Risk factors and outcomes associated with surgical site infections after craniotomy or craniectomy

Clinical article

Full access

Object

Many studies that have evaluated surgical site infections (SSIs) after craniotomy or craniectomy (CRANI) did not use robust methods to assess risk factors for SSIs or outcomes associated with SSIs. The authors conducted the current study to identify risk factors for SSIs after CRANI procedures and to evaluate outcomes attributed to SSIs.

Methods

The authors performed a nested case-control study of patients undergoing CRANI procedures between 2006 and 2010 at the University of Iowa Hospitals and Clinics. They identified 104 patients with SSIs and selected 312 controls. They collected data from medical records and used multivariate analyses to identify risk factors and outcomes associated with SSIs.

Results

Thirty-two percent of SSIs were caused by Staphylococcus aureus, 88% were deep incisional or organ space infections, and 70% were identified after discharge. Preoperative length of stay (LOS) ≥ 1 day was the only significant patient-related factor in the preoperative model (OR 2.1 [95% CI 1.2–3.4]) and in the overall model (OR 1.9 [95% CI 1.1–3.3]). Procedure-related risk factors that were significant in the overall model included Gliadel wafer use (OR 6.7 [95% CI 2.5–18.2]) and postoperative CSF leak (OR 3.5 [95% CI 1.4–8.5]). The preoperative SSI risk index, including body mass index, previous brain operation, chemotherapy on admission, preoperative LOS, procedure reason, and preoperative glucose level, had better predictive efficacy (c-statistic = 0.664) than the National Healthcare Safety Network risk index (c-statistic = 0.547; p = 0.004). Surgical site infections were associated with increased LOS during the initial hospitalizations (average increase of 50%) or readmissions (average increase of 100%) and with an increased risk of readmissions (OR 7.7 [95% CI 4.0–14.9]), reoperations (OR 36 [95% CI 14.9–87]), and death (OR 3.4 [95% CI 1.5–7.4]).

Conclusions

Surgeons were able to prospectively assess a patient's risk of SSI based on preoperative risk factors and they could modify some processes of care to lower the risk of SSI. Surgical site infections substantially worsened patients' outcomes. Preventing SSIs after CRANI could improve patient outcomes and decrease health care utilization.

Abbreviations used in this paper:ASA = American Society of Anesthesiologists; CRANI = craniotomy or craniectomy; ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification; LOS = length of stay; NHSN = National Healthcare Safety Network; SSI = surgical site infection; UIHC = University of Iowa Hospitals and Clinics.

Abstract

Object

Many studies that have evaluated surgical site infections (SSIs) after craniotomy or craniectomy (CRANI) did not use robust methods to assess risk factors for SSIs or outcomes associated with SSIs. The authors conducted the current study to identify risk factors for SSIs after CRANI procedures and to evaluate outcomes attributed to SSIs.

Methods

The authors performed a nested case-control study of patients undergoing CRANI procedures between 2006 and 2010 at the University of Iowa Hospitals and Clinics. They identified 104 patients with SSIs and selected 312 controls. They collected data from medical records and used multivariate analyses to identify risk factors and outcomes associated with SSIs.

Results

Thirty-two percent of SSIs were caused by Staphylococcus aureus, 88% were deep incisional or organ space infections, and 70% were identified after discharge. Preoperative length of stay (LOS) ≥ 1 day was the only significant patient-related factor in the preoperative model (OR 2.1 [95% CI 1.2–3.4]) and in the overall model (OR 1.9 [95% CI 1.1–3.3]). Procedure-related risk factors that were significant in the overall model included Gliadel wafer use (OR 6.7 [95% CI 2.5–18.2]) and postoperative CSF leak (OR 3.5 [95% CI 1.4–8.5]). The preoperative SSI risk index, including body mass index, previous brain operation, chemotherapy on admission, preoperative LOS, procedure reason, and preoperative glucose level, had better predictive efficacy (c-statistic = 0.664) than the National Healthcare Safety Network risk index (c-statistic = 0.547; p = 0.004). Surgical site infections were associated with increased LOS during the initial hospitalizations (average increase of 50%) or readmissions (average increase of 100%) and with an increased risk of readmissions (OR 7.7 [95% CI 4.0–14.9]), reoperations (OR 36 [95% CI 14.9–87]), and death (OR 3.4 [95% CI 1.5–7.4]).

Conclusions

Surgeons were able to prospectively assess a patient's risk of SSI based on preoperative risk factors and they could modify some processes of care to lower the risk of SSI. Surgical site infections substantially worsened patients' outcomes. Preventing SSIs after CRANI could improve patient outcomes and decrease health care utilization.

More than 100,000 craniotomy and craniectomy procedures are performed in the US each year. Surgical site infections (SSIs) complicate from 2.2% among low-risk patients to 4.7% among high-risk patients undergoing these procedures (http://hcupnet.ahrq.gov).8,15 Most SSIs after a craniotomy or craniectomy (CRANI) affect organ spaces (that is, subgaleal space, subdural space, cranial bone, or brain). Consequently, these infections often require surgical treatment and increase morbidity, mortality, and cost.

Patient-related risk factors for SSIs after CRANI identified by multivariate analyses include male sex,18,19,35 age 70 years or older,35 nontrauma reason for the procedure,19,32 and high American Society of Anesthesiologists (ASA) score.20 Procedure-related risk factors include emergency procedure,17 no antimicrobial prophylaxis,19 longer procedure duration,12,14,18,19 postoperative CSF leakage,17–19,35 and early reoperation.17,19,30,32 However, inclusion criteria for these studies varied and most studies evaluated specific subgroups of patients (for example, elective procedures) or specific infections (such as meningitis). Additionally, some studies did not assess potentially important risk factors such as comorbidities, skin preparation, and antimicrobial prophylaxis. Three studies assessed postoperative outcomes of SSIs after CRANI, but the investigators did not adjust for patients' preexisting conditions.16,26,30

At the University of Iowa Hospitals and Clinics (UIHC), the SSI rates after CRANI procedures ranged from 2.6% (low risk) to 5.2% (high risk). We previously found that procedures begun before the skin antiseptic dried and procedures during which Gliadel wafers (Arbor Pharmaceuticals, Inc.) were implanted were associated with an increased risk of SSIs.5 However, we did not evaluate some potential risk factors for SSIs and we did not evaluate outcomes. Thus, we performed this nested case-control study to 1) identify risk factors for SSIs after CRANI, 2) develop a preoperative SSI risk index, and 3) evaluate patient outcomes attributed to SSIs after CRANI.

Methods

Study Population

We studied patients undergoing CRANI in the UIHC's Department of Neurosurgery from January 1, 2006, through December 31, 2010. We excluded patients who had infections at the time of their CRANI or who died within 2 days after CRANI. Patients were included once if they had undergone 1 or more procedures. We included all patients with infections meeting the National Healthcare Safety Network (NHSN) definition of SSI (cases).7 We selected 312 controls from the patients without SSIs who had ≥ 1 follow-up visit or phone call. We frequency matched controls to cases by procedure period (6 months/period) so that cases and controls were distributed equally over 10 procedure periods. For outcome analyses, we included patients undergoing CRANI from January 1, 2006, through June 30, 2010; a general internist (A.S.K.) reviewed their preoperative notes and assigned McCabe and Jackson severity of illness scores.6,23

Data Collection

The CRANI procedures were identified by the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) procedure codes (Appendix Table 1). We retrospectively collected data from patients' medical records regarding SSIs, potential risk factors for SSIs (Appendix Table 2), duration of hospitalization, readmissions, and reoperations at the UIHC, which were related to complications of the initial CRANI procedures and occurred during the 1st postoperative year. We entered the data directly into a Microsoft Access database and validated the data entry.

Data Analysis

We analyzed de-identified data using SAS software (version 9.3, SAS Institute, Inc.). We performed bivariate analyses to test the association between each potential risk factor and SSIs and the association between SSIs and each outcome. We used univariate logistic regression modeling to assess the association between SSIs and quantitative risk factors and the chi-square test or Fisher's exact test for categorical variables.

To identify factors significantly associated with SSIs, we put all factors with p < 0.15 in the bivariate analyses or with a priori clinical significance into a multivariate logistic regression model. We used backward elimination to identify factors for a final multivariate model; factors with p < 0.15 were included in the final model. We then tested clinically relevant 2-way interactions for potential inclusion. We assessed collinearity of the quantitative variables by examining the Pearson correlation coefficient between pairs of variables, and we evaluated multicollinearity by computing the variance inflation factor for each variable. To assess the predictive ability of the final multivariate model to discriminate between cases and controls, we computed the c-statistic using both the conventional method and a cross-validation method.33 We used the Hosmer-Lemeshow goodness-of-fit test to assess the adequacy of the model fit.

We computed Kaplan-Meier estimates of the survival functions for each outcome, and we used the log-rank test to compare time-to-event curves for cases and controls. The time to the postoperative outcome (that is, event) was calculated from the date the CRANI was done to the date the postoperative outcome occurred. For readmission and reoperation, we evaluated patients who survived their 1st postoperative year and considered patients who did not have one of these outcomes within 1 year after the CRANI as censored observations. For postoperative death, we considered patients who survived 1 year after the initial CRANI as censored observations. We evaluated outcomes attributed to SSIs with multivariate linear regression (for length of stay [LOS]) or logistic regression (for readmission, reoperation, and death) to control for patients' preexisting conditions and for procedure period. Although controls were frequency matched to cases based on procedure period, we checked for possible residual confounding by considering it in the bivariate analyses. Procedure period was included as a covariate in the multivariate outcome model when the corresponding bivariate association was such that p < 0.15. Length of stay was log-transformed to meet the normality assumption.

Ethics Approval

The institutional review board at the University of Iowa Carver College of Medicine approved the study.

Results

Risk Factor Study

During the study period, neurosurgeons performed 2919 CRANI procedures in 2541 patients, of whom 104 (4.1%) acquired SSIs. We selected 312 patients without SSIs as controls. Common signs and symptoms of SSIs included purulent drainage (52.9%); temperature > 38°C (41.4%); localized swelling, redness, or heat (29.8%); and incisional pain or tenderness (21.1%). Fifty-four patients with SSIs (cases) (51.9%) had organ/space infections (30 with meningitis), 38 (36.5%) had deep incisional infections, and 12 (11.5%) had superficial incisional infections. Of the SSIs, 70% were identified after discharge, 60% during readmissions, and 10% during postoperative clinic or emergency room visits. The time to the onset of SSIs ranged from 2 to 221 days after CRANI (median 20 days). Cultures obtained in 83.7% of cases were positive for organisms. Staphylococcus aureus alone (23.1%) or in combination with other organisms (8.7%) was the most common etiological agent (Appendix Table 3). Gram-negative organisms alone (15.4%) or in combination with other organisms (11.5%) caused 26.9% of the SSIs.

Patients and controls were similar with respect to numerous demographic, preoperative, and operative variables (Table 1). Compared with controls, patients with SSIs were more likely to have worse comorbidity scores, previous brain operations, chemotherapy on admission, preoperative glucose levels ≥ 100 mg/dl, nontrauma reason for the CRANI procedure, longer preoperative hospitalizations, an NHSN risk score ≥ 2, bone flaps fixed with plates and screws (that is, rigid plated), Gliadel wafer implants, and postoperative CSF leaks.

TABLE 1:

Bivariate associations between potential risk factors and SSIs after CRANI*

Patient-Related FactorsCasesControlsOR (95% CI)p Value
no. of patients104312
median age in yrs52.052.51.0 (1.0–1.0)0.59
male sex54 (51.9)168 (53.9)0.9 (0.6–1.4)0.73
current or past smoking44 (42.3)128 (41.0)1.1 (0.7–1.7)0.82
BMI
 median in kg/m228.226.91.0 (0.99–1.0)0.18
 >30 kg/m243 (41.4)100 (33.0)1.4 (0.9–2.3)0.12
comorbidities
 peripheral vascular diseases3 (2.9)1 (0.3)9.2 (1.0–89.8)0.05
 diabetes12 (11.5)39 (12.5)0.9 (0.5–1.8)0.80
 Charlson comorbidity index ≥255 (52.9)129 (41.4)1.6 (1.0–2.5)0.04
MJ severity of illness scoren = 93n = 279
 rapidly fatal14 (15.1)69 (24.7)0.6 (0.3–1.2)0.15
 ultimately fatal30 (32.3)62 (22.2)1.5 (0.9–2.5)0.17
 nonfatal49 (52.7)148 (53.1)reference
preop variables
 any previous brain op33 (31.7)56 (18.0)2.1 (1.3–3.5)0.003
 any previous brain radiation7 (6.7)8 (2.6)2.7 (1.0–7.8)0.05
 steroid use20 (19.2)82 (26.3)0.7 (0.4–1.2)0.15
 chemotherapy on admission5 (4.8)1 (0.3)15.7 (1.8–136.1)0.004
 median glucose in mg/dl§120.5111.00.04
 glucose ≥100 mg/dl§75 (75.0)181 (63.1)1.8 (1.1–2.9)0.03
procedure reason
 trauma12 (11.5)68 (21.8)reference
 tumor57 (54.8)134 (43.0)2.4 (1.2–4.8)0.01
 bleed13 (12.5)40 (12.8)1.8 (0.8–4.4)0.17
 other reasons22 (21.2)70 (22.4)1.8 (0.8–3.9)0.15
nontrauma procedure reason92 (88.5)244 (78.2)2.1 (1.1–4.1)0.02
preop LOS
 mean in days2.2 ± 4.51.0 ± 2.61.1 (1.0–1.2)0.002
 ≥1 day46 (44.2)80 (25.6)2.3 (1.4–3.7)0.0004
procedure done w/in 24 hrs after admission38 (36.5)126 (40.4)0.8 (0.5–1.3)0.49
ASA score >264 (61.5)179 (57.4)1.2 (0.8–1.9)0.46
procedure-related factors
 using PVP as skin preparation92 (96.8)289 (96.3)1.2 (0.3–4.3)0.82
 prophylactic antimicrobial agents**
 cefazolin10 (9.6)17 (5.6)1.8 (0.8–4.1)0.15
 vancomycin38 (36.5)89 (29.2)1.4 (0.9–2.2)0.16
 nafcillin53 (51.0)172 (56.4)0.8 (0.5–1.3)0.34
op duration
 mean in mins211.4 ± 119.8191.7 ± 124.51.0 (1.0–1.0)0.16
 >225 mins42 (40.4)102 (32.7)1.4 (0.9–2.2)0.15
NHSN risk index score ≥225 (24.0)48 (15.4)1.7 (1.0–3.0)0.04
rigid-plated bone flap45 (43.3)98 (31.4)1.7 (1.0–2.6)0.03
Gliadel wafer use15 (14.4)7 (2.2)7.3 (2.9–18.6)<0.0001
intraop transfusion13 (12.5)39 (12.5)1.0 (0.5–2.0)1.00
CSF drain
 ventricular/lumbar drain27 (26.0)58 (18.6)1.2 (0.7–2.2)0.45
 other drain22 (21.2)102 (32.7)0.6 (0.4–1.1)0.09
 no drain55 (52.9)152 (48.7)reference
postop CSF leakage12 (11.5)11 (3.5)3.6 (1.5–8.4)0.002
mean postop glucose level, mg/dl††156.9 ± 49.9148.6 ± 44.40.12

Statistics are number of patients (%) or mean value ± SD, unless otherwise indicated. BMI = body mass index; MJ = McCabe and Jackson; PVP = povidone-iodine.

