The Influence of Perioperative Risk Factors and Therapeutic Interventions on Infection Rates After Spine Surgery

A Systematic Review

Abstract

Study Design. 

Systematic review.

Objective. 

The objectives of this systematic review were to determine the patient and perioperative risk factors that contribute to infections after spine surgery and to examine the level of evidence to support the use of therapeutic interventions to reduce infection rates.

Summary of Background Data. 

Infection continues to be one of the most common and feared complications after spine surgery. As such, it is used as a sentinel event for quality assurance processes. It is clear that the causes of infections after spine surgery are multifactorial and numerous patient- and procedure-related factors have been proposed as contributory elements. In addition, numerous perioperative adjuncts have been suggested to reduce infection rates.

Methods. 

A systematic review of the English-language literature (published between January 1990 and June 2009) was undertaken to identify articles examining risk factors associated with and adjunct treatment measures for preventing surgical-site infections. Two independent reviewers assessed the level of evidence quality using the Grading of Recommendations Assessment, Development, and Evaluation criteria, and disagreements were resolved by consensus.

Results. 

Of the 127 articles identified, 32 met the criteria to undergo full-text review. Individual patient, operative, and perioperative variables have been identified that are associated with increased infection rates (i.e., older age, obesity, diabetes, malnutrition, