Risk and Protective Factors Associated with Surgical Infections among Spine Patients

Search strategy and inclusion criteria

Two independent investigators (MS, AB) conducted a literature review. The Cumulative Index to Nursing and Allied Health Literature (CINAHL), PubMed, Ovid Medline, and Ebsco databases were searched for articles meeting the review criteria published from January 2003 through October 2015. A hand search of articles was conducted by the same investigators. The search strategy implemented combinations of the following terms: (“spine surgery”) AND (“infection” OR “infection risk” OR “post-operative infection”) AND (“orthopedic”). The terms (“patient risk factors”) OR (“infection prevention factors”) OR (“risk factors”) were used as MESH terms in the search. There were 842 articles identified using the search strategy. The articles included, based on inclusion and exclusion criteria, were determined using the PRISMA method for identification of the articles.

Of the 842 identified articles, those selected for full review were included if a main outcome addressed at least one of the following: (1) Patient-related infection risk factors in orthopedic spinal operation; (2) surgery-specific infection risk factors in orthopedic spinal operation; (3) operation-protective factors in orthopedic spinal operation; (4) educational tools to assist spine surgeons in preventing SSIs; (5) epidemiology of SSIs in spinal operation; (6) evaluation of the efficacy of specific interventions in reducing infection rates.

The types of studies reviewed were limited to: Systematic reviews, meta-analyses, retrospective case/chart reviews, retrospective case control studies, retrospective cohort studies, and prospective randomized clinical trials. Both quantitative and qualitative articles were included. A hand search was completed subsequently, and two additional articles meeting these inclusion criteria were identified. In all, 29 studies were identified that met the inclusion criteria. The review was limited to studies in the English Language.

For each of the 29 studies included, using the PRISMA criteria for inclusion and exclusion, the following information was extracted: Author(s); year of publication; article title; journal of publication; objectives of the study; sample size and characteristics; study design; study results; study limitations; and level of evidence. Specific information focused on protective and risk factors relative to SSI are summarized in Tables 1–3.

Analysis

Quantitative analysis was not used because of the heterogeneity of the studies' design. Data collected from each study included: Study characteristics (authors, year of publication, article title, journal of publication); objectives of the study; sample size and characteristics; study design; study results (SSI protective factors, SSI risk factors, infection rates); study limitations; and level of evidence.

Patient associated risk factors that increase SSI in post-operative orthopedic spine patients

A number of patient-specific and operation-specific variables have been demonstrated to confer an increased risk of post-operative SSIs among patients undergoing orthopedic spine operation. Review of the current literature identified the most cited patient risk factors for post-operative SSI as high body mass index (BMI)/obesity (BMI >30 or 35), followed by diabetes mellitus (Table 1). The literature documents a number of other factors associated with increased risk of SSI after an orthopedic spinal operation.

In a 2013 meta-analysis, Abdallah et al. [7] evaluated BMI as an independent risk factor for SSI after controlling for diabetes mellitus and for heterogeneity in definitions of obesity and BMI groupings in the literature. In this study, the authors calculated BMI as a risk factor using a model in which slope estimate units were used to establish SSI risk associated with five-unit increases in BMI. The results of this study found that increased BMI has a statistically significant positive association with an increased risk of SSI. When adjusted for confounders such as diabetes mellitus, the authors found that every five-unit increase in BMI resulted in a 21% increase in risk of spinal SSI [7].

The authors attributed this increase in risk to greater retraction forces necessary for surgical incision closure in patients with a greater amount of subcutaneous tissue. The study team hypothesized that increases in dead-space on surgical incision closure may yield increased tissue necrosis, poor vascular perfusion, and impaired bacterial oxidative killing, thus increasing the overall risk for a post-operative SSI among obese patients.

The link between obesity and post-operative SSI is not universally accepted. For example, Watanabe et al. [8] found no significant increase in SSI among obese patients who had undergone spinal operation. Further, this study demonstrated no evidence of increased SSI risk associated with advanced age (>60 y), gender, or smoking history, three commonly cited risk factors found in a number of studies (Table 1).

