Cause
Infection Rate (Cases With Fusion) Control: 2.6%
All posterior fusion cases
1732 Treated: 0.2% ( P < 0.0001)
O'Neill eta/,''2011
Cervical, thoracic, lumbar
Posterior trauma fusion cases
110
Molinari ef a/," 2012
Cervical, thoracic, lumbar
All spine cases
663
Strom ei a/,'»2013
Lumbar
All posterior cases
165
Control: 13% Treated: 0% (P = 0.02) Treated: 1.20% Control: 12% Treated: 0% (P = 0.0008)
Strom eia/,'*2013
Cervical
All posterior fusion cases
171
Control: 10.9% Treated: 2.5% (P = 0.0384)
Caroomefa/,"'2013
Cervical
Posterior multilevel spondylotic myelopathy fusion cases
including diabetes, steroid use, previous spinal surgery, obesity, smoking, duration of anesthesia, estimated blood loss, drug use,' blood transfusion, coronary artery disease, osteoporosis, durai tear, and chronic obstructive pulmonary disease.' Vancomycin-treated patients trended toward higher blood loss and increased number of levels treated. These factors would be expected to increase the risk of deep infection in the treatment group. However, both groups of 8 patients with postoperative infections had a balanced distribution of chnical and operative variables (Table 2), and our logistic regression analysis controlling for risk factors and operative details did not show an impact for vancomycin in affecting incidence of infection. Furthermore, we reviewed clinical records, operative and nursing notes, and surveyed all surgeons and did not identify any changes in practice patterns or external factors that changed during these two time periods before and after the routine use of vancomycin powder. Simple linear regression analysis by month showed no significant relationship between infection rate and date of operation (P = 0.714). The limitations inherent in our study have implications for its interpretation. A power analysis demonstrated that there was sufficient sample size to detect a change in the rate of infection seen in prior studies using vancomycin powder. However, given that the event rate for postoperative infection is small, this study is not powered to detect subtle differences. This study was limited by its retrospective design, focus on deformity diseases, and resulting confounding factors that may have been undetected. We also focused on deep wound complications within the 30-day postoperative period, not accounting for superficial or possible late infections. Also, the study involved 9 surgeons that have similar but nonexact techniques. Note the distribution of infections showed that no one surgeon had a significantly greater infection rate. Also, both cohorts centered on the introduction of vancomycin powder, which initially began because of surgeon's discretion and may have introduced selection bias. Additionally, serum vancomycin levels were not measured, which could have Spine
112
Control: 15% Treated: 0% (P = 0.007)
aided in proper dosing and assessing its duration of impact. However, in our study, we did not detect toxicity or complications from the routine use of vancomycin powder. Finally, drains were common in this series, and could have drained or limited the vancomycin concentration.''' Despite these limitations, our study demonstrated that the use of locally applied vancomycin powder did not reduce the incidence of 30-day postoperative deep SSI rate in spinal deformity fusion. This study is in contrary to prior spinal studies, which have reported a decrease in SSI with the addition of perioperative intrawound vancomycin powder.
CONCLUSION Despite our conclusions, our institution is continuing to use and investigate the use of intrawound spinal vancomycin powder because our sample size is too small to detect any subtle but beneficial effects of this intervention. We continue to use 2 g for larger surgical procedures, but also question whether a larger dose may be beneficial. A randomized prospective study will be performed to corroborate our findings.
Key Points •
In this retrospective review of 306 patients undergoing posterior thoracolumbar or lumbar deformity fusion, the infection rate did not significantly change (untreated 5.3% vs. treated 5.1%, P = 0.936) after the application of local vancomycin powder.
•
The lack of efficacy of intrawound vancomycin to reduce surgery site infections for spinal deformity is contrary to results for other spine surgical procedures demonstrating that it may not be able to overcome the increased infection risk factors seen in patients with deformity or that increased dosages may be needed. www.spinejournal.com
183
Spine •
SURGERY
Further prospective studies are needed to evaluate if intrawound vancomycin powder combined with other treatnnent modalities may benefit patients undergoing spinal deformity surgery.
References 1. Pull ter Gunne AF, van Laarhoven CJ, Cohen DB. Surgical site infection after osteotomy of the adult spine: does type of osteotomy matter? Spine} 20W;W:410-6. 2. Kirkland KB, Briggs JP, Trivette SL, et al. The impact of surgicalsite infections in the 1990s: attributable mortality, excess length of hospitalization, and extra costs. Infect Control Hosp Epidemiol 1999;20:725-30. 3. Koutsoumbelis S, Hughes AP, Girardi FP, et al. Risk factors for postoperative infection following posterior lumbar instrumented arthrodesis. / Bone Joint Surg Am 2011;93:1627-33. 4. Massie JB, Heller JG, Abitbol JJ, et al. Postoperative posterior spinal wound infections. Clin Orthop Relat Res 1992;99-108. 5. Wimmer C, Gluch H, Franzreb M, et al. Predisposing factors for infection in spine surgery: a survey of 850 spinal procedures. / S/?¿nal Disord I998;n:124-S. 6. Whitehouse JD, Friedman ND, Kirkland KB, et al. The impact of surgical-site infections following orthopedic surgery at a community hospital and a university hospital: adverse quality of life, excess length of stay, and extra cost. Infect Control Hosp Epidemiol 2002;23:\83-9. 7. Klevens RM, Edwards JR, Richards GL, Jr, et al. Estimating health care-associated infections and deaths in U.S. hospitals, 2002. Public Health Rep 2007;122:160-6. 8. Anderson DJ, Kaye KS, Glassen D, et al. Strategies to prevent surgical site infections in acute care hospitals. Infect Control Hosp £p;ciem;o/2008;29(suppl 1):S51-61. 9. Lazar HL, Barlam T, Gabral H. The effect of topical vancomycin applied to sternotomy incisions on postoperative serum vancomycin levels. / Card Surg 2011;26:461-5. 10. Mohammed S, Pisimisis GT, Daram SP, et al. Impact of intraoperative administration of local vancomycin on inguinal wound complications. ] Vase Surg 2013;57:1079-83.
184
www.spinejournal.com
Experience With Intrawound Vancomycin • Martin et al
11. Molinari RW, Khera OA, Molinari WJ. Prophylactic intraoperative powdered vancomycin and postoperative deep spinal wound infection: 1,512 consecutive surgical cases over a 6-year period. Eur Spi«e/ 2012;21(suppl 4):S476-82. 12. Ozcan AV, Demir M, Onem G, et al. Topical versus systemic vancomycin for deep sternal wound infection caused by methicillinresistant Staphylococcus aureus in a rodent experimental model. Tex Heart Inst] 2006;33:107-10. 13. Strom RG, Pacione D, Kalhorn SP, et al. Decreased risk of wound infection after posterior cervical fusion with routine local application of vancomycin powder. Spine (Phila Pa 1976) 2013. 14. Sweet FA, Roh M, Sliva C. Intrawound application of vancomycin for prophylaxis in instrumented thoracolumbar fusions: efficacy, drug levels, and patient outcomes. Spine (Phila Pa 1976) 2011;36:2084-8. 15. Vander Salm TJ, Okike ON, Pasque MK, et al. Reduction of sternal infection by application of topical vancomycin. / Thorac Cardiovasc Surg 19S9;9S:618-22. 16. Garoom G, Tullar JM, Benton FG, Jr, et al. Intrawound vancomycin powder reduces surgical site infections in posterior cervical fusion. Spine (Phila Pa 1976) 2013. 17. O'Neill KR, Smith JG, Abtahi AM, et al. Reduced surgical site infections in patients undergoing posterior spinal stabilization of traumatic injuries using vancomycin powder. Spine ] 2011;ll: 641-6. 18. Strom RG, Pacione D, Kalhorn SP, et al. Lumbar laminectomy and fusion with routine local application of vancomycin powder: decreased infection rate in instrumented and non-instrumented cases. Clin Neurol Neurosurg 2013. 19. Klein JD, Hey LA, Yu GS, et al. Perioperative nutrition and postoperative complications in patients undergoing spinal surgery. Spine (Phila Pa 1976) 1996;21:2676-82. 20. Olsen MA, Nepple JJ, Riew KD, et al. Risk factors for surgical site infection following orthopaedic spinal operations. / Bone Joint Surg Am 2008;90:62-9. 21. Schimmel JJ, Horsting PP, de Kleuver M, et al. Risk factors for deep surgical site infections after spinal fusion. Eur Spine J 2010;19:1711-9. 22. Mangram AJ, Horan TG, Pearson ML, et al. Guideline for prevention of surgical site infection, 1999. Hospital Infection Gontrol Practices Advisory Gommittee. Infect Control Hosp Epidemiol 1999;20:250-78; quiz 279-80.
January 2014
Copyright of Spine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