Data on BMI were missing for 9 controls.

Odds ratios and p values were calculated by univariate logistic regression.

We recorded the last glucose level obtained during the 30-day period before the procedure. Data were missing for 4 cases and 25 controls.

Data on skin preparation were missing for 9 cases and 12 controls.

Data on antimicrobial prophylaxis were missing for 7 controls. The antimicrobial agents were administered either alone or in combination with other agents.

We recorded the first glucose level obtained within the first 30 days after the procedure. Data were missing for 6 cases and 27 controls.

In the multivariate model, preoperative LOS ≥ 1 day was the only preoperative risk factor for SSIs that was significant (Table 2). Any previous brain operation, chemotherapy on admission, nontrauma procedure reason, and a preoperative glucose level ≥ 100 mg/dl were marginally associated with increased risk of SSIs (p ≤ 0.10). We used 100 mg/dl as the cut point for glucose level because it had the highest odds ratio and the lowest p value among the 3 cut points (that is, 100, 120, and 150 mg/dl) considered in the bivariate analysis. The preoperative model had fair predictive efficacy (c = 0.663; cross-validated c = 0.607). The overall model, which included both patient-related and procedure-related factors, identified preoperative LOS ≥ 1 day, Gliadel wafer implants, and postoperative CSF leakage as factors that achieved significance at the 0.05 level (Table 3). Chemotherapy on admission, preoperative glucose level ≥ 100 mg/dl, and procedure duration > 225 minutes (the duration cut point for the NHSN risk index)3 were marginally associated with increased risk of SSIs (p ≤ 0.10). The overall model had good predictive efficacy (c = 0.716; cross-validated c = 0.677).

TABLE 2:

Preoperative multivariate model assessing patient-related factors for SSIs after CRANI*

Patient-Related FactorsEstimateOR (95% CI)p ValueRisk Points
BMI >30 kg/m20.401.5 (0.9–2.5)0.124
any previous brain op0.481.6 (0.9–2.9)0.095
chemotherapy on admission2.047.7 (0.8–74.0)0.0820
preop LOS ≥1 day0.732.1 (1.2–3.4)0.0057
nontrauma procedure reason0.621.9 (0.9–3.7)0.086
preop glucose level ≥100 mg/dl0.481.6 (0.9–2.8)0.085

n = 378; c-statistic = 0.663; cross-validated c-statistic = 0.607; goodness-of-fit test, p = 0.90.

Data on BMI were missing for 9 controls.

We recorded the last glucose level obtained during the 30-day period before the procedure. Data were missing for 4 cases and 25 controls.

TABLE 3:

Overall multivariate model assessing patient-related and procedure-related risk factors for surgical site infections after CRANI*

Risk FactorsEstimateOR (95% CI)p Value
patient-related factors
 BMI >30 kg/m20.371.4 (0.9–2.4)0.15
 chemotherapy on admission1.856.3 (0.7–58.6)0.10
 preop LOS ≥1 day0.671.9 (1.1–3.3)0.01
 peripheral vascular disease1.826.2 (0.6–66.7)0.13
 preop glucose level ≥100 mg/dl0.551.7 (1.0–3.1)0.06
procedure-related factors
 op duration >225 mins0.451.6 (0.9–2.6)0.08
 Gliadel wafer implants1.906.7 (2.5–18.2)0.0002
 postop CSF leakage1.243.5 (1.4–8.5)0.007

n = 378; c-statistic = 0.716; cross-validated c-statistic = 0.677; goodness-of-fit test, p = 0.65.

Data on BMI were missing for 9 controls.

We recorded the last glucose level obtained during the 30-day period before procedure. Data were missing for 4 cases and 25 controls.

To develop a preoperative SSI risk index, we multiplied regression coefficient estimates from the preoperative model by 10 and rounded them to establish the points assigned for each risk factor (Table 2). The risk points for each factor were added to obtain the risk score, which predicted SSIs significantly better (c = 0.664; cross-validated c = 0.622) than the NHSN risk index (c = 0.547; p = 0.004) (Fig. 1). We categorized patients into 3 risk classes based on cut points that provided the highest c-statistic (Table 4). Patients in risk Class 2 or 3 had an increased risk of SSI over those in risk Class 1. Our risk index categorized 28% of cases and 8% of controls in the high-risk group (Class 3) compared with 24% of cases and 15% of controls for NHSN's risk index (Class 2 or 3).

Fig. 1.
Fig. 1.

Receiver operating characteristic (ROC) curves associated with the preoperative risk index score and the NHSN risk index score. The preoperative risk index had a better predictive efficacy (c-statistic = 0.664) than the NHSN risk index (c-statistic = 0.547; p = 0.004).

TABLE 4:

Estimated ORs for SSIs after CRANI using risk classes based on the preoperative risk index score

Risk ClassRisk Index ScoreOR (95% CI)*p Value*% Cases% ControlsPredictive Probability for SSIs
1<10reference12.030.612.4%
210–192.5 (1.3–4.9)0.00860.061.526.0%
3≥209.0 (4.0–20.5)<0.000128.07.956.0%

Calculated by univariate logistic regression.

The predictive probability could be interpreted as positive predictive value.

Outcome Study

The outcome study included 93 cases and 279 controls who underwent CRANI between January 1, 2006, and June 30, 2010. On average, relative to controls, patients with SSIs received antimicrobial agents on more days after CRANI during their initial hospitalizations and had longer hospital LOS during either initial hospitalizations or readmissions (Table 5). Because LOS was log-transformed and analyzed by linear regression, we should interpret the results as the multiplicative LOS increase in average LOS attributed to SSIs. For example, a multiplicative increase of 1.4 for LOS during the initial hospitalization indicated that an SSI will attribute to a 40% increase in the average LOS. Patients with SSIs were more likely than controls to have ≥ 2 readmissions, to have ≥ 2 reoperations to treat complications, and to die within 1 year after their procedures (Table 5).

TABLE 5:

Bivariate associations between 1-year postoperative outcomes and SSIs after CRANI*

OutcomeCases (n = 93)Controls (n = 279)Multiplicative Increase (95% CI) Attributed to SSIOR (95% CI) for SSIp Value
quantitative
 postop antimicrobial days7.8 ± 13.13.7 ± 9.42.0 (1.3–3.0)0.001
 postop LOS, days
  initial hospitalization12.2 ± 15.27.9 ± 11.01.4 (1.2–1.8)0.002
  readmission days9.0 ± 10.32.6 ± 12.32.2 (1.6–3.0)<0.0001
  both21.2 ± 15.710.4 ± 19.82.7 (2.2–3.3)<0.0001
 postop survival days315.1 ± 101.6324.0 ± 106.31.1 (0.9–1.4)0.40
binary
 readmission§50 (72.5)73 (30.5)6.0 (3.3–10.9)<0.0001
 no. of readmissions ≥2§19 (27.5)20 (8.4)4.2 (2.1–8.4)<0.0001
 reoperations§61 (88.4)62 (25.9)21.8 (9.9–48.0)<0.0001
 no. of reoperations ≥2§32 (46.4)15 (6.3)12.9 (6.4–26.1)<0.0001
 postop death24 (25.8)40 (14.3)2.1 (1.2–3.7)0.01

Statistics are number of patients (%) or mean ± SD.

Quantitative outcomes were log-transformed and analyzed by univariate linear regression.

Postoperative antimicrobial days excluding the first 2 days after the procedure during the initial hospitalization.

Readmissions and reoperations were evaluated for 69 cases and 239 controls who survived the 1st postoperative year.

The time from the CRANI to hospital discharge was ≥ 20 days for about 25% of cases and 15% of controls (Fig. 2A; p = 0.04). Among patients who survived ≥ 1 year, about 45% of cases and 15% of controls were readmitted within 50 days after their procedures (Fig. 2B; p < 0.0001), and about 65% of cases and 15% of controls had reoperations within 50 days of their procedures (Fig. 2C; p < 0.0001). Time-to-event curves assessing the time from the CRANI to death were different, but we did not compare the 2 curves because they crossed around the 100th postoperative day (Fig. 2D). Before the 100th day, the probability of survival was higher for cases than controls; thereafter, the probability of survival was higher for controls than cases. Acute medical conditions may have increased the risk of early death among controls because controls were more likely than cases to have head trauma or intracranial bleeding (34% vs 24%). We did a subgroup analysis by indication for procedure. Among patients with acute conditions of trauma or intracranial bleeding, the time-to-event curves of cases and controls did not cross and they were not significantly different (log-rank test, p = 0.87). Among patients with other indications, the 2 curves crossed at the 100th postoperative day, but the curves were almost identical before that. Cases were more likely than controls to die after the 100th postoperative day.

Fig. 2.
Fig. 2.

Kaplan-Meier time-to-event curves of postoperative outcomes within 1 year of craniotomy/craniectomy procedures. Postoperative LOS (A), readmissions (B), reoperations (C), and postoperative death (D).

Among 93 cases and 279 controls included in the outcome study, cases were more likely than controls to have longer preoperative LOS (2.1 ± 4.2 days vs 1.0 ± 2.7 days; p = 0.02), Charlson comorbidity indices ≥ 2 (54.8% vs 41.6%; p = 0.03),4 and nontraumatic reasons for CRANI (90.3% vs 78.1%; p = 0.009). Older age and worse Mc-Cabe and Jackson severity of illness scores were not significantly associated with SSIs. However, these factors could be clinically important. Thus, we included these factors and procedure month period in the multivariate models for each outcome to control for potential confounding effects. Compared with controls, patients with SSIs tended to remain in the hospital longer during their initial visits and during readmissions (Table 6). Surgical site infections were associated with an increased risk of readmission, reoperation, and postoperative death.

TABLE 6:

Multivariate analysis for postoperative outcomes associated with SSIs after CRANI*

OutcomeMultiplicative Increase (95% CI) Attributed to SSIOR (95% CI) for SSIp Value
qualitative
 postop LOS
  initial hospitalization1.5 (1.2–1.8)0.0001
  readmission days2.0 (1.4–2.8)0.0001
  both2.7 (2.2–3.3)<0.0001
binary
 readmissions§7.7 (4.0–14.9)<0.0001
 reoperations§36.0 (14.9–87.0)<0.0001
 death3.4 (1.5–7.4)0.003

Each outcome model was adjusted for preoperative LOS of more than 1 day, age, Charlson comorbidity index, nontrauma procedure reason, McCabe and Jackson severity of illness score, and procedure month period.

Quantitative outcomes were log-transformed and analyzed by multivariate linear regression.

Binary outcomes were analyzed by logistic regression analysis.

Readmissions and reoperations were evaluated for 69 cases and 239 controls who survived the 1st postoperative year.

Discussion

This study is unique in that we used robust analytical methods to assess risk factors and outcomes for all SSIs after all CRANI procedures. We comprehensively evaluated many risk factors for SSIs, controlled for potential confounding in multivariate analyses, and used appropriate analyses to evaluate postoperative outcomes of SSIs. Therefore, compared with previous studies, our study should provide less biased association estimates for risk factors and outcomes potentially related to SSIs after CRANI.

Patient-Related Risk Factors

Preoperative LOS was the only significant patient-related risk factor identified in the multivariate model. In contrast, studies by Korinek et al. and Lietard et al. identified preoperative LOS as a risk factor in bivariate analyses but not in multivariate analyses.19,22 These differences may be related to differences in the variables included in the studies and to differences in the patient populations. Neither Korinek et al. nor Lietard et al. collected data on patient comorbidities or on laboratory test results. In addition, Lietard et al. included patients who underwent spine operations, drain placements, or peripheral nerve operations, which may have different risk factors for SSIs than CRANI. In our study, CRANI done to treat brain tumors and Charlson comorbidity indices ≥ 2 were associated with an increased risk of preoperative hospital stay ≥ 1 day (OR 1.9 [95% CI 1.3–2.9]; OR 2.8 [95% CI 1.8–4.3]) and of SSIs (OR 1.6 [95% CI 1.0–2.5]; OR 1.6 [95% CI 1.0–2.5]). Thus, prolonged preoperative LOS may be a marker for patients with severe underlying conditions, which increased their risk of SSIs.

A history of previous brain operation, chemotherapy on admission, and preoperative glucose level achieved the significance level in our bivariate analyses but not in the multivariate analysis. However, these variables are clinically important. Moreover, Korinek's multivariate analysis found that a neurosurgical procedure in the 30 days before CRANI increased the risk of SSIs.17

In our study, patients who underwent CRANI to treat brain tumors were more likely than patients who underwent CRANI for other reasons to be on a chemotherapy regimen at admission and they had an increased risk of acquiring SSIs. Thus, the procedure reason could be an effect modifier for the association between other preoperative risk factors and SSIs. However, none of the interaction terms for the procedure reason and other preoperative factors was significant. Thus, the procedure reason did not modify the manner in which prolonged preoperative LOS, previous brain operation, and chemotherapy on admission were associated with an increased risk of SSIs.

Patients who have neurosurgical procedures often receive corticosteroids, which may elevate their glucose levels. Most prior studies have not found an association between diabetes or hyperglycemia and SSIs after neurosurgical procedures.9,20,24,30 Glycemic control has been associated with SSI rates after coronary artery bypass procedures.11,21 Diabetes was not a risk factor for SSIs in our study but a preoperative glucose level ≥ 100 mg/dl was marginally associated with a 1.6-fold increase in the odds of SSIs. Given that glucose levels are modifiable, further study of this potential risk factor is warranted.

Procedure-Related Risk Factors

Our overall model indicated that procedure-related factors were more strongly associated with SSIs than were patient-related factors. Gliadel wafer implants were the strongest risk factor for SSIs even after adjusting for chemotherapy on admission. We previously found an 8.4-fold increase in the odds of SSIs associated with Gliadel wafers, and Subach et al. found a 13.5-fold increase in the odds.5,36 However, Attenello et al. and Brem et al. did not find increased SSI rates associated with these implants.1,3 The differing results may be related, in part, to differences in study populations and to the fact that previous investigators did not account for potential confounders such as steroids, chemotherapy, or radiation therapy. The UIHC's neurosurgeons used Gliadel wafers in patients with advanced-stage brain tumors (for example, recurrent glioblastoma). Patients with Gliadel wafers were more likely than those without wafer implants to have Charlson comorbidity indices ≥ 2 (100% vs 41%; p < 0.0001) and McCabe and Jackson scores of fatal (100% vs 44%; p < 0.0001) and, thus, may have had increased risk for SSIs. We did not collect data on the tumor stage (for example, recurrent high-grade glioma), but we might have controlled for some confounding effect by including chemotherapy on admission in the overall multivariate model. Our neurosurgeons thought that most patients who received Gliadel wafers and subsequently acquired SSIs required multiple procedures to treat the infections and had bad outcomes. On the basis of both our results and their own experiences, our neurosurgeons significantly decreased Gliadel wafer use. However, we should not conclude that Gliadel wafers directly caused the SSIs because association did not establish causation.

Similar to our study, Korinek alone and with colleagues and Shinoura et al. found that CSF leakage increased the risk of SSIs after CRANI,17–19,35 whereas Kourbeti et al. found that CSF leakage was not associated with meningitis after CRANI.20 Special surgical techniques,13,27 synthetic absorbent sealants to close the wound or dura mater,10,37 or draining CSF during procedures25 may help prevent CSF leakage. Most of these interventions were used during procedures to remove pituitary tumors25 or vestibular schwannomas,13 and these methods may not be appropriate for other CRANI procedures. Like Gliadel wafers, CSF leakage might not cause SSIs. Instead SSIs and CSF leakage may both be measures of impaired wound healing. Future studies of techniques for obtaining watertight skin closures or for sealing the dura should assess whether they prevent CSF leakage and whether they prevent SSIs after CRANI.

Surgical Site Infection Risk Index

A risk index consisting of patient-related factors can help neurosurgeons identify patients at high risk for SSIs and allow infection preventionists to stratify patients' intrinsic risk of SSIs so they can compare SSI rates between hospitals or surgeons. The NHSN risk index had low efficacy in predicting whether patients would acquire SSIs after CRANI because the 3 factors (ASA score, wound contamination class, and operation duration) in the NHSN risk index often are not associated with SSIs after CRANI.5,17,19,32 Sánchez-Arenas and colleagues' risk index, which included the presence of chronic disease and nontrauma as the procedure reason, had better predictive discrimination than the NHSN risk index (c = 0.625 vs c = 0.558).32 However, this risk index may not be generalizable because the study did not include patients who received implants. Our risk index's predictive efficacy was superior to that of the NHSN risk index. Moreover, neurosurgeons could use it to determine the a priori risk of an SSI, regardless of the reason for the CRANI procedure. The cross-validated c-statistic (0.622) suggests how the index might perform on a validation sample. But this preoperative SSI risk index should be validated in another large patient population before it is applied widely.

Postoperative Outcomes

Studies by Kasatpibal et al. and O'Keeffe et al. found excess postoperative hospital stays of 31 days16 and 25 days,26 respectively, attributable to SSIs after CRANI procedures. Reichert et al. found that 30% of patients with SSIs died in the hospital but no control patient died.30 However, these analyses were not controlled for some possible confounders. Thus, bias may have affected their results. Unlike prior studies, we performed multivariate analyses to assess whether SSIs were associated with postoperative outcomes after adjustment for covariates that were potentially associated with these outcomes. We found that SSIs were associated with significantly prolonged LOS during initial hospitalizations and readmissions and with significantly increased risk of readmission, reoperation, and death. Similarly, our previous study found that 85% of patients with SSIs required additional operations to treat infections.5

Organisms Causing Surgical Site Infections

Staphylococcus aureus caused almost one-third of SSIs after CRANI, similar to the published finding that S. aureus caused 14%–51% of SSIs.5,17–19 Nasal carriers of S. aureus are at increased risk for S. aureus SSIs,28,38,39 and studies of other surgical specialties have shown that screening for S. aureus nasal carriage and decolonizing carriers significantly decreased S. aureus SSIs.2,29,31,34 A recent meta-analysis suggested that a bundle including screening for S. aureus nasal carriage, decolonizing carriers with mupirocin ointment, and chlorhexidine bathing before procedures reduced S. aureus SSIs.34 However, no published study has evaluated the effectiveness of S. aureus decolonization in reducing SSIs after neurosurgical procedures. A well-designed, quasi-experimental study using time-series analysis or a randomized clinical trial assessing the effectiveness of S. aureus decolonization in decreasing SSIs after CRANI would be a valuable contribution to the literature.

Gram-negative bacteria, which usually are not part of the normal skin flora, caused over 25% of SSIs. These infections could potentially be prevented by administering prophylactic agents that cover gram-negative bacteria. In our study, more than half of the patients received prophylaxis with nafcillin alone (49%) or in combination with other agents that cover gram-positive bacteria (5%), and 30% of patients received vancomycin alone (13%) or in combination with other agents that cover gram-positive bacteria (17%). Unlike cefazolin, the agent recommended for perioperative prophylaxis, these agents are not active against gram-negative bacteria. In our unpublished study of spinal operations, we found an increased risk of SSIs for patients receiving nafcillin or vancomycin as prophylaxis (Behen AZ, Pottinger JM, Chiang HY, et al: Risk factors for surgical site infections after spine operations. Presented at the 21st Annual Scientific Meeting of Society for Healthcare Epidemiology of America, 2011). On the basis of these observations, neurosurgeons at the UIHC began using cefazolin for all neurosurgical patients who are not allergic to cefazolin and who do not carry methicillin-resistant S. aureus. We are monitoring to determine if changing the prophylactic agent has decreased the SSI rate.

Study Limitations

This study has several limitations. First, we did not actively follow up on patients and we collected data from medical records. Thus, we might have missed patients who acquired SSIs or other adverse outcomes but did not seek care at the UIHC. Most patients who acquired SSIs should have been captured by our surveillance because infection preventionists at other hospitals in Iowa usually inform the UIHC's infection preventionists if patients who had operations at the UIHC are admitted with SSIs to their hospitals. Additionally, our neurosurgeons follow up on patients for at least 3 months, and most patients with serious surgical complications are referred back to the UIHC. Thus, we believe that the undetected outcomes were likely to have been minor and were likely to have occurred among patients without SSIs, which could have caused us to overestimate outcomes attributed to SSIs. Second, we did not match cases and controls by the indication for procedure, which makes it difficult to interpret the time-to-event curve for postoperative death. Acute conditions of trauma or intracranial bleeding might have attributed to early death among controls. When we completed a subgroup analysis for patients with acute conditions, the time-to-event curves of cases and controls did not cross and were not significantly different. For patients with other indications, the 2 curves remained to be crossed at the 100th postoperative day, but unlike the curves for all patients, they were very similar before the 100th postoperative day. The subgroup analyses revealed that the association between SSIs and postoperative death could be affected by the indication of the procedure, which should be taken into account when doing survival analysis for patients undergoing CRANI. Third, we did not validate the study results in another patient cohort. The patients admitted to the UIHC tend to have more severe medical conditions than patients admitted to community hospitals because the UIHC is a quaternary care, referral medical center. Therefore, our results may not be generalizable and should be validated in other patient populations.

Conclusions

Procedure-related factors were the strongest predictors of SSIs. Most patient-related risk factors reflect the patients' intrinsic risk of SSIs, which may be difficult to modify. Hyperglycemia could be modified, but further study is needed to determine if it is associated with an increased risk of SSI after CRANI. Staphylococcus aureus nasal carriage, Gliadel wafer implants, and postoperative CSF leakage may be modifiable. Given the adverse outcomes associated with SSIs and current reimbursement issues, studies of interventions to modify these risk factors could benefit patients and neurosurgery programs significantly.

Acknowledgment

We thank Barbara A. Zilles from Health Care Information Systems at the University of Iowa Hospitals and Clinics for creating the patient list for the current study.

Disclosure

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author contributions to the study and manuscript preparation include the following. Conception and design: Chiang, Pottinger, Greenlee, Howard, Herwaldt. Acquisition of data: Chiang, Kamath, Pottinger. Analysis and interpretation of data: Chiang. Drafting the article: Chiang. 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: Chiang. Statistical analysis: Chiang, Cavanaugh. Study supervision: Cavanaugh, Herwaldt.

This article contains some figures that are displayed in color online but in black-and-white in the print edition.

Portions of this work were presented in poster form at the 45th Annual Meeting of the Society for Epidemiologic Research, Minneapolis, Minnesota, June 27–30, 2012, and at IDWeek, San Diego, California, October 17–21, 2012.

References

  • 1

    Attenello FJMukherjee DDatoo GMcGirt MJBohan EWeingart JD: Use of Gliadel (BCNU) wafer in the surgical treatment of malignant glioma: a 10-year institutional experience. Ann Surg Oncol 15:288728932008

  • 2

    Bode LGKluytmans JAWertheim HFBogaers DVandenbroucke-Grauls CMRoosendaal R: Preventing surgical-site infections in nasal carriers of Staphylococcus aureus. N Engl J Med 362:9172010

  • 3

    Brem HPiantadosi SBurger PCWalker MSelker RVick NA: Placebo-controlled trial of safety and efficacy of intraoperative controlled delivery by biodegradable polymers of chemotherapy for recurrent gliomas. Lancet 345:100810121995

  • 4

    Charlson MEPompei PAles KLMacKenzie CR: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40:3733831987

  • 5

    Chiang HYSteelman VMPottinger JMSchlueter AJDiekema DJGreenlee JD: Clinical significance of positive cranial bone flap cultures and associated risk of surgical site infection after craniotomies or craniectomies. Clinical article. J Neurosurg 114:174617542011

  • 6

    Cosgrove SEQi YKaye KSHarbarth SKarchmer AWCarmeli Y: The impact of methicillin resistance in Staphylococcus aureus bacteremia on patient outcomes: mortality, length of stay, and hospital charges. Infect Control Hosp Epidemiol 26:1661742005

  • 7

    Division of Healthcare Quality Promotion National Center for Preparedness Detection and Control of Infectious Diseases: Surgical Site Infection (SSI) Event. The National Healthcare Safety Network (NHSN) Manual: Patient Safety Component Protocol AtlantaCDC2009. 9-19-14(http://www.cdc.gov/nhsn/PDFs/pscManual/pscManual_current.pdf) [Accessed September 24 2013]

  • 8

    Edwards JRPeterson KDMu YBanerjee SAllen-Bridson KMorrell G: National Healthcare Safety Network (NHSN) report: data summary for 2006 through 2008, issued December 2009. Am J Infect Control 37:7838052009

  • 9

    Erman TDemirhindi HGöçer AITuna MIldan FBoyar B: Risk factors for surgical site infections in neurosurgery patients with antibiotic prophylaxis. Surg Neurol 63:1071132005

  • 10

    Ferroli PAcerbi FBroggi MSchiariti MAlbanese ETringali G: A novel impermeable adhesive membrane to reinforce dural closure: a preliminary retrospective study on 119 consecutive high-risk patients. World Neurosurg 79:5515572013

  • 11

    Furnary APZerr KJGrunkemeier GLStarr A: Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg 67:3523621999

  • 12

    Gaberel TBorgey FThibon PLesteven CLecoutour XEmery E: Surgical site infection associated with the use of bovine serum albumine-glutaraldehyde surgical adhesive (BioGlue) in cranial surgery: a case-control study. Acta Neurochir (Wien) 153:1561632011

  • 13

    Goddard JCOliver ERLambert PR: Prevention of cerebrospinal fluid leak after translabyrinthine resection of vestibular schwannoma. Otol Neurotol 31:4734772010

  • 14

    Hardy SJNowacki ASBertin MWeil RJ: Absence of an association between glucose levels and surgical site infections in patients undergoing craniotomies for brain tumors. Clinical article. J Neurosurg 113:1611662010

  • 15

    Healthcare Cost and Utilization Project Agency for Healthcare Research and Quality: Overview of the Nationwide Inpatient Sample (NIS) (http://www.hcup-us.ahrq.gov/nisoverview.jsp) [Accessed September 24 2013]

  • 16

    Kasatpibal NThongpiyapoom SNarong MNSuwalak NJamulitrat S: Extra charge and extra length of postoperative stay attributable to surgical site infection in six selected operations. J Med Assoc Thai 88:108310912005

  • 17

    Korinek AM: Risk factors for neurosurgical site infections after craniotomy: a prospective multicenter study of 2944 patients. The French Study Group of Neurosurgical Infections, the SEHP, and the C-CLIN Paris-Nord. Neurosurgery 41:107310811997

  • 18

    Korinek AMBaugnon TGolmard JLvan Effenterre RCoriat PPuybasset L: Risk factors for adult nosocomial meningitis after craniotomy: role of antibiotic prophylaxis. Neurosurgery 59:1261332006

  • 19

    Korinek AMGolmard JLElcheick ABismuth Rvan Effenterre RCoriat P: Risk factors for neurosurgical site infections after craniotomy: a critical reappraisal of antibiotic prophylaxis on 4,578 patients. Br J Neurosurg 19:1551622005

  • 20

    Kourbeti ISJacobs AVKoslow MKarabetsos DHolzman RS: Risk factors associated with postcraniotomy meningitis. Neurosurgery 60:3173262007

  • 21

    Latham RLancaster ADCovington JFPirolo JSThomas CS Jr: The association of diabetes and glucose control with surgical-site infections among cardiothoracic surgery patients. Infect Control Hosp Epidemiol 22:6076122001

  • 22

    Lietard CThébaud VBesson GLejeune B: Risk factors for neurosurgical site infections: an 18-month prospective survey. Clinical article. J Neurosurg 109:7297342008

  • 23

    McCabe WRJackson GG: Gram-negative bacteremia. I. Etiology and ecology. Arch Intern Med 110:8478551962

  • 24

    McClelland S IIIHall WA: Postoperative central nervous system infection: incidence and associated factors in 2111 neurosurgical procedures. Clin Infect Dis 45:55592007

  • 25

    Mehta GUOldfield EH: Prevention of intraoperative cerebrospinal fluid leaks by lumbar cerebrospinal fluid drainage during surgery for pituitary macroadenomas. Clinical article. J Neurosurg 116:129913032012

  • 26

    O'Keeffe ABLawrence TBojanic S: Oxford craniotomy infections database: a cost analysis of craniotomy infection. Br J Neurosurg 26:2652692012

  • 27

    Park JSKong DSLee JAPark K: Intraoperative management to prevent cerebrospinal fluid leakage after microvascular decompression: dural closure with a “plugging muscle” method. Neurosurg Rev 30:1391422007

  • 28

    Perl TMGolub JE: New approaches to reduce Staphylococcus aureus nosocomial infection rates: treating S. aureus nasal carriage. Ann Pharmacother 32:1 SupplS7S161998

  • 29

    Rao NCannella BACrossett LSYates AJ JrMcGough RL IIIHamilton CW: Preoperative screening/decolonization for Staphylococcus aureus to prevent orthopedic surgical site infection: prospective cohort study with 2-year follow-up. J Arthroplasty 26:150115072011

  • 30

    Reichert MCMedeiros EAFerraz FA: Hospital-acquired meningitis in patients undergoing craniotomy: incidence, evolution, and risk factors. Am J Infect Control 30:1581642002

  • 31

    Richer SLWenig BL: The efficacy of preoperative screening and the treatment of methicillin-resistant Staphylococcus aureus in an otolaryngology surgical practice. Otolaryngol Head Neck Surg 140:29322009

  • 32

    Sánchez-Arenas RRivera-García BEGrijalva-Otero IJuárez-Cedillo Tdel Carmen Martínez-García MRangel-Frausto S: Factors associated with nosocomial surgical-site infections for craniotomy in Mexico City hospitals. Cir Cir 78:5132010

  • 33

    SAS Institute: Usage Note 39724: ROC analysis using validation data and crossvalidation (http://support.sas.com/kb/39/724.html) [Accessed September 24 2013]

  • 34

    Schweizer MPerencevich EMcDanel JCarson JFormanek MHafner J: Effectiveness of a bundled intervention of decolonization and prophylaxis to decrease Gram positive surgical site infections after cardiac or orthopedic surgery: systematic review and meta-analysis. BMJ 346:f27432013

  • 35

    Shinoura NYamada ROkamoto KNakamura O: Early prediction of infection after craniotomy for brain tumours. Br J Neurosurg 18:5986032004

  • 36

    Subach BRWitham TFKondziolka DLunsford LDBozik MSchiff D: Morbidity and survival after 1,3-bis(2-chloroethyl)-1-nitrosourea wafer implantation for recurrent glioblastoma: a retrospective case-matched cohort series. Neurosurgery 45:17231999

  • 37

    von der Brelie CSoehle MClusmann HR: Intraoperative sealing of dura mater defects with a novel, synthetic, self adhesive patch: application experience in 25 patients. Br J Neurosurg 26:2312352012

  • 38

    von Eiff CBecker KMachka KStammer HPeters G: Nasal carriage as a source of Staphylococcus aureus bacteremia. N Engl J Med 344:11162001

  • 39

    Wenzel RPPerl TM: The significance of nasal carriage of Staphylococcus aureus and the incidence of postoperative wound infection. J Hosp Infect 31:13241995

Appendix Table 1:

ICD-9-CM procedure codes for CRANI procedures

ICD-9-CM Procedure CodeDefinition
01.12open biopsy of cerebral meninges
01.14open biopsy of brain
01.21incision & drainage of cranial sinus
01.22removal of intracranial neurostimulator lead(s)
01.23reopening of craniotomy site
01.24other craniotomy
01.25other craniectomy
01.28placement of intracerebral catheter(s) via bur hole(s)
01.31incision of cerebral meninges
01.32lobotomy & tractotomy
01.39other incision of brain
01.41operations on thalamus
01.42operations on globus pallidus
01.51excision of lesion or tissue of cerebral meninges
01.52hemispherectomy
01.53lobectomy of brain
01.59other excision or destruction of lesion or tissue of brain
02.11simple suture of dura mater of brain
02.12other repair of cerebral meninges
02.13ligation of meningeal vessel
02.14choroid plexectomy
02.91lysis of cortical adhesions
02.92repair of brain
02.93implantation or replacement of intracranial neurostimulator lead(s)
07.51exploration of pineal
07.52incision of pineal gland
07.53partial excision of pineal gland
07.54total excision of pineal gland
07.59other operations on pineal gland
07.61partial excision of pituitary gland, transfrontal approach
07.62partial excision of pituitary gland, transsphenoidal approach
07.63partial excision of pituitary gland, unspecified approach
07.64total excision of pituitary gland, transfrontal approach
07.65total excision of pituitary gland, transsphenoidal approach
07.68total excision of pituitary gland, other specified approach
07.69total excision of pituitary gland, unspecified approach
07.71exploration of pituitary fossa
07.72incision of pituitary gland
07.79other
38.01incision of vessel, intracranial vessels
38.11endarterectomy, intracranial vessels
38.31resection of vessel w/ anastomosis, intracranial vessels
38.41resection of vessel w/ replacement, intracranial vessels
38.51ligation & stripping of varicose veins, intracranial vessels
38.61other excision of vessel, intracranial vessels
38.81other surgical occlusion of vessels, intracranial vessels
39.28extracranial-intracranial vascular bypass
Appendix Table 2:

Variables evaluated in the current study

Patient-Related FactorsProcedure-Related Factors
agepreop antimicrobial prophylaxis
sexskin preparation
Charlson comorbidity indexperiop laboratory test results (e.g., glucose, hemoglobin, white blood cell count)
type of health insuranceprocedure duration
smoking historyGliadel wafer implant
BMIsurgeons
history of brain operation or radiationICD-9-CM procedure codes
scheduling of procedures (i.e., emergency, urgent, scheduled)periop transfusion
preop LOSintracranial pressure monitoring
reason for CRANI (i.e., tumor, hemorrhage, trauma, biopsy, infection, other)postop CSF leakage w/in 30 days after CRANI
wound classificationperiop ventricular or lumbar drains
ASA score
ICD-9-CM diagnosis codes
McCabe & Jackson severity of illness score
medications on admission (chemotherapy, steroid use, immunosuppressive drugs, hypoglycemic agents, & insulin)
Appendix Table 3:

Organisms causing SSIs after CRANI by depth of infection*

Organism(s)No. of Cases
Superficial IncisionalDeep IncisionalOrgan/Space
Acinetobacter spp.010
alpha hemolytic streptococcus, not enterococcus001
alpha hemolytic streptococcus, not Enterococcus & Neisseria spp.001
Candida albicans101
CoNS138
CoNS & Enterobacter aerogenes001
CoNS & Propionibacterium acnes014
CoNS & Pseudomonas aeruginosa001
CoNS & Staphylococcus lugdunensis010
Escherichia coli021
E. aerogenes022
Enterobacter cloacae011
E. cloacae & Acinetobacter baumannii010
Haemophilus influenzae011
Klebsiella pneumoniae011
P. acnes024
P. acnes & alpha hemolytic streptococcus, not enterococcus001
P. acnes & Corynebacterium spp.010
P. acnes & K. pneumoniae010
probable Peptostreptococcus spp.010
P. aeruginosa010
P. aeruginosa & Enterococcus spp.100
Staphylococcus aureus188
S. aureus & gram-positive rods suggestive of diphtheroids & C. albicans001
S. aureus & P. acnes020
S. aureus & Peptostreptococcus spp. & Citrobacter koseri001
S. aureus & Prevotella spp.001
S. aureus & Serratia marcescens001
MRSA151
MRSA & E. cloacae010
MRSA & Enterococcus spp. & E. coli & Peptostreptococcus spp.010
MRSA & P. aeruginosa001
Stenotrophomonas maltophilia001
mixed flora200
no growth2111
not cultured300
total123854

CoNS = coagulase-negative staphylococci; MRSA = methicillin-resistant Staphylococcus aureus.

Gram-negative bacteria.

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

Article Information

Address correspondence to: Hsiu-Yin Chiang, Ph.D., Department of Internal Medicine, University of Iowa Carver College of Medicine, 200 Hawkins Dr., Iowa City, IA 52242. email: hsiu-yin-chiang@uiowa.edu.

Please include this information when citing this paper: published online November 8, 2013; DOI: 10.3171/2013.9.JNS13843.

© AANS, except where prohibited by US copyright law.

Headings

Figures

  • View in gallery

    Receiver operating characteristic (ROC) curves associated with the preoperative risk index score and the NHSN risk index score. The preoperative risk index had a better predictive efficacy (c-statistic = 0.664) than the NHSN risk index (c-statistic = 0.547; p = 0.004).

  • View in gallery

    Kaplan-Meier time-to-event curves of postoperative outcomes within 1 year of craniotomy/craniectomy procedures. Postoperative LOS (A), readmissions (B), reoperations (C), and postoperative death (D).

References

1

Attenello FJMukherjee DDatoo GMcGirt MJBohan EWeingart JD: Use of Gliadel (BCNU) wafer in the surgical treatment of malignant glioma: a 10-year institutional experience. Ann Surg Oncol 15:288728932008

2

Bode LGKluytmans JAWertheim HFBogaers DVandenbroucke-Grauls CMRoosendaal R: Preventing surgical-site infections in nasal carriers of Staphylococcus aureus. N Engl J Med 362:9172010

3

Brem HPiantadosi SBurger PCWalker MSelker RVick NA: Placebo-controlled trial of safety and efficacy of intraoperative controlled delivery by biodegradable polymers of chemotherapy for recurrent gliomas. Lancet 345:100810121995

4

Charlson MEPompei PAles KLMacKenzie CR: A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 40:3733831987

5

Chiang HYSteelman VMPottinger JMSchlueter AJDiekema DJGreenlee JD: Clinical significance of positive cranial bone flap cultures and associated risk of surgical site infection after craniotomies or craniectomies. Clinical article. J Neurosurg 114:174617542011

6

Cosgrove SEQi YKaye KSHarbarth SKarchmer AWCarmeli Y: The impact of methicillin resistance in Staphylococcus aureus bacteremia on patient outcomes: mortality, length of stay, and hospital charges. Infect Control Hosp Epidemiol 26:1661742005

7

Division of Healthcare Quality Promotion National Center for Preparedness Detection and Control of Infectious Diseases: Surgical Site Infection (SSI) Event. The National Healthcare Safety Network (NHSN) Manual: Patient Safety Component Protocol AtlantaCDC2009. 9-19-14(http://www.cdc.gov/nhsn/PDFs/pscManual/pscManual_current.pdf) [Accessed September 24 2013]

8

Edwards JRPeterson KDMu YBanerjee SAllen-Bridson KMorrell G: National Healthcare Safety Network (NHSN) report: data summary for 2006 through 2008, issued December 2009. Am J Infect Control 37:7838052009

9

Erman TDemirhindi HGöçer AITuna MIldan FBoyar B: Risk factors for surgical site infections in neurosurgery patients with antibiotic prophylaxis. Surg Neurol 63:1071132005

10

Ferroli PAcerbi FBroggi MSchiariti MAlbanese ETringali G: A novel impermeable adhesive membrane to reinforce dural closure: a preliminary retrospective study on 119 consecutive high-risk patients. World Neurosurg 79:5515572013

11

Furnary APZerr KJGrunkemeier GLStarr A: Continuous intravenous insulin infusion reduces the incidence of deep sternal wound infection in diabetic patients after cardiac surgical procedures. Ann Thorac Surg 67:3523621999

12

Gaberel TBorgey FThibon PLesteven CLecoutour XEmery E: Surgical site infection associated with the use of bovine serum albumine-glutaraldehyde surgical adhesive (BioGlue) in cranial surgery: a case-control study. Acta Neurochir (Wien) 153:1561632011

13

Goddard JCOliver ERLambert PR: Prevention of cerebrospinal fluid leak after translabyrinthine resection of vestibular schwannoma. Otol Neurotol 31:4734772010

14

Hardy SJNowacki ASBertin MWeil RJ: Absence of an association between glucose levels and surgical site infections in patients undergoing craniotomies for brain tumors. Clinical article. J Neurosurg 113:1611662010

15

Healthcare Cost and Utilization Project Agency for Healthcare Research and Quality: Overview of the Nationwide Inpatient Sample (NIS) (http://www.hcup-us.ahrq.gov/nisoverview.jsp) [Accessed September 24 2013]

16

Kasatpibal NThongpiyapoom SNarong MNSuwalak NJamulitrat S: Extra charge and extra length of postoperative stay attributable to surgical site infection in six selected operations. J Med Assoc Thai 88:108310912005

17

Korinek AM: Risk factors for neurosurgical site infections after craniotomy: a prospective multicenter study of 2944 patients. The French Study Group of Neurosurgical Infections, the SEHP, and the C-CLIN Paris-Nord. Neurosurgery 41:107310811997

18

Korinek AMBaugnon TGolmard JLvan Effenterre RCoriat PPuybasset L: Risk factors for adult nosocomial meningitis after craniotomy: role of antibiotic prophylaxis. Neurosurgery 59:1261332006

19

Korinek AMGolmard JLElcheick ABismuth Rvan Effenterre RCoriat P: Risk factors for neurosurgical site infections after craniotomy: a critical reappraisal of antibiotic prophylaxis on 4,578 patients. Br J Neurosurg 19:1551622005

20

Kourbeti ISJacobs AVKoslow MKarabetsos DHolzman RS: Risk factors associated with postcraniotomy meningitis. Neurosurgery 60:3173262007

21

Latham RLancaster ADCovington JFPirolo JSThomas CS Jr: The association of diabetes and glucose control with surgical-site infections among cardiothoracic surgery patients. Infect Control Hosp Epidemiol 22:6076122001

22

Lietard CThébaud VBesson GLejeune B: Risk factors for neurosurgical site infections: an 18-month prospective survey. Clinical article. J Neurosurg 109:7297342008

23

McCabe WRJackson GG: Gram-negative bacteremia. I. Etiology and ecology. Arch Intern Med 110:8478551962

24

McClelland S IIIHall WA: Postoperative central nervous system infection: incidence and associated factors in 2111 neurosurgical procedures. Clin Infect Dis 45:55592007

25

Mehta GUOldfield EH: Prevention of intraoperative cerebrospinal fluid leaks by lumbar cerebrospinal fluid drainage during surgery for pituitary macroadenomas. Clinical article. J Neurosurg 116:129913032012

26

O'Keeffe ABLawrence TBojanic S: Oxford craniotomy infections database: a cost analysis of craniotomy infection. Br J Neurosurg 26:2652692012

27

Park JSKong DSLee JAPark K: Intraoperative management to prevent cerebrospinal fluid leakage after microvascular decompression: dural closure with a “plugging muscle” method. Neurosurg Rev 30:1391422007

28

Perl TMGolub JE: New approaches to reduce Staphylococcus aureus nosocomial infection rates: treating S. aureus nasal carriage. Ann Pharmacother 32:1 SupplS7S161998

29

Rao NCannella BACrossett LSYates AJ JrMcGough RL IIIHamilton CW: Preoperative screening/decolonization for Staphylococcus aureus to prevent orthopedic surgical site infection: prospective cohort study with 2-year follow-up. J Arthroplasty 26:150115072011

30

Reichert MCMedeiros EAFerraz FA: Hospital-acquired meningitis in patients undergoing craniotomy: incidence, evolution, and risk factors. Am J Infect Control 30:1581642002

31

Richer SLWenig BL: The efficacy of preoperative screening and the treatment of methicillin-resistant Staphylococcus aureus in an otolaryngology surgical practice. Otolaryngol Head Neck Surg 140:29322009

32

Sánchez-Arenas RRivera-García BEGrijalva-Otero IJuárez-Cedillo Tdel Carmen Martínez-García MRangel-Frausto S: Factors associated with nosocomial surgical-site infections for craniotomy in Mexico City hospitals. Cir Cir 78:5132010

33

SAS Institute: Usage Note 39724: ROC analysis using validation data and crossvalidation (http://support.sas.com/kb/39/724.html) [Accessed September 24 2013]

34

Schweizer MPerencevich EMcDanel JCarson JFormanek MHafner J: Effectiveness of a bundled intervention of decolonization and prophylaxis to decrease Gram positive surgical site infections after cardiac or orthopedic surgery: systematic review and meta-analysis. BMJ 346:f27432013

35

Shinoura NYamada ROkamoto KNakamura O: Early prediction of infection after craniotomy for brain tumours. Br J Neurosurg 18:5986032004

36

Subach BRWitham TFKondziolka DLunsford LDBozik MSchiff D: Morbidity and survival after 1,3-bis(2-chloroethyl)-1-nitrosourea wafer implantation for recurrent glioblastoma: a retrospective case-matched cohort series. Neurosurgery 45:17231999

37

von der Brelie CSoehle MClusmann HR: Intraoperative sealing of dura mater defects with a novel, synthetic, self adhesive patch: application experience in 25 patients. Br J Neurosurg 26:2312352012

38

von Eiff CBecker KMachka KStammer HPeters G: Nasal carriage as a source of Staphylococcus aureus bacteremia. N Engl J Med 344:11162001

39

Wenzel RPPerl TM: The significance of nasal carriage of Staphylococcus aureus and the incidence of postoperative wound infection. J Hosp Infect 31:13241995

TrendMD

Metrics

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 358 358 67
PDF Downloads 246 246 42
EPUB Downloads 0 0 0

PubMed

Google Scholar