Documented in 12 studies examined in this review, diabetes mellitus (types I and/or II) was the second most commonly cited patient-related risk factor associated with SSI in orthopedic spinal operation [8–20]. Variation in odds ratio (OR) calculations for diabetes mellitus as a risk factor differed, however, across the studies reviewed. In a 2008 retrospective, case-controlled study of 2,316 patients who underwent spinal operation, a diagnosis of diabetes mellitus was associated with an increased risk of post-operative SSI. The study concluded that persons with diabetes mellitus have an increased risk for a post-operative SSI developing that was more than eight times greater than persons without diabetes mellitus (OR = 8.4; confidence interval [CI], 95% = 3.5–19.8; p < 0.001) [16]. Further, the authors of this study found that pre-operative serum glucose levels greater than 125 mg/dL and/or post-operative serum glucose levels greater than 200 mg/dl as independent risk factors of SSI (OR = 3.3; CI 95% = 1.5–7.3; p = 0.004) [16].

Similarly, in another study aimed at identifying factors associated with increased risk of SSI, Olsen et al. [20] found that a serum glucose level ≥200 mg/dL at any time during the peri-operative phase (pre, intra, post) was significantly associated with an increased risk of an SSI developing among patients who had undergone laminectomy of spinal fusion surgery (n = 41) compared with a sub-sample of patients in whom an SSI had not developed (n = 179) (15, 36.6% of infected persons; OR = 3.0; CI, 95% = 1.4–6.3; p = 0.003.

Although many of the studies reviewed found that diabetes mellitus was associated positively with an increased risk of SSI among surgical spine patients, conflicting evidence relative to the SSI and diabetes mellitus based on the type exists. For example, in a retrospective study of data (N = 24,774) obtained from the Veterans Affairs National Surgical Improvement Program Database of patients who had undergone a spinal decompression and fusion, the risk associated with insulin dependent diabetes mellitus was found to be much lower. Patients with insulin dependent diabetes mellitus had a significantly increased risk (OR = 1.50; 95% CI 1, 13–2.00) for SSI compared with those without diabetes mellitus or with diet-controlled diabetes mellitus [19]. Patients with diabetes mellitus who controlled their glucose levels with modifications to diets were not found to have an increased risk for SSI [19]. There was no significant risk of SSIs associated with patients with diabetes mellitus who controlled glucose levels by taking oral medications (OR = 0.89; 95% CI 0.68–1.16) [19].

Other studies reviewed concluded that regardless of the type of diabetes mellitus, the presence of diabetes mellitus in itself was associated with an increased risk of SSI among patients undergoing spinal operation. For example, a study by Kurtz et al. [15] documented that diabetes mellitus (regardless of type) is an independent risk factor for early deep tissue infection (defined as infection within two years of surgery). The researchers, however, found no association between diabetes mellitus and overall infection risk.

A 2015 study by Lewkonia et al. [21] examined patient-specific risk factors associated with the onset of delayed infections (>90 d) after lumbar spinal fusion operation. A retrospective case review of 5,770 consecutive patients who had undergone spinal operation between the years 2000 and 2008 (with a total of 162 post-operative infections), SSIs had a delayed onset of three months or more (n = 7, 4.3%). Known risk factors were present in six of the seven patients in whom an SSI had developed [21]. The risk factors associated with delayed infection among the patients in whom an SSI developed included: Surgical complication resulting in significant incision drainage, early remote methicillin-resistant Staphylococcus aureus (MRSA) infection, post-operative sepsis, history of infection after a procedure, neurogenic bladder, spinal abscesses, and intravenous (IV) drug use [21].

Additional patient-specific risk factors identified by more than one of the reviewed studies included: Advanced age, current or history of cigarette smoking, alcohol abuse, cancer, malnutrition, incontinence, high American Society of Anesthesiologists (ASA) class, hypertension, renal disease, anemia/low pre-operative hematocrit, and history of SSI (Table 1).

Operation-associated risk factors associated with increased SSI among surgical spine patients

A number of operation-specific risk factors have been documented to increase the risk of SSI among post-operative orthopedic spine patients (Table 2). In the articles reviewed, the risk factor most often documented with an association for increased risk of post-operative SSI was increased length of operation [8,9,15,17,20,22]. Although many studies identified increased estimated blood loss (EBL) to be a risk factor for SSI, this factor may be patient-related versus operation-specific; therefore, EBL was not determined to be the factor most often documented as an operation-associated risk factor. Six independent studies documented that increased length of operation was associated with increased risk of development of a post-operative SSI among patients undergoing orthopedic spinal operation (Table 2). As the length of surgery increased, the risk of SSI increased.

Many of the studies reviewed concluded that the type of procedure and the indication for the operation were associated with different risks of post-operative infection. In a study of 108,419 patients, Smith et al. [23] found that a patient undergoing a surgical procedure for acute osteomyelitis/discitis had the greatest risk of post-operative infection in the adult population with a post-operative SSI developing in 5.1% (n = 866) of patients studied. Among pediatric patients, a surgical procedure to treat kyphosis was found to be the greatest risk factor for post-operative SSI with 5.4% of surgeries resulting in infection (n = 1,555). Patients undergoing spinal operation for a degenerative disease were documented as having the lowest risk of SSI: 1.4% for adults (n = 46,434) and 0.9% in the pediatric population (n = 654) [23].

Watanabe et al. [8] concluded that the operation-specific risk factor with the greatest correlation to post-operative SSI was a surgery performed for a trauma-related indication, (compared with an elective surgical procedure). In this retrospective review, the team found that traumatic indications for operation were associated with almost 10 times greater risk of SSI compared with non-trauma–related indications for spinal operation (OR = 9.42; 95% CI 1.59–55.73; p = 0.013) [8].

Minimally invasive techniques were associated with lesser infection rates (2.4%–0.5%) in a study by Chahoud et al. [9]. The findings of Smith et al. [23] support the finding by Chahoud et al. [9]. Smith et al. [23] documented a 1.1% SSI rate in open procedures compared with a 0.4% SSI rate among patients who had undergone a surgical procedure using a minimally invasive technique. Smith et al. [23] based their findings on a large sample (N = 108,419) of procedures.

A number of studies examined the relationship between infection rate and the location of the spinal operation. Procedures involving the thoracic spine were documented as having the greatest risk of infection with 2.1% (N = 524) among patients in whom a post-operative SSI developed, while in comparison, the infection rate for lumbar procedures was 1.6% (N = 33,910) and 0.8% (N = 11,674) for cervical procedures [9]. Cizik et al. [10] reached a similar conclusion, determining relative risk of SSI in thoracic versus cervical surgery to be 2.57 (95% CI, 1.20–5.60; p = 0.01), and relative risk for lumbosacral surgery versus cervical surgery to be 2.03 (95% CI 1.10–4.05).

The surgical approach and the use of implants are identified by many of the studies as factors associated with increased risk of SSI among surgical spine patients. Four studies [11,12,17,20] found that spinal operation performed using a posterior approach significantly increased the overall risk of an SSI, while three studies found that surgical procedures involving the use of implants were associated with an increased risk of SSI [9,14,23]. In contrast, Olsen et al. [20] found no evidence that the use of implants increased the risk of SSI among patients undergoing spinal operation. The designs and methods of analysis across these studies differed and may account for some of the differences in findings and conclusions.

Seven studies identified a blood transfusion and higher than expected EBL as independent factors associated with increased risk of post-operative SSI among surgical spine patients (Table 2).

One study included in this review investigated gender differences (male vs. female) and risk of post-operative SSI [24]. The researchers concluded that evidence exists to support an association between male gender and an increased risk of death after spinal operation; however, no difference exists between males and females relative to post-operative complications or rates of infection (OR = 1.04; 95% CI 0.95–1.13; p = 0.42) [24].

Protective factors against SSI among orthopedic spine patients (Table 3)

Of the studies reviewed, two identified rates of infection after the implementation of a bundle protocol aimed to decrease post-operative SSIs. Clinically meaningful reductions in SSIs were documented by Ballard et al. [25] after the implementation of an evidence-based bundle termed “Target Zero.” The multi-disciplinary developed bundle aimed to decrease infection rates among pediatric patients undergoing elective spinal operation.

The task force, a team at Children's Hospital Colorado, identified five pre-operative interventions: (1) A 4% chlorhexidine wash by the patient 24 hours before operation; (2) MRSA nasal culturing 30 days before operation; (3) 2% chlorhexidine wipe to the site the day of operation; (4) the use of pre-operative warming blankets; and (5) prophylactic antibiotics. They identified five intra-operative initiatives: (1) Proper gowning/gloving technique for line placement; (2) appropriate use of antibiotic agents before incision; (3) appropriate antibiotic re-dosing; (4) minimizing operating room personnel to essential personnel only during the operation; and (5) use of warming blankets; and three post-operative interventions: (1) discontinuation of antibiotic agents 24 hours post-operation; (2) removal of drains within 48 hours post-operation; and (3) prompt initiation of pulmonary therapy [25].

Ballard et al. [25] estimated that the implementation of the evidence-based bundle protocol prevented an SSI in one of every 16 high-risk pediatric patients who had undergone spinal operation (defined as patients with a diagnosis of cerebral palsy, spina bifida, muscle disease, paralytic deformities, or those with vertebral column resections) and one of every 30 low-risk pediatric patients (defined as patients undergoing elective spinal operation) [25]. Staff compliance with the implementation of each element in the bundle was identified by the authors as the most important factor determinant of bundle optimization and efficacy. Further, collective compliance decreased overall infection rates among all pediatric patients undergoing spinal operation, regardless of diagnosis, from 7.8% to 4.5%, reflecting an overall relative risk reduction of 43% [25].

Researchers from the orthopedic department at the Aristotle University of Thessalonki, Greece, in 2006 developed a bundle aimed to decrease SSIs among spine patients [22]. The research team identified etiologic factors associated with increased risk of SSI, and the bundle aimed to minimize risks based on etiologic factors. The bundle, termed the “Nine P's Protocol,” focused on minimizing risk-factors associated with the following:

(1) Patient factors, (ex. advanced age, malnutrition, smoking, diabetes mellitus, etc.); (2) personnel involved (surgeon, assistants, nurses, residents); (3) place (wards, operating theater, ICU); (4) pre-operative length of stay; (5) procedure; (6) prosthetics; (7) prophylaxis (antibiotic agents) (8) packed red blood cells (number of units transfused); and (9) pus culture results (purulent fluid). With consistent implementation of these nine elements outlined in the protocol, the authors found an overall decrease in infection rates from 16.7% to 3.6%. A significant decrease in post-operative spinal SSI was found with increased compliance of a new bundle protocol (p < 0.05). [22].

Watanabe et al. [8] found that the implementation of a single intervention rather than a bundle composed of numerous interventions effectively reduced rates of post-operative SSI among patients undergoing spinal operation. The authors concluded that the amount of saline incision irrigation used during operation was associated with post-operative SSI rates. The team concluded that saline lavage with a mean volume of irrigation greater than 2,000 mL/h was protective against SSI (OR 0.08; 95% CI 0.01–0.61; p = 0.015). Irrigation with volume of saline greater than 2,000 mL/h decreased the risk of SSI below 10% when compared with incision irrigation in which the surgeon and/or team used less than 1,000 mL/h of physiologic saline [8].

The role of antimicrobial surgical preparations in reducing post-operative SSI

A significant body of research exists highlighting the use of appropriate (and, at times, differing) antimicrobial preparations in minimizing the risk of post-operative infection among surgical spine patients. Three studies included in this review documented a decrease in SSI with pre-operative application of chlorhexidine alcohol.

In a randomized, prospective trial comparing the efficacy between two solutions used for pre-operative skin preparation, chlorhexidine-alcohol scrub was compared with povidone-iodine scrub in reducing post-operative infection rates [26]. There were 849 patients undergoing clean-contaminated procedures included in the study. Chlorhexidine-alcohol solutions were found to be superior in reducing SSIs, thus more protective than povidone-iodine against both superficial incisional infections (4.2% infection rate vs. 8.6% infection rate, p = 0.008) and against deep incisional infections (1% infection rate vs. 3% infection rate, p = 0.05) [26].

A study conducted by Savage and Anderson [3] found that the use of chlorhexidine—in the form of a pre-operative bathing cloth—significantly reduced the observed bacterial count before operation. The team observed a lower positive bacterial culture rate (66%) in the chlorhexidine treatment group compared with the control group (94%, p = 0.008). Ferry et al. [12] found that chlorhexidine scrub offers greater protection against bacterial micro-organisms (and possibly SSIs) because of the prolonged properties of the chlorhexidine in continuous reduction of micro-organisms on the skin post-application in comparison with povidone-iodine. Further, the team documented that the chlorhexidine preparation was associated with less potential for iodine toxicity and skin irritation compared with povidone-iodine [12].

In a study conducted at Washington University in St. Louis [27], the efficacy of three antimicrobial protocols aimed to prevent post-operative SSI were compared following cervical spine procedures. In this study, infection rates were compared across three groups—one group received IV peri-operative antibiotic agents only, the second group received IV antibiotic agents in addition to alcohol foam and surgical drains, and a third group received IV antibiotic agents, alcohol foam and surgical drains, and vancomycin powder applied to the incision. The team determined that surgical drains reduced infection risk from 1.86% to 0.003%. In addition, the team found that the addition of vancomycin powder, 500 mg, added directly to the surgical incision as an adjunct to the use of alcohol foam and surgical drains resulted in zero infections in the 195 patients in which the protocol was implemented [27].

In a 2015 literature review exploring the prophylactic use of intra-site vancomycin powder to prevent SSIs in patients undergoing spinal operation, researchers concluded that only anecdotal evidence exists to support the efficacy of using intra-site vancomycin powder during the surgical procedure to prevent SSIs among patients undergoing spinal operation [28]. The authors of this review concluded that additional studies must be conducted to establish evidence demonstrating the efficacy of this treatment and further studies must be conducted to determine a standard dosage regimen for the use of intra-site vancomycin powder, should it be demonstrated that vancomycin powder is an effective agent in reducing SSIs among spine patients when used intra-operatively [28].

A number of studies have also studied the use of povidone-iodine in the prevention of SSI. In a prospective cohort study by Chang et al. [29], patients undergoing primary instrumented lumbosacral posterolateral fusion for treatment of degenerative spinal disorders were randomly divided into two groups—Group 1 was treated with incision irrigation of 0.35% povidone-iodine solution followed by physiologic saline just before instrumentation, and Group 2 was treated with physiologic incision incision irrigation without the 0.35% povidone-iodine. Among the 107 patients treated with povidone-iodine, no SSIs occurred. By comparison, six infections occurred in the 109 patients treated with physiologic saline alone [29]. This study demonstrated that in comparison with physiologic saline, povidone-iodine is more effective in reducing SSIs in patients undergoing primary instrumented lumbosacral posterolateral spinal fusion for degenerative disease. The study, however, did not describe differences based on levels of fusion.

In another study by Cheng et al. [30], the efficacy of dilute povidone-iodine solution applied to the incision before incision closure in reducing SSIs was investigated. In this study, patients were randomly assigned to two groups. Patients in Group 1 received surgical incision irrigation with dilute 3.5% povidone-iodine solution while the patients in Group 2 served as controls and did not receive 3.5% povidone-iodine incision incision irrigation. The investigators followed patients post-operatively (X = 15.5, range = 6–24 mo). Similar to the results of the Chang et al. [29] study, the intervention (povidone-iodine) patients experienced no incision infections (N = 208) compared with six infections in the control group (N = 206, p = 0.0072) [30]. In both arms of the study by Cheng et al. [30], the average age was in the lower to mid-60s; therefore, it is unknown whether the results would be similar in younger patients undergoing spinal operation.

