Surgical site infections (SSIs) are one of the most frequent complications following spine surgery with a reported incidence in the range of 10% depending on case complexity.1–7 SSIs increase the risk of additional peri-operative morbidity and mortality and can have an adverse effect on functional outcome. They also place a burden on the health care system with prolonged hospitalizations, reoperations, and increased utilization of outpatient care and resources.6–10 Increasingly, Methicillin-resistant Staphylococcus aureus (MRSA) infections are recognized as a plurality of SSIs following spine surgery. These infections are notoriously difficult to eradicate and many culminate in chronic pain and other disabilities for individuals who sustain these types of complications.11–13
Since the early 1990s, the importance of preoperative antibiotic administration has been universally accepted with current recommendations advocating completion of the prophylactic regimen within 1 hour of surgical incision.14–16 Between 1990 and 2010, our institution employed a standard protocol for preoperative antibiotic prophylaxis consisting of 1 g of cefazolin unless contraindicated due to allergy. In 2010, our institution internally reviewed SSIs for all disciplines. The unpublished data indicated that a substantial portion of SSIs resulted from MRSA across orthopedic subspecialties, including joint arthroplasty and spine surgery. Following a collaboration between specialties including orthopedics, neurosurgery, anesthesia, and infectious disease, a new regimen consisting of intravenous vancomycin in addition to cefazolin was promulgated as a standard prophylactic protocol. The protocol applied to all patients undergoing orthopedic surgeries requiring placement of any surgical implants.
Vancomycin (weight-based dosing) and 1 g cefazolin were administered within 1 hour of skin incision and were given for 24 hours during the perioperative period. If the patient had an allergy to vancomycin, cefazolin, or any antibiotic with cross reactivity, clindamycin was used as a replacement. As the protocol was uniformly applied at the institutional level in 2010, this allowed for an opportunity to study the efficacy of this intervention in the form of a natural experiment. The ultimate impact of this change in prophylactic antibiotic regimen on SSIs following instrumented spinal fusions remains unknown at present, as do consequent alterations in the organisms responsible for postsurgical infections in the period following the standardized use of preoperative vancomycin for all patients.
In this context, we sought to assess whether administration of prophylactic vancomycin in addition to cefazolin decreased revision surgeries for postoperative infection as well as the need for revisions as a whole. We hypothesized that there would be no alteration in infection rates following introduction of the new prophylactic protocol. As far as we are aware, this is the first investigation to consider the effect of a standardized regimen incorporating vancomycin as a prophylactic agent in patients undergoing instrumented spinal fusion.
MATERIALS AND METHODS
We queried our institutional registry,17–21 using International Classification of Diagnosis, 9th revision (ICD-9) procedure codes to identify all clean, nonrevision, instrumented spine surgical procedures performed between 2005 and 2015. Implemented in 1991, our hospital registry is a comprehensive repository of all clinical and surgical encounters and has been used in the past in investigations regarding orthopedic practice,18 spine surgical care,17,19,20 and healthcare policy.20,21 The means through which data is collected and accessed from this registry has been extensively described in prior research.17–21
We conducted a prepost analysis evaluating the effect of the mandated change in antibiotic prophylaxis regimen at the institutional level on SSIs following instrumented spinal fusion surgery. As the change in antibiotic prophylaxis protocol was implemented in 2010 we collected data on all eligible patients over the course of 5 years (2005–2009) prior and 5 years (2011–2015) following the initiation of this protocol. Cases performed in 2010 were excluded to allow for a wash-out period. We further excluded patients who did not complete 12-months of postoperative surveillance, defined as an office visit or other healthcare encounter within a 30-day window on either side of the 1-year time point following surgery.
All individuals eligible for inclusion had their medical records abstracted and age at the time of surgery, biologic sex, number of comorbidities classified using the Deyo modified Charlson score,22 and preoperative diagnosis were recorded. Location of the surgery, approach, intraoperative complications, and the use of vancomycin powder were also captured. Patients were surveyed for the occurrence of a revision spine surgical procedure for any cause, as well as revision surgeries specifically performed for a diagnosis of infection. Revision surgical procedures were considered to be any intervention that warranted a return to the operating room including irrigation and debridement, removal of instrumentation, and revision fusions. In the event of a revision surgery performed for a diagnosis of infection, intraoperative culture results were also obtained. Culture results were categorized as MRSA, Methicillin-sensitive Staphylococcus aureus (MSSA), or other organisms including gram-negative organisms, multiple organisms, and multidrug-resistant organisms.
In this investigation, the primary outcome consisted of revision spine surgical intervention performed for a diagnosis of infection. Revision spine surgical interventions performed for any cause were considered secondarily. The rationale behind this approach was that ensuring consideration of all revision surgical procedures would allow for increased sensitivity, ensuring that revisions performed for an occult infection, or where diagnosis for an SSI was incorrectly recorded would not be summarily missed. The institutional prophylactic antibiotic protocol at the time of surgery was the primary predictor variable: cefazolin alone for 2005 to 2009 and cefazolin with vancomycin for 2011 to 2015. Logistic regression analysis was utilized to adjust for confounding. We evaluated initial unadjusted results for the prophylactic antibiotic protocol on the outcomes of interest followed by sequential adjustments for case-mix (sociodemographic factors and diagnosis) and surgical characteristics (location and surgical approach).23 Changes in culture results for individuals treated for a SSI were evaluated using chi-square analysis.
In all statistical tests age was considered continuously with comorbid conditions categorized as 0, 1, and ≥2. Preoperative diagnosis was classified as degenerative (e.g., degenerative disc disease, stenosis, spondylolisthesis), trauma, and other. Location of the surgical intervention was defined as cervical, thoracic or lumbosacral and approach classified as anterior, posterior, or combined. Regression results were presented using odds ratios (ORs) and 95% confidence intervals (CI) along with P values. Significant results were considered to be those with OR and 95% CIs exclusive of 1.0 with P < 0.05. Discrimination and calibration of final logistic models were assessed using the c-statistic and Hosmer–Lemeshow goodness of fit test, respectively.23 Statistical testing was performed using STATA 15.0 (STATA Corp, College Station, TX). This investigation was deemed exempt by our institutional IRB prior to commencement.
We identified 3231 patients eligible for inclusion with 1275 treated in 2005 to 2009 and 1956 in 2011 to 2015. The average age of the cohort as a whole was 56.6 (SD 14.6) with 56% (n = 1809) female and 46% (n = 1496) considered smokers. Seventy-seven percent (n = 2497) of the entire population received surgery for a degenerative disorder. The mean postoperative surveillance was 25.4 months (range 12 mo to 13 yr). Patients treated in 2011 to 2015 were found to be older, more likely to smoke (Table 1) and had a greater proportion of lumbosacral surgeries (Table 2). Individuals who received surgery in 2011 to 2015 were also less likely to experience an intraoperative complication (2% vs. 3% in 2005–2009; P = 0.004). The use of vancomycin powder was documented in 145 cases (8%) in 2011 to 2015, but was not encountered for any procedures in the years 2005 to 2009.
Revision surgeries performed for a diagnosis of infection were reduced from a rate of 4% (n = 57) in the period 2005 to 2009 to 2% (n = 44) over 2011 to 2015 (P < 0.001; Table 2). At the same time, the incidence of revision surgeries for any cause was also reduced (14% in 2005–2009 vs. 9% in 2011–2015; P < 0.001). MRSA was identified in 25% (n = 14) of SSIs in 2005 to 2009 and 28% (n = 12) in 2011 to 2015, while MSSA occurred in 23% (n = 13) and 23% (n = 10) over the same time-periods, respectively.
Unadjusted estimates determined that the change in prophylactic antibiotic regimen reduced the odds of a revision surgery for infection by 51% (OR 0.49; 95% CI 0.33, 0.73; P = 0.001; Table 3). This determination was relatively unchanged following adjustment for case mix (OR 0.50; 95% CI 0.33, 0.75; P = 0.001) as well as case mix and surgical characteristics (OR 0.50; 95% CI 0.33, 0.76; P = 0.001). The final model was well calibrated to the data (P = 0.75) and explained 77% of the variation in revision surgery for infection. Furthermore, no significant difference in the organisms responsible for SSI was identified between the two time periods (P = 0.22).
Regarding revisions performed for any cause, our unadjusted estimates identified a 36% reduction in the odds of such interventions associated with the change in prophylactic antibiotic regimen (OR 0.64; 95% CI 0.51, 0.80; P < 0.001; Table 3). This estimate similarly remained unaltered after adjusting for differences in case mix alone (OR 0.64; 95% CI 0.51, 0.80; P < 0.001) as well as case mix and surgical characteristics (OR 0.63; 95% CI 0.50, 0.79; P < 0.001). The final model was also well calibrated to the data (P = 0.48) and explained 69% of the variation in the incidence of revision surgery.
In recent decades, numerous interventions have been trialed as a means of reducing SSIs following instrumented spinal fusion. These have ranged from universal screening of MRSA carriers preoperatively, to the application of vancomycin powder to the surgical bed and the use of antibiotic infused irrigation prior to wound closure.1–3,24,25 In most instances, these practices vary between surgeons, even among individuals who operate at the same institution.1–3,24 Our study involves a prepost analysis of institutional data from the years 2005 to 2009 and 2011 to 2015. This natural experiment was engendered by a change in the standard preoperative antibiotic prophylaxis regimen at our hospital in 2010 that formalized the use of intravenous vancomycin in conjunction with cefazolin.
We were able to include over 3000 individuals treated with an instrumented spinal fusion in this analysis. The sociodemographic and surgical characteristics are in line with other reports in the literature as are the rates of SSIs and revision surgeries as a whole.1–3,6,7,11,24–26 These facts speak to the external validity of our results and indicate the potential for translation to other centers. Following implementation of the new antibiotic prophylaxis regimen, our results indicate that the rate of SSIs requiring a surgical procedure decreased by 50%, while revisions as a whole were reduced by 36%. Importantly, the standardized use of vancomycin did not result in an inordinate increase in SSIs caused by atypical bacteria or multidrug-resistant organisms.
While our models were well calibrated to the source data, only 76% of the variation in revision surgeries for SSIs was explained by the factors considered here. Thus, we cannot attribute the observed reduction in infection rates entirely to the change in preoperative antibiotic prophylaxis. Further, although intended to primarily impact the prevalence of SSIs associated with MRSA, the rates of MRSA and MSSA associated infections were unchanged between the time periods under study. Although there was no significant difference in the organisms responsible for SSI, overall, the absolute reductions were greatest in the heterogenous other category, which includes atypical bacteria, slow growing organisms, and multidrug-resistant pathogens. This may indicate that a combined vancomycin/cefazolin protocol could exert prophylaxis for infectious agents beyond MRSA and MSSA.
These realities speak to a multifactorial etiology for the observed postprotocol reduction in SSIs, including secular trends in infection prevention measures, surgical techniques, and the application of vancomycin powder to the surgical field.1–3,24,25 As the use of vancomycin powder was restricted to the time-period in which the vancomycin and cefalozin regimen was in effect, it could not be adjusted for in our regression tests. The fact that vancomycin powder was employed in less than 10% of cases in the postprotocol period, however, mitigates its role as a substantial confounder in this setting. While vancomycin powder has demonstrated some efficacy in terms of reducing wound infection rates in spine surgery,1–3,24 this appears to have been confined to selective application in high-risk cases. A randomized trial of more than 900 patients treated with vancomycin powder as compared with systemic prophylaxis found no significant difference in SSIs.25 This is the first effort we are aware of to consider a uniform institutional protocol that employs the use of intravenous vancomycin and cefazolin as prophylactic agents. The protocol appears to be both safe and effective and did not lead to a significant alteration in the spectrum of organisms responsible for SSIs at our hospital.
We acknowledge several limitations associated with this work, including its retrospective nature and the fact that the study relied on chart abstraction. Errors in reporting or coding could have resulted in an information bias. Additionally, individuals who transferred their care to other facilities or required urgent treatment at an outside center would not have been documented as sustaining a complication in this analysis. Clinically silent infections, or those managed with antibiotics alone in an outpatient setting could similarly have escaped notice given our methodologic approach. Thus, the documented incidence of revision surgery or SSIs may be underestimates, although the fact that the rates are comparable to those reported in the literature1–3,6,7,11,24–26 is reassuring to some extent. The investigation could likewise be underpowered to detect meaningful differences in the organisms responsible for SSIs. To some extent revision procedures are a time varying measure that may increase with time. The fact that revision rates for all causes decreased from 9% to 7% in the period under study while a more robust reduction was appreciated in revision surgery for SSIs indicates that improvement in this arena is probably more than a secular trend. We were limited to the inclusion of variables captured in our institutional registry and electronic medical records and cannot consider other facets such as surgeon experience, manner of training, the role of trainees in surgical care and decision making, as well as decision making around the need for a revision procedure.17–21 Last, as a quaternary academic referral center, findings specific to our institution may not necessarily be translatable to other facilities, especially hospitals that have a markedly different profile of organisms responsible for SSIs or those that perform a dramatically different scope of spine surgical procedures.
Despite these limitations, we believe that this natural experiment has shown some utility for a preoperative prophylactic antibiotic regimen of vancomycin and cefazolin, including meaningful reductions in revision procedures, as well as those performed for SSI. While encouraging in these respects, the decision to adopt this practice at other institutions likely should be based on careful consideration of the types of organisms commonly responsible for SSIs, the overall infection rate, and the instrumented spine surgical procedures regularly performed. In the event that such an approach is selected, our institutional experience supports that the adoption of a vancomycin/cefazolin protocol is safe and does not adversely impact the organism profile following implementation.
Key PointsWe conducted a prepost analysis (2005–2009 vs. 2011–2015) evaluating the effect of the mandated use of intravenous vancomycin and cefazolin on SSIs following instrumented spinal fusions that was implemented in 2010.Revision surgeries performed for a diagnosis of infection were reduced from a rate of 4% (n = 57) in the period 2005 to 2009 to 2% (n = 44) over 2011 to 2015 (P < 0.001). At the same time, the incidence of revision surgeries for any cause was also reduced (14% in 2005–2009 vs. 9% in 2011–2015; P < 0.001).In adjusted analysis, the odds of a revision procedure for SSI were reduced by 50% following introduction of the protocol (OR 0.50; 95% CI 0.33, 0.76).We found some utility for a preoperative prophylactic antibiotic regimen of vancomycin and cefazolin, including meaningful reductions in revision procedures performed for SSIs.
1. Radcliff KE, Neusner AD, Millhouse PW, et al. What is new in the diagnosis and prevention of spine surgical site infections. Spine J
2. Pahys JM, Pahys JR, Cho SK, et al. Methods to decrease postoperative infections following posterior cervical spine surgery
. J Bone Joint Surg Am
3. Van Hal M, Lee J, Laudermilch D, et al. Vancomycin
powder regimen for prevention of surgical site infection
in complex spine surgeries. Clin Spine Surg
4. Coello R, Charlett A, Wilson J, et al. Adverse impact of surgical site infections in English hospitals. J Hosp Infect
5. Kirkland KB, Briggs JP, Trivette SL, et al. The impact of surgical-site infections in the 1990 s: attributable mortality, excess length of hospitalization, and extra costs. Infect Control Hosp Epidemiol
6. Pak LM, Fogel HA, Chaudhary MA, et al. Outpatient spine clinic utilization is associated with reduced emergency department visits following spine surgery
. Spine (Phila Pa 1976)
2018; 43: E836–E841.
7. Schoenfeld AJ, Carey PA, Cleveland AW 3rd, et al. Patient factors, comorbidities, and surgical characteristics that increase mortality and complication risk after spinal arthrodesis: a prognostic study based on 5,887 patients. Spine J
8. Stone PW. Economic burden of healthcare-associated infections: an American perspective. Expert Rev Pharmacoecon Outcomes Res
9. Abdul-Jabbar A, Berven SH, Hu SS, et al. Surgical site infections in spine surgery
. Spine (Phila Pa 1976)
10. Messina AF, Berman DM, Ghazarian SR, et al. The management and outcome of spinal implant-related infections in pediatric patients: a retrospective review. Pediatr Infect Dis J
11. Pull ter Gunne AF, van Laarhoven CJ, Cohen DB. Incidence of surgical site infection
following adult spinal deformity surgery: an analysis of patient risk. Eur Spine J
12. Moran GJ, Krishnadasan A, Gorwitz RJ, et al. Methicillin-resistant S. aureus infections among patients in the emergency department. N Engl J Med
13. Panlilio AL, Culver DH, Gaynes RP, et al. Methicillin-resistant Staphylococcus aureus in U.S. hospitals, 1975-1991. Infect Control Hosp Epidemiol
14. Classen DC, Evans RS, Pestotnik SL, et al. The timing of prophylactic administration of antibiotics and the risk of surgical-wound infection. N Engl J Med
15. Steinberg JP, Braun BI, Hellinger WC, et al. Trial to Reduce Antimicrobial Prophylaxis Errors (TRAPE) Study Group. Timing of antimicrobial prophylaxis and the risk of surgical site infections: results from the Trial to Reduce Antimicrobial Prophylaxis Errors. Ann Surg
16. Van Kasteren ME, Mannien J, Ott A, et al. Antibiotic prophylaxis and the risk of surgical site infections following total hip arthroplasty: timely administration is the most important factor. Clin Infect Dis
17. Schoenfeld AJ, Harris MB, McGuire KJ, et al. Mortality in elderly patients with hyperostotic disease of the cervical spine following fracture: An age- and sex-matched study. Spine J
18. Janssen SJ, van der Heijden AS, van Dijke M, et al. 2015 Marshall Urist Young Investigator Award: prognostication in patients with long bone metastases: does a boosting algorithm improve survival estimates? Clin Orthop Relat Res
19. Schoenfeld AJ, Bono CM, Reichmann WM, et al. Type II odontoid fractures of the cervical spine: do treatment type and medical co-morbidities affect mortality in elderly patients? Spine (Phila Pa 1976)
20. Schoenfeld AJ, Tobert DG, Le HV, et al. Utility of adding magnetic resonance imaging to computed tomography alone in the evaluation of cervical spine injury: a propensity matched analysis. Spine (Phila Pa 1976)
21. Tobert DG, Le HV, Blucher JA, et al. The clinical implications of adding computed tomography angiography in the evaluation of cervical spine fractures: a propensity matched analysis. J Bone Joint Surg Am
2018; accepted for publication.
22. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol
23. Long JS, Freese J. Regression Models for Categorical Dependent Variables Using STATA. 2nd Edition. College Station, TX: STATA Press; 2006.
24. Hill BW, Emohare O, Song B, et al. The use of vancomycin
powder reduces surgical reoperation in posterior instrumented and noninstrumented spinal surgery. Acta Neurochir (Wien)
25. Tubaki VR, Rajasekaran S, Shetty AP. Effects of using intravenous antibiotic only versus local intrawound vancomycin
antibiotic powder application in addition to intravenous antibiotics on postoperative infection in spine surgery
in 907 patients. Spine (Phila Pa 1976)
26. Golinvaux NS, Basques BA, Bohl DD, et al. Comparison of 368 patients undergoing surgery for lumbar degenerative spondylolisthesis from the SPORT trial with 955 from the NSQIP database. Spine (Phila Pa 1976)
Keywords:Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.
MRSA; spinal fusion; spine surgery; surgical antibiotic prophylaxis; surgical site infection; vancomycin