Risk factors for deep surgical site infection following thoracolumbar spinal surgery

Free access

OBJECTIVE

Surgical site infection (SSI) following spine surgery causes major morbidity and greatly impedes functional recovery. In the modern era of advanced operative techniques and improved perioperative care, SSI remains a problematic complication that may be reduced with institutional practices. The objectives of this study were to 1) characterize the SSI rate and microbial etiology following spine surgery for various thoracolumbar diseases, and 2) identify risk factors that were associated with SSI despite current perioperative management.

METHODS

All patients treated with thoracic or lumbar spine operations on the neurosurgery service at the University of California, San Francisco from April 2012 to April 2016 were formally reviewed for SSI using the National Healthcare Safety Network (NHSN) guidelines. Preoperative risk variables included age, sex, BMI, smoking, diabetes mellitus (DM), coronary artery disease (CAD), ambulatory status, history of malignancy, use of preoperative chlorhexidine gluconate (CHG) showers, and the American Society of Anesthesiologists (ASA) classification. Operative variables included surgical pathology, resident involvement, spine level and surgical technique, instrumentation, antibiotic and steroid use, estimated blood loss (EBL), and operative time. Multivariable logistic regression was used to evaluate predictors for SSI. Odds ratios and 95% confidence intervals were reported.

RESULTS

In total, 2252 consecutive patients underwent thoracolumbar spine surgery. The mean patient age was 58.6 ± 13.8 years and 49.6% were male. The mean hospital length of stay was 6.6 ± 7.4 days. Sixty percent of patients had degenerative conditions, and 51.9% underwent fusions. Sixty percent of patients utilized presurgery CHG showers. The mean operative duration was 3.7 ± 2 hours, and the mean EBL was 467 ± 829 ml. Compared to nonfusion patients, fusion patients were older (mean 60.1 ± 12.7 vs 57.1 ± 14.7 years, p < 0.001), were more likely to have an ASA classification > II (48.0% vs 36.0%, p < 0.001), and experienced longer operative times (252.3 ± 120.9 minutes vs 191.1 ± 110.2 minutes, p < 0.001). Eleven patients had deep SSI (0.49%), and the most common causative organisms were methicillin-sensitive Staphylococcus aureus and methicillin-resistant S. aureus. Patients with CAD (p = 0.003) or DM (p = 0.050), and those who were male (p = 0.006), were predictors of increased odds of SSI, and presurgery CHG showers (p = 0.001) were associated with decreased odds of SSI.

CONCLUSIONS

This institutional experience over a 4-year period revealed that the overall rate of SSI by the NHSN criteria was low at 0.49% following thoracolumbar surgery. This was attributable to the implementation of presurgery optimization, and intraoperative and postoperative measures to prevent SSI across the authors’ institution. Despite prevention measures, having a history of CAD or DM, and being male, were risk factors associated with increased SSI, and presurgery CHG shower utilization decreased SSI risk in patients.

ABBREVIATIONS ASA = American Society of Anesthesiologists; CAD = coronary artery disease; CHG = chlorhexidine gluconate; CI = confidence interval; DM = diabetes mellitus; EBL = estimated blood loss; LOS = length of stay; MIS = minimally invasive surgery; MRSA = methicillin-resistant Staphylococcus aureus; MRSE = methicillin-resistant Staphylococcus epidermidis; MSSA = methicillin-sensitive S. aureus; MSSE = methicillin-sensitive S. epidermidis; NHSN = National Healthcare Safety Network; OR = odds ratio; SSI = surgical site infection.

OBJECTIVE

Surgical site infection (SSI) following spine surgery causes major morbidity and greatly impedes functional recovery. In the modern era of advanced operative techniques and improved perioperative care, SSI remains a problematic complication that may be reduced with institutional practices. The objectives of this study were to 1) characterize the SSI rate and microbial etiology following spine surgery for various thoracolumbar diseases, and 2) identify risk factors that were associated with SSI despite current perioperative management.

METHODS

All patients treated with thoracic or lumbar spine operations on the neurosurgery service at the University of California, San Francisco from April 2012 to April 2016 were formally reviewed for SSI using the National Healthcare Safety Network (NHSN) guidelines. Preoperative risk variables included age, sex, BMI, smoking, diabetes mellitus (DM), coronary artery disease (CAD), ambulatory status, history of malignancy, use of preoperative chlorhexidine gluconate (CHG) showers, and the American Society of Anesthesiologists (ASA) classification. Operative variables included surgical pathology, resident involvement, spine level and surgical technique, instrumentation, antibiotic and steroid use, estimated blood loss (EBL), and operative time. Multivariable logistic regression was used to evaluate predictors for SSI. Odds ratios and 95% confidence intervals were reported.

RESULTS

In total, 2252 consecutive patients underwent thoracolumbar spine surgery. The mean patient age was 58.6 ± 13.8 years and 49.6% were male. The mean hospital length of stay was 6.6 ± 7.4 days. Sixty percent of patients had degenerative conditions, and 51.9% underwent fusions. Sixty percent of patients utilized presurgery CHG showers. The mean operative duration was 3.7 ± 2 hours, and the mean EBL was 467 ± 829 ml. Compared to nonfusion patients, fusion patients were older (mean 60.1 ± 12.7 vs 57.1 ± 14.7 years, p < 0.001), were more likely to have an ASA classification > II (48.0% vs 36.0%, p < 0.001), and experienced longer operative times (252.3 ± 120.9 minutes vs 191.1 ± 110.2 minutes, p < 0.001). Eleven patients had deep SSI (0.49%), and the most common causative organisms were methicillin-sensitive Staphylococcus aureus and methicillin-resistant S. aureus. Patients with CAD (p = 0.003) or DM (p = 0.050), and those who were male (p = 0.006), were predictors of increased odds of SSI, and presurgery CHG showers (p = 0.001) were associated with decreased odds of SSI.

CONCLUSIONS

This institutional experience over a 4-year period revealed that the overall rate of SSI by the NHSN criteria was low at 0.49% following thoracolumbar surgery. This was attributable to the implementation of presurgery optimization, and intraoperative and postoperative measures to prevent SSI across the authors’ institution. Despite prevention measures, having a history of CAD or DM, and being male, were risk factors associated with increased SSI, and presurgery CHG shower utilization decreased SSI risk in patients.

ABBREVIATIONS ASA = American Society of Anesthesiologists; CAD = coronary artery disease; CHG = chlorhexidine gluconate; CI = confidence interval; DM = diabetes mellitus; EBL = estimated blood loss; LOS = length of stay; MIS = minimally invasive surgery; MRSA = methicillin-resistant Staphylococcus aureus; MRSE = methicillin-resistant Staphylococcus epidermidis; MSSA = methicillin-sensitive S. aureus; MSSE = methicillin-sensitive S. epidermidis; NHSN = National Healthcare Safety Network; OR = odds ratio; SSI = surgical site infection.

In Brief

Surgical site infection (SSI) remains a notable cause of postoperative morbidity and mortality with current operative techniques and antiseptic protocols, indicating a clear need for improved understanding and reduction of the risk factors. Over a 4-year period at a major institution, the authors determined the rate of deep SSI in 2252 consecutive patients. They demonstrated that male sex, coronary artery disease, and diabetes were associated with increased SSI, whereas the implementation of presurgical chlorhexidine showers reduced SSI.

Surgical site infection (SSI) is a problematic complication following spine surgery, with reported rates ranging from 0.7% to 10%.10,23,24,27,28,31 The diagnosis and treatment of SSI can be difficult, and it frequently necessitates reoperation for wound debridement and/or instrumentation removal, prolonged hospitalization, and an extended antibiotic course. For these reasons, SSI remains a notable cause of postoperative morbidity, mortality, and high acuity of medical care, with treatment costs that range from $15,800 to $43,900.5,9,21,35

In the context of an aging population and an increase in the number of spinal procedures performed,19,38 the need to improve our understanding and prevention of SSI remains important. The current standards of practice support the use of prophylactic antibiotics to decrease SSI after spinal surgery.20,37 Indeed, there is level II and III evidence on the efficacy of a single preoperative dose of antibiotics followed by intraoperative redosing as needed.29 The consensus over the source of SSI is that native skin microflora cause the majority of SSIs after spine procedures. Despite advanced antiseptic protocols, including the use of presurgery antiseptic showers and antibiotics,13–17 current practices have failed to eliminate SSI, and the reported rates vary from one institution to another.

Several risk factors have been studied in spine surgery patients that can possibly predispose patients to SSI after spine operations, including diabetes mellitus (DM), smoking, obesity, and prolonged operative time.7,23,29 For a number of reasons, including variations in neurosurgical practices and the lack of database granularity, the capacity to identify risk factors that increase SSI has been limited. To address this gap in knowledge, we systematically evaluated all patients who underwent thoracic or lumbar spine operations from 2012 to 2016 and identified candidate risk factors that were independently associated with SSI. We hypothesize that demographics, comorbidities (including but not limited to DM), and operative variables may be associated with risk of SSI, and that utilization of presurgical chlorhexidine gluconate (CHG) showers by patients may be associated with reduced risks of SSI.

Methods

Study Design and Data Collection

The study received IRB review and approval from the University of California, San Francisco. The study contained fully de-identified patient information, and therefore patient consent was not required. A total of 4266 spine surgeries were performed during a period between April 2012 and April 2016 that met the NHSN criteria on SSI follow-up after a spine operation, as described below. All spine procedures were performed at a single, high-volume, academic facility by a neurosurgical team, as previously described.6 The prevention strategies of SSI implemented in our spine surgery patients are illustrated in Fig. 1. A presurgical CHG shower protocol was implemented in November 2013 for all patients treated at our institution. Patients were instructed to shower with CHG liquid soap daily for a minimum of three times, including two times the nights prior to surgery and once the morning of admission for surgery (Fig. 2).

FIG. 1.
FIG. 1.

Flowchart depicting the perioperative and intraoperative preventive measures of SSI for patients undergoing spinal surgery at our institution. ADL = activities of daily living.

FIG. 2.
FIG. 2.

Instructional handout given to patients at the preoperative appointment to demonstrate the CHG wash protocol. Image credit: 2013–2016 University of California, Office of the President (PI: Torriani). Developing standardized operative bundles to decrease surgical site infections (SSI). All rights reserved. Copyright UC Health. Published with permission. Figure is available in color online only.

All patients who underwent thoracic and/or lumbar surgery with SSI surveillance were included in the study. Patients with cervical and cervicothoracic surgery, and surgeries extending beyond the thoracolumbar region, were not included in this study. Per our institutional protocol, all patients regardless of sex received standard pre-incision clipping of their hair as part of standard sterilization. We have included this in the Methods section. Demographics (age, sex), comorbidities (BMI, DM, coronary artery disease [CAD], baseline malignancy, smoking status, ambulation status, use of preoperative CHG showers, American Society of Anesthesiologists [ASA] classification), surgical pathology (degenerative, deformity, tumor, trauma, or other), spinal surgical level, intraoperative variables (revision, fusion, resident involvement, open versus minimally invasive surgery [MIS], operative approach, estimated blood loss [EBL], operative duration), antibiotic and steroid use, and hospital length of stay (LOS) were documented for each patient. The data reporting adhered to Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) recommendations.33

Infection Classification and Outcome Monitoring

The rate and total number of SSIs per quarter were identified in accordance with the National Healthcare Safety Network (NHSN) SSI event identification guidelines, and consistent with our prior publication.6 In this study, NHSN operative procedure categories included all spinal fusions, laminectomies, and refusions of the spine. Our institution’s Department of Hospital Epidemiology and Infection Control reviewed records for patients undergoing spine surgery, as identified by ICD-9 procedure codes (Supplemental Table 1). Additionally, case records were independently reviewed to identify additional cases of SSI. Patient records associated with the following were identified through an electronic surveillance system: 1) positive culture of tissue, body fluid, wound, or blood; 2) return to surgery for incision and debridement; or 3) ICD-9 diagnosis codes suggestive of SSI (996.6, 996.69, 998.51, and 998.59). Records meeting any of the above criteria were then reviewed in detail by a hospital infection preventionist who was not associated with the current study. SSIs were classified as either superficial or deep. Criteria for superficial SSIs were defined as infections occurring within 30 days of surgery and involving only skin and subcutaneous tissue, while criteria for deep SSIs were infections that occurred within 30 days of surgery and involved deeper layers (e.g., fascial and subfascial muscle layers). SSIs from fusion procedures were defined as occurring within 90 days of surgery.

Statistical Analysis

Categorical variables were presented as proportions, and continuous variables using means and standard deviations. Univariate analyses used Pearson’s chi-square test. The Fisher exact test was used for analyses with cell counts < 5. Variables with p values < 0.20 were incorporated into the multivariable model. The outcome variable was SSI, and multivariable analysis of SSI was performed using binary logistic regression. Odds ratios (ORs) and their associated 95% confidence intervals (CIs) were reported. Significance was assessed at p ≤ 0.05. Statistical analyses were executed using the Statistical Package for Social Sciences (version 25, IBM Corp.).

Results

Demographics, Comorbidities, and Operative Variables

In total, 4266 patients were treated with spinal procedures between April 2012 and April 2016. Of these, 2252 patients had thoracic or lumbar surgeries and were included for subsequent analyses. Summary data of demographic, comorbidity, and operative variables for all patients are listed in Table 1. The mean patient age was 58.6 ± 13.8 years, 49.6% of patients were male, and the mean BMI was 28.6 ± 6.3 kg/m2. The comorbidities included DM (7.9%), CAD (6.1%), history of malignancy (9.6%), smoking (10.8%), and impaired ambulation (19.4%). Sixty percent of patients utilized a presurgical CHG shower protocol, and 42.2% of patients were classified as ASA class > 2. Primary spinal pathologies were degenerative conditions (61.9%), tumor (16.8%), deformity (13.5%), trauma (2.0%), and other (5.8%).

TABLE 1.

Characteristics of all patient cohorts following thoracic or lumbar operations, and univariate analysis of patients with and without SSI

ParameterAll (%)No SSI (%)SSI (%)p Value
No. of patients2252224111
Mean age ± SD, yrs58.6 ± 13.858.6 ± 13.761.1 ± 15.30.553
Males1118 (49.6)1108 (49.4)10 (90.9)0.006
Mean BMI ± SD, kg/m228.6 ± 6.328.6 ± 6.329.1 ± 7.10.801
DM178 (7.9)175 (7.8)3 (27.3)0.050
CAD138 (6.1)134 (6.0)4 (36.4)0.003
History of malignancy215 (9.5)214 (9.5)1 (9.1)0.996
Smoker242 (10.8)241 (10.8)1 (9.1)0.858
Ambulation1837 (81.6)1828 (81.6)11 (100)0.983
ASA classification0.132
 ≤ II1301 (57.8)1292 (57.7)9 (81.8)
 > II951 (42.2)949 (42.3)2 (18.2)
Presurgical CHG shower1347 (59.8)1346 (60.1)1 (9.1)0.001
Surgical pathology0.637
 Degenerative1394 (61.9)1388 (61.9)6 (54.5)
 Deformity303 (13.5)300 (13.4)3 (27.3)
 Tumor379 (16.8)377 (16.8)2 (18.2)
 Trauma46 (2.0)46 (2.1)0 (0)
 Other130 (5.8)130 (5.8)0 (0)
Resident involvement1753 (77.8)1744 (77.8)9 (81.8)0.750
Open vs MIS surgery0.267
 Open1763 (78.3)1757 (78.4)7 (63.6)
 MIS489 (21.7)484 (21.6)4 (36.4)
Revision283 (12.6)282 (12.6)1 (9.1)0.727
Fusion1168 (51.9)1162 (51.9)7 (63.6)0.551
Surgical level0.413
 ≤ 5-level thoracic/thoracolumbar236 (10.5)236 (10.5)0 (0)
 ≤ L2–S11789 (79.4)1779 (79.4)9 (81.8)
 > L2–S1 & < T10–pelvis179 (7.9)177 (7.9)2 (18.2)
 ≥ T10–pelvis48 (2.1)49 (2.2)0 (0)
Surgical approach0.011
 Posterior1709 (75.9)1701 (75.9)7 (63.6)
 Anterior324 (14.4)324 (14.5)0 (0)
 Lateral78 (3.5)76 (3.4)2 (18.2)
 2-stage141 (6.3)140 (6.2)2 (18.2)
Intraop antibiotics2156 (95.7)2145 (95.7)11 (100)
Postop antibiotics1493 (66.3)1488 (66.4)5 (45.5)0.143
Pre/postop steroids984 (43.7)980 (43.7)4 (36.4)0.765
EBL, L0.196
 < 11855 (85.9)1846 (85.9)8 (72.7)
 ≥ 1305 (14.1)303 (14.1)3 (27.3)
Op duration, hrs0.263
 < 2421 (18.7)417 (18.6)3 (27.3)
 2–51352 (60.1)1349 (60.2)4 (36.4)
 > 5478 (21.2)474 (21.2)4 (36.4)
Mean LOS ± SD, days6.6 ± 7.46.6 ± 7.46.1 ± 2.70.806
Boldface type indicates statistical significance.

According to the spinal level of procedure, 79.4% of operations were ≤ L2–S1 (Table 1). The rate of resident involvement in the surgeries was 77.8%. Twenty-two percent of operations were MISs, and 12.6% of surgeries were revisions. By operative approach, 75.9% were posterior, 14.4% anterior, 3.5% lateral, and 6.3% 2-stage surgeries. The rate of intraoperative antibiotic use was 95.7% and postoperative antibiotic use was 66.3%. Perioperative steroid use during the pre- or postoperative period was 43.7%. The mean operative duration was 3.7 ± 2 hours, and the mean EBL was 467 ± 829 ml.

Risk Factors of SSIs

The results of the univariate analysis for the risk factors associated with SSI are provided in Table 1. In total, 11 patients (0.49%) experienced postoperative SSIs. The SSI rates for men and women were 0.9% and 0.1%, respectively, using the chi-square test (p = 0.006). Patients with DM (1.7% vs 0.4%, p = 0.050) and CAD (2.9% vs 0.3%, p = 0.003) also had higher rates of SSI. Patients using preoperative CHG showers had lower rates of SSI (1.1% vs 0.1%, p = 0.001). Also in the univariate analysis, patients receiving postoperative antibiotics were associated with lower rates of SSI (0.3% vs 0.8%, p = 0.143), and EBL ≥ 1 L with a higher SSI rate (1.0% vs 0.4%, p = 0.196). Table 2 shows the characteristics of the patient cases with SSI and the cultured microorganisms. All of the identified infections constituted deep SSIs involving the fascial and subfascial tissues as defined by the NHSN guidelines. Of the 11 patients with SSIs, all required reoperation for wound debridement, tissue culture, and wound washout. Four of these patients also needed hardware removal and revision operation. The pathogens were methicillin-sensitive Staphylococcus aureus (MSSA; n = 2), methicillin-resistant S. aureus (MRSA; n = 2), vancomycin-resistant Enterococcus faecium (n = 2), Propionibacterium acnes (n = 2), methicillin-sensitive Staphylococcus epidermidis (MSSE; n = 1), methicillin-resistant S. epidermidis (MRSE) (n = 1), and unknown (n = 1). Subgroup analysis of patients pre- and post-CHG protocol implementation in November 2013 demonstrated that SSI was lower postimplementation (OR 0.09, 95% CI 0.11–0.70) while controlling for other patient variables in the multivariate analysis.

TABLE 2.

Summary data of patients with SSIs following thoracic or lumbar spinal surgery

Age (yrs), SexDMCADASA ClassCHG ShowerPathologyLevelFusionApproachEBL (L)TransfusionOp Time (hrs)Culture
76, MYesYesIINoDegenerative≤ L2–S1Yes2-stage> 12 units pRBC> 5MSSE
80, MNoYesIINoDeformity≤ L2–S1NoPosterior< 1No< 2MRSE
52, FNoNoIINoTumor≤ L2–S1YesLateral< 1No> 5Vancomycin-resistant E. faecium
72, MNoYesIINoDeformity> L2–S1, < T10–pelvisYes2-stage> 12 units pRBC> 5MSSA
34, MNoNoIINoDegenerative≤ L2–S1YesLateral< 1No2–5Unknown*
51, MYesNoIIINoDegenerative≤ L2–S1NoPosterior< 1No< 2Vancomycin-resistant E. faecium
52, MNoNoINoDegenerative≤ L2–S1NoPosterior< 1No2–5P. acnes
73, MNoNoIINoDegenerative≤ L2–S1NoPosterior< 1No< 2MSSA
73, MNoNoIIINoTumor> L2–S1, < T10–pelvisYes2-stage< 1No2–5MRSA
43, MNoNoIIYesDegenerative≤ L2–S1YesPosterior> 12 units pRBC> 5MRSA
66, MYesYesIINoDeformity≤ L2–S1YesPosterior< 1No< 2P. acnes
pRBC = packed red blood cells.

Patient received antibiotics preoperatively; therefore, intraoperative cultures were negative. However, pathology demonstrated microabscesses.

After adjusting for all variables with p values < 0.20 in the univariate analysis (sex, DM, CAD, presurgical CHG shower, ASA classification, surgical approach, EBL, and postoperative antibiotic), on multivariable analysis (Supplemental Table 2) male sex had an OR of 11.75 (95% CI 1.45–96.46, p = 0.021) for SSI compared to female sex (Fig. 3A), and patients with CAD had an OR of 8.78 (95% CI 1.99–38.72, p = 0.004) for SSI compared to patients without CAD (Fig. 3B). In addition, patients with presurgical CHG showers had an OR of 0.08 (95% CI 0.01–0.67, p = 0.019) for SSI compared to patients without CHG showers (Fig. 3C). DM did not reach statistical significance for predicting SSI on multivariable regression.

FIG. 3.
FIG. 3.

Bar graphs depicting the rate of SSI in all patients who underwent thoracic or lumbar spine surgery at our institution. A: The rate of SSI between male and female patients. B: The rate of SSI between patients with and without CAD. C: The rate of SSI between patients with and without CHG showers. *p < 0.05.

Subgroup Analysis of Patients With Nonfusion and Fusion Surgery

In total, 1168 patients underwent spinal fusion surgery, of which there were 7 SSIs (0.6%) (Table 3). A total of 1084 patients had nonfusion surgery, and there were 4 SSIs (0.4%). There was no statistical difference in the rate of SSI between these two cohorts (Fig. 4A) as well as according to the region of spine surgery (Fig. 4B). However, we observed that patients undergoing fusion were significantly older (mean 60.1 ± 12.7 vs 57.1 ± 14.7 years, p < 0.001), had a higher proportion of men (52.7% vs 46.3%, p = 0.002), and were more likely to have an ASA classification > II (48.0% vs 36.0%, p < 0.001). Fusion patients were significantly more likely to undergo open spinal surgery (80.2% vs 76.2%, p = 0.021) as opposed to MIS, and had longer operative durations (mean 252.3 ± 120.9 vs 191.1 ± 110.2 minutes, p < 0.001).

TABLE 3.

Subgroup analysis of patient cohorts with nonfusion and fusion surgery of the thoracic or lumbar spine

ParameterNonfusion (%)Fusion (%)p Value
No. of patients10841168
Mean age ± SD, yrs57.1 ± 14.760.1 ± 12.7<0.001
Males502 (46.3)616 (52.7)0.002
Mean BMI ± SD, kg/m228.8 ± 6.528.5 ± 6.10.299
DM82 (7.6)96 (8.2)0.585
CAD59 (5.4)79 (6.8)0.218
History of malignancy96 (8.9)119 (10.2)0.350
Smoker111 (10.2)131 (11.2)0.451
Ambulation892 (82.3)945 (80.9)0.399
CHG shower625 (57.7)722 (61.8)0.044
ASA classification<0.001
 ≤ II694 (64.0)607 (52.0)
 > II390 (36.0)561 (48.0)
Surgical pathology0.301
 Degenerative683 (63.0)711 (60.9)
 Deformity136 (12.5)167 (14.3)
 Tumor173 (16.0)206 (17.6)
 Trauma27 (2.5)19 (1.6)
 Other65 (6.0)65 (5.6)
Resident involvement849 (78.3)904 (77.4)0.598
Open vs MIS surgery0.021
 Open826 (76.2)937 (80.2)
 MIS258 (23.8)231 (19.8)
Revision137 (12.6)146 (12.5)0.921
Surgical level0.768
 ≤ 5-level thoracic/thoraco-  lumbar108 (10.0)128 (11.0)
 ≤ L2–S1871 (80.4)918 (78.6)
 > L2–S1 & < T10–pelvis82 (7.6)97 (8.3)
 ≥ T10–pelvis23 (2.1)25 (2.1)
Surgical approach<0.001
 Posterior1084 (100)625 (53.5)
 Anterior0 (0)324 (27.8)
 Lateral0 (0)78 (6.7)
 2-stage0 (0)141 (12.1)
Intraop antibiotics1035 (95.5)1121 (96.0)0.560
Postop antibiotics710 (65.5)783 (67.0)0.440
Pre/postop steroids459 (42.3)525 (45.0)0.206
EBL, L0.071
 < 1900 (87.3)955 (84.6)
 ≥ 1131 (12.7)174 (15.4)
Op duration, hrs<0.001
 < 2328 (30.3)93 (8.0)
 2–5592 (54.6)760 (65.1)
 > 5164 (15.1)314 (26.9)
Mean LOS ± SD, days6.6 ± 7.76.6 ± 7.10.937
Boldface type indicates statistical significance.
FIG. 4.
FIG. 4.

Bar graphs depicting the rate of SSI in patients according to the type of surgery. A: The rate of SSI between fusion and nonfusion patients. B: The rate of SSI by spinal location.

Discussion

SSI after spinal surgery is a major complication that results in increased patient morbidity, mortality, and healthcare burden.5,9,18,35 Prior studies have demonstrated SSI rates ranging widely from 0.7% to 10.0%, depending on the patient population, pathology, and type of procedure.1,10,23,24,27,28,31,36 Unfortunately, large administrative registries lack the granularity to account for a number of perioperative and intraoperative variables that can affect the infection rate among institutions. In addition to patient characteristics and operative variables playing a role, the location of spine surgery also affects the rate of SSI, with thoracic (2.1%) and lumbar (1.6%) procedures more prone to SSI than cervical surgery (0.8%).30 The present study captures these variables and further clarifies their effect on SSI following spine surgery.

To date, only a handful of studies have analyzed an entire population of spine surgery patients at a major academic institution. This methodology adjusts for confounding variables and helps to provide insights regarding the efficacy of current standards of practice. In our retrospective cohort study of 2252 patients treated with thoracolumbar spine surgery at 1 facility, we were able to determine the absolute SSI rate as well as the rate of SSI by patient subgroup. Furthermore, we identified independent risk factors that increased SSI. Utilizing the NHSN operative procedure guidelines to identify SSI events, our overall SSI rate of 0.49% over a 4-year time span was lower than the rates generally reported in the literature. The surveillance period for SSI was 30 days after surgery, and 90 days for fusion operations. The low rate of SSI in the present study can possibly be attributed to concerted strategies in the past decade to reduce SSI29,34 and the implementation of standardized multistep prevention measures across our institution (Fig. 1). These measures included extensive patient education and preventive strategies from before hospitalization to follow-up after surgery. In addition, there are ongoing efforts through antiseptic CHG showers to reduce the skin microflora of preoperative spine patients.6 While evidence regarding CHG shower is lacking in spine surgery and in other surgical specialties,15,26,39 in our spine patients presurgical CHG showering decreased SSI from 1.1% to 0.1%. Given that S. aureus and S. epidermidis bacteria found in normal skin flora are the frequent source of SSIs, an antiseptic shower regimen may indeed decrease the infection risk of the patient at baseline relative to using unmedicated soap.

For the reasons mentioned above, the low number of SSIs over the 4-year period illustrates a strength in the evolving field of spine surgery, as well as a limitation in identifying multivariate predictors. While our study remains exploratory and in need of confirmation, nevertheless our data are important to report given the large sample size. The infection rate varied between men (0.9%) and women (0.1%). Reasons for this difference could be multifactorial, including variations in the propensity for skin microbial colonization and the effect of differential hormonal influence on the acute phase of wound healing.29,34 Few studies have reported the impact of patient sex on SSI following spine surgery. Evidence in other surgical specialties provides precedent for sex differences in SSI rates that could be procedure-specific. For example, women have lower infection rates following abdominal surgery, orthopedic procedures, and trauma than men do.4,8,11,17 In patients with central catheters, men were more likely to have bacterial colonization of the surrounding skin, and hair growth can also interfere with wound dressing and the healing process.8 Lastly, sex hormones are known to modulate inflammation and result in marked differences in the wound healing process. In particular, androgens have been associated with impaired cutaneous re-epithelialization.12–14,25

Several prior studies have explored the risk factors associated with SSI after spine surgery,2,10,24,27,31,32 which vary by institution as well as by study design. In the modern era of spine surgery, with improved perioperative medical management and preventive measures, our institutional experience suggests that smoking status and operative time are not risk factors for increased infection. However, in addition to DM affecting postoperative healing, we found that CAD is associated with increased odds of SSI following thoracolumbar spine surgery. Patients with CAD (2.9%) had a higher infection rate than those without the comorbidity (0.3%). CAD can be an infection risk for several reasons. It has been associated with prolonged recovery, which predisposes the patient to greater risks of postoperative complications.3 Patients with cardiac disease may be prone to experience anemia with surgical blood loss,16,22 and understanding the link between postsurgical transfusion and SSI is a topic of interest in future investigations. Despite prophylactic antibiotic administration, these patients may be at risk for decreased delivery of antibiotics in addition to hypoperfusion and poor blood supply of the surgical wound secondary to general vasculopathy.14

We found that patients with diabetes experienced a higher rate of SSI (1.7%) than those without diabetes (0.4%, comparable to previous publications2,10,24,27). For our patients, the target for presurgery hemoglobin A1C score was less than 7.5%, with the exception of operations that could not be delayed. Fusion was not associated with increased rate of SSIs and prolonged LOS in this study, which can vary with discharge practices and patient pathology. Patients with tumor or deformity pathologies make up a small sample relative to degenerative diseases, with wide variability in management in these patients. Pathology-specific risk factors are in need of validation through larger patient populations. Of the 11 infection cases, 4 patients had a history of steroid use and 1 with immunosuppressant medication use, which may be a risk factor for increased infection. Patients with postoperative antibiotics for drain prophylaxis had a lower rate of SSI (0.3%) than patients without (0.8%), but this difference was not statistically significant. Larger studies—without compromising data granularity—can help clarify the effect of postoperative drain prophylaxis. This suggests that male sex and CAD may be more important predictors of SSI in the thoracic or lumbar spinal procedure population. Of note, at our center, we ask patients undergoing spine surgery to quit smoking perioperatively.

Limitations and Strengths

The limitations of our findings arise from the retrospective nature of our study. The rate of SSI was quite low in our study. While conferring the benefits of modern surgical technique and infection control protocols, such as those at our institution (Figs. 1 and 2), this low number of SSIs also limits statistical power and the identification of validated multivariable predictors, which await the results of larger studies on the topic. Our reported rate of SSI following spine surgery was also lower than prior reports. Reasons for this include differences in patient population, spinal pathology, and institutional practices.38 SSIs occurring outside of the surveillance window per the NHSN guidelines were not included in our study. Gradual changes to surgical practice patterns and patient populations over time can limit the generalizability of the results reported herein. Of interest for future investigations are other variables that could influence the wound healing process, such as chronic immunosuppressant drug use and preoperative nutritional status of the patient. Patients who received surgeries starting in November 2013 were provided detailed instructions and CHG liquid soap in preparation for surgery. While we did not quantify the rate of patient adherence, CHG shower decreased SSI following implementation of the protocol, as described in our prior publication.6 The strengths of the current study were the degree of detail and completeness of clinical information obtained for every patient to help assess the magnitude of risk that was associated with each factor.

Conclusions

In a cohort of 2252 consecutive patients undergoing surgery for thoracolumbar spine disorders at our institution, the rate of SSI according to the NHSN criteria was low at 0.49% and attributable to the use of standardized patient optimization and education from before surgery to follow-up after discharge from the hospital, as well as due to implementation of an antisepsis preoperative CHG showering protocol across our institution. We found that being male and having a history of CAD were risk factors for increased SSI in addition to DM, whereas preoperative CHG shower utilization decreased SSI. Furthermore, obesity and smoking history were not associated with infections in our patient population, likely from preoperative optimization. Prospective studies are needed to further investigate causal factors of SSI, and to evaluate strategies for infection reduction in this cohort of at-risk patients.

Disclosures

Dr. A. K. Chan reports support of non–study-related clinical or research effort from Orthofix. Dr. Dhall reports being a consultant to DePuy Synthes. Dr. Clark reports being a consultant to NuVasive. Dr. Chou reports being a consultant to Medtronic and Globus, and receiving a small royalty from Globus. Dr. Ames reports being a consultant to DePuy Synthes, Medtronic, Stryker, Medicrea, K2M, and Biomet Zimmer; receiving royalties from Stryker, Biomet Zimmer, DePuy Synthes, NuVasive, Next Orthosurgical, K2M, and Medicrea; conducting research for Titan Spine, DePuy Synthes, and ISSG; being on the editorial board of Operative Neurosurgery; receiving grant funding from SRS; being on the Executive Committee of ISSG; and being the director of Global Spine Analytics. Dr. Mummaneni reports being a consultant to DePuy Spine, Globus, and Stryker; direct stock ownership in Spinicity/ISD; receiving statistical analysis for study/writing or editorial assistance on the manuscript from ISSG; receiving support of non–study-related clinical or research effort from the NREF; receiving royalties from DePuy Spine, Thieme Publishing, and Springer Publishing; and receiving honoraria from Spineart.

Author Contributions

Conception and design: Deng, AK Chan, Oh, Liu, Yue, Dhall, Clark, Chou, Ames, Mummaneni. Acquisition of data: AK Chan, Ammanuel, AY Chan, Skrehot, Edwards, Kondapavulur, Nichols, Liu, Dhall, Clark, Chou, Ames, Mummaneni. Analysis and interpretation of data: Deng, AK Chan, Ammanuel, AY Chan, Skrehot, Kondapavulur, Nichols, Yue. Drafting the article: Deng, AK Chan, Edwards, Nichols, Yue, Chou, Mummaneni. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Deng. Statistical analysis: Deng, Liu. Administrative/technical/material support: Ammanuel, AY Chan, Skrehot, Edwards, Kondapavulur, Nichols, Liu, Mummaneni. Study supervision: AK Chan, Oh, Yue, Dhall, Clark, Chou, Ames, Mummaneni.

Supplemental Information

Online-Only Content

Supplemental material is available with the online version of the article.

References

  • 1

    Abbey DMTurner DMWarson JSWirt TCScalley RD: Treatment of postoperative wound infections following spinal fusion with instrumentation. J Spinal Disord 8:2782831995

    • Search Google Scholar
    • Export Citation
  • 2

    Abdul-Jabbar ATakemoto SWeber MHHu SSMummaneni PVDeviren V: Surgical site infection in spinal surgery: description of surgical and patient-based risk factors for postoperative infection using administrative claims data. Spine (Phila Pa 1976) 37:134013452012

    • Search Google Scholar
    • Export Citation
  • 3

    Bekelis KDesai ABakhoum SFMissios S: A predictive model of complications after spine surgery: the National Surgical Quality Improvement Program (NSQIP) 2005–2010. Spine J 14:124712552014

    • Search Google Scholar
    • Export Citation
  • 4

    Brandt CSohr DBehnke MDaschner FRüden HGastmeier P: Reduction of surgical site infection rates associated with active surveillance. Infect Control Hosp Epidemiol 27:134713512006

    • Search Google Scholar
    • Export Citation
  • 5

    Calderone RRGarland DECapen DAOster H: Cost of medical care for postoperative spinal infections. Orthop Clin North Am 27:1711821996

    • Search Google Scholar
    • Export Citation
  • 6

    Chan AKAmmanuel SGChan AYOh TSkrehot HCEdwards CS: Chlorhexidine showers are associated with a reduction in surgical site infection following spine surgery: an analysis of 4266 consecutive surgeries. Neurosurgery [epub ahead of print]2018

    • Search Google Scholar
    • Export Citation
  • 7

    Chen SAnderson MVCheng WKWongworawat MD: Diabetes associated with increased surgical site infections in spinal arthrodesis. Clin Orthop Relat Res 467:167016732009

    • Search Google Scholar
    • Export Citation
  • 8

    Cohen BChoi YJHyman SFuruya EYNeidell MLarson E: Gender differences in risk of bloodstream and surgical site infections. J Gen Intern Med 28:131813252013

    • Search Google Scholar
    • Export Citation
  • 9

    Deng HYue JKOrdaz ARivera EJSuen CGSing DC: Cervical fusion for degenerative disease: a comprehensive cost analysis of hospital complications in the United States from 2002 to 2014. J Craniovertebr Junction Spine 9:1401472018

    • Search Google Scholar
    • Export Citation
  • 10

    Fang AHu SSEndres NBradford DS: Risk factors for infection after spinal surgery. Spine (Phila Pa 1976) 30:146014652005

  • 11

    Gibbons CBruce JCarpenter JWilson APWilson JPearson A: Identification of risk factors by systematic review and development of risk-adjusted models for surgical site infection. Health Technol Assess 15:1156iii–iv 2011

    • Search Google Scholar
    • Export Citation
  • 12

    Gilliver SCAshworth JJAshcroft GS: The hormonal regulation of cutaneous wound healing. Clin Dermatol 25:56622007

  • 13

    Grossman CJ: Interactions between the gonadal steroids and the immune system. Science 227:2572611985

  • 14

    Guo SDipietro LA: Factors affecting wound healing. J Dent Res 89:2192292010

  • 15

    Hayek L: A placebo-controlled trial of the effect of two preoperative baths or showers with chlorhexidine detergent on postoperative wound infection rates. J Hosp Infect 13:2022041989

    • Search Google Scholar
    • Export Citation
  • 16

    Hollis RHSingletary BAMcMurtrie JTGraham LARichman JSHolcomb CN: Blood transfusion and 30-day mortality in patients with coronary artery disease and anemia following noncardiac surgery. JAMA Surg 151:1391452016

    • Search Google Scholar
    • Export Citation
  • 17

    Langelotz CMueller-Rau CTerziyski SRau BKrannich AGastmeier P: Gender-specific differences in surgical site infections: an analysis of 438,050 surgical procedures from the German National Nosocomial Infections Surveillance System. Viszeralmedizin 30:1141172014

    • Search Google Scholar
    • Export Citation
  • 18

    Levi ADDickman CASonntag VK: Management of postoperative infections after spinal instrumentation. J Neurosurg 86:9759801997

  • 19

    Liu CYZygourakis CCYoon SKliot TMoriates CRatliff J: Trends in utilization and cost of cervical spine surgery using the National Inpatient Sample Database, 2001 to 2013. Spine (Phila Pa 1976) 42:E906E9132017

    • Search Google Scholar
    • Export Citation
  • 20

    Malis LI: Prevention of neurosurgical infection by intraoperative antibiotics. Neurosurgery 5:3393431979

  • 21

    McGirt MJParker SLLerner JEngelhart LKnight TWang MY: Comparative analysis of perioperative surgical site infection after minimally invasive versus open posterior/transforaminal lumbar interbody fusion: analysis of hospital billing and discharge data from 5170 patients. J Neurosurg Spine 14:7717782011

    • Search Google Scholar
    • Export Citation
  • 22

    Nappi J: Anemia in patients with coronary artery disease. Am J Health Syst Pharm 60 (14 Suppl 3):S4S82003

  • 23

    Olsen MAMayfield JLauryssen CPolish LBJones MVest J: Risk factors for surgical site infection in spinal surgery. J Neurosurg 98 (2 Suppl):1491552003

    • Search Google Scholar
    • Export Citation
  • 24

    Olsen MANepple JJRiew KDLenke LGBridwell KHMayfield J: Risk factors for surgical site infection following orthopaedic spinal operations. J Bone Joint Surg Am 90:62692008

    • Search Google Scholar
    • Export Citation
  • 25

    Pergola CRogge ADodt GNorthoff HWeinigel CBarz D: Testosterone suppresses phospholipase D, causing sex differences in leukotriene biosynthesis in human monocytes. FASEB J 25:337733872011

    • Search Google Scholar
    • Export Citation
  • 26

    Prabhu ASKrpata DMPhillips SHuang LCHaskins INRosenblatt S: Preoperative chlorhexidine gluconate use can increase risk for surgical site infections after ventral hernia repair. J Am Coll Surg 224:3343402017

    • Search Google Scholar
    • Export Citation
  • 27

    Pull ter Gunne AFCohen DB: Incidence, prevalence, and analysis of risk factors for surgical site infection following adult spinal surgery. Spine (Phila Pa 1976) 34:142214282009

    • Search Google Scholar
    • Export Citation
  • 28

    Rechtine GRBono PLCahill DBolesta MJChrin AM: Postoperative wound infection after instrumentation of thoracic and lumbar fractures. J Orthop Trauma 15:5665692001

    • Search Google Scholar
    • Export Citation
  • 29

    Shaffer WOBaisden JLFernand RMatz PG: An evidence-based clinical guideline for antibiotic prophylaxis in spine surgery. Spine J 13:138713922013

    • Search Google Scholar
    • Export Citation
  • 30

    Smith JSShaffrey CISansur CABerven SHFu K-MGBroadstone PA: Rates of infection after spine surgery based on 108,419 procedures: a report from the Scoliosis Research Society Morbidity and Mortality Committee. Spine (Phila Pa 1976) 36:5565632011

    • Search Google Scholar
    • Export Citation
  • 31

    Tominaga HSetoguchi TKawamura HKawamura INagano SAbematsu M: Risk factors for unavoidable removal of instrumentation after surgical site infection of spine surgery: a retrospective case-control study. Medicine (Baltimore) 95:e51182016

    • Search Google Scholar
    • Export Citation
  • 32

    Tsubouchi NFujibayashi SOtsuki BIzeki MKimura HOta M: Risk factors for implant removal after spinal surgical site infection. Eur Spine J 27:248124902018

    • Search Google Scholar
    • Export Citation
  • 33

    von Elm EAltman DGEgger MPocock SJGøtzsche PCVandenbroucke JP: The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg 12:149514992014

    • Search Google Scholar
    • Export Citation
  • 34

    Watters WC IIIBaisden JBono CMHeggeness MHResnick DKShaffer WO: Antibiotic prophylaxis in spine surgery: an evidence-based clinical guideline for the use of prophylactic antibiotics in spine surgery. Spine J 9:1421462009

    • Search Google Scholar
    • Export Citation
  • 35

    Whitmore RGStephen JStein SCCampbell PGYadla SHarrop JS: Patient comorbidities and complications after spinal surgery: a societal-based cost analysis. Spine (Phila Pa 1976) 37:106510712012

    • Search Google Scholar
    • Export Citation
  • 36

    Wimmer CGluch HFranzreb MOgon M: Predisposing factors for infection in spine surgery: a survey of 850 spinal procedures. J Spinal Disord 11:1241281998

    • Search Google Scholar
    • Export Citation
  • 37

    Young RFLawner PM: Perioperative antibiotic prophylaxis for prevention of postoperative neurosurgical infections. A randomized clinical trial. J Neurosurg 66:7017051987

    • Search Google Scholar
    • Export Citation
  • 38

    Zygourakis CCLiu CYKeefe MMoriates CRatliff JDudley RA: Analysis of national rates, cost, and sources of cost variation in adult spinal deformity. Neurosurgery 82:3783872018

    • Search Google Scholar
    • Export Citation
  • 39

    Zywiel MGDaley JADelanois RENaziri QJohnson AJMont MA: Advance pre-operative chlorhexidine reduces the incidence of surgical site infections in knee arthroplasty. Int Orthop 35:100110062011

    • Search Google Scholar
    • Export Citation

If the inline PDF is not rendering correctly, you can download the PDF file here.

Article Information

Contributor Notes

Correspondence Hansen Deng: University of Pittsburgh Medical Center, Pittsburgh, PA. hdneurosurg@gmail.com.INCLUDE WHEN CITING Published online November 1, 2019; DOI: 10.3171/2019.8.SPINE19479.

H.D. and A.K.C. contributed equally to this work.

Disclosures Dr. A. K. Chan reports support of non–study-related clinical or research effort from Orthofix. Dr. Dhall reports being a consultant to DePuy Synthes. Dr. Clark reports being a consultant to NuVasive. Dr. Chou reports being a consultant to Medtronic and Globus, and receiving a small royalty from Globus. Dr. Ames reports being a consultant to DePuy Synthes, Medtronic, Stryker, Medicrea, K2M, and Biomet Zimmer; receiving royalties from Stryker, Biomet Zimmer, DePuy Synthes, NuVasive, Next Orthosurgical, K2M, and Medicrea; conducting research for Titan Spine, DePuy Synthes, and ISSG; being on the editorial board of Operative Neurosurgery; receiving grant funding from SRS; being on the Executive Committee of ISSG; and being the director of Global Spine Analytics. Dr. Mummaneni reports being a consultant to DePuy Spine, Globus, and Stryker; direct stock ownership in Spinicity/ISD; receiving statistical analysis for study/writing or editorial assistance on the manuscript from ISSG; receiving support of non–study-related clinical or research effort from the NREF; receiving royalties from DePuy Spine, Thieme Publishing, and Springer Publishing; and receiving honoraria from Spineart.
Headings
Figures
  • View in gallery

    Flowchart depicting the perioperative and intraoperative preventive measures of SSI for patients undergoing spinal surgery at our institution. ADL = activities of daily living.

  • View in gallery

    Instructional handout given to patients at the preoperative appointment to demonstrate the CHG wash protocol. Image credit: 2013–2016 University of California, Office of the President (PI: Torriani). Developing standardized operative bundles to decrease surgical site infections (SSI). All rights reserved. Copyright UC Health. Published with permission. Figure is available in color online only.

  • View in gallery

    Bar graphs depicting the rate of SSI in all patients who underwent thoracic or lumbar spine surgery at our institution. A: The rate of SSI between male and female patients. B: The rate of SSI between patients with and without CAD. C: The rate of SSI between patients with and without CHG showers. *p < 0.05.

  • View in gallery

    Bar graphs depicting the rate of SSI in patients according to the type of surgery. A: The rate of SSI between fusion and nonfusion patients. B: The rate of SSI by spinal location.

TrendMD
Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 344 344 344
PDF Downloads 412 412 412
EPUB Downloads 0 0 0
PubMed
Google Scholar