Savage and Anderson [3] found no clinical evidence that preparation with one antimicrobial agent is more effective in reducing the rate of SSIs among surgical spine patients when compared with others. Citing conflicting results across studies, the authors concluded that no evidence exists to support a difference in antimicrobial growth inhibition when comparing povidone-iodine, chlorhexidine, and 10% povidone-iodine (Betadine^®) solutions [3]. The authors did, however, conclude that incision irrigation with dilute Betadine alone or with vancomycin powder before incision closure may be effective in decreasing the incidence of SSIs. One major limitation of the Betadine study is that the study team used bacterial counts as the outcome measure for the efficacy of the antimicrobial agents and did not use actual SSI rates of patients who had undergone spinal operation. Therefore, it cannot be inferred if one antimicrobial agent was more efficacious in reducing actual SSI rates among the patients in the studies.

Incision irrigation with a combination of povidone-iodine and hydrogen peroxide solution was investigated as a potential solution efficacious in minimizing the risk of SSI. Ulivieri et al. [31] analyzed prospective and retrospective clinical data of patients undergoing spinal operation with this type of incision irrigation and compared infection rates with patients undergoing similar surgical procedures who did not receive this irrigant. In 490 patients treated with this regimen, there were no instances of incision infection, compared with seven deep infections in 460 patients (rate of 1.5%) who did not receive irrigation with this solution [31].

An increase in rate of SSI after orthopedic spinal operation has been found to be associated with nasal colonization with S. aureus. A systematic review of the literature by Savage and Anderson [3] found that a defined pre-screening program for S. aureus colonization followed by eradication with mupirocin ointment and chlorhexidine soap/shower in carriers significantly decreases the risk of a post-operative SSI developing among surgical spine patients (0.19% vs. 0.45%, p = 0.0093) when compared with persons who did not participate in a defined pre-screening, preventative program.

The use of antibiotic prophylaxis is common practice in patients undergoing orthopedic operation. Studies emphasize the appropriate use of antibiotic agents to decrease risk of infection. Ramo et al [32] found that patients receiving inadequate doses of cefazolin (≤20 mg/kg) had a significantly greater risk of infection (18%) compared with those who received evidence-based doses, which are >20 mg/kg (infection rate 6.2%, OR = 3.3, p = 0.0002). The investigators found no difference in infection rates with patients who received gentamicin therapy in conjunction with cefazolin when compared with the cefazolin only treatment group. No significant difference in rates of infection related to timing and post-operative duration of antibiotic agents was found.

Specifically, there was no difference in rate of SSI found among patients who had received only one dose of antibiotic agents within 60 minutes before the incision, those who had also received re-dosing of the antibiotic agent within four hours, or between those who received continued prolongation of antibiotic agents for 48 hours after operation. The aforementioned factors have been recommended practices but were not found to be significant in reducing the rate of SSIs among the orthopedic patients undergoing spine operation in this study [32].

In a retrospective review conducted by Kanayama et al. [33], the investigators compared rates of SSIs in patients receiving a single dose of an antibiotic agent on the day of operation with rates in patients receiving multiple doses for 5–7 days after operation. The authors documented an overall risk of SSI to be 0.7% (N = 1,597) with a risk of 0.8% (N = 1,133) in the group treated with multiple doses of first-generation cephalosporins compared with 0.4% (N = 282) in the single dose group. In addition, resistant strains of bacteria were found in 83.3% of the patients with SSIs in the multiple dose group compared with 0% of patients with SSIs in the single antibiotic dose group [33]. This study was limited by the fact that the authors defined SSI as a post-operative incision infection requiring further surgical intervention, necessarily underestimating SSIs in this population because superficial infections were managed by antibiotic agents and local incision care alone and thus were not considered SSIs by the authors of this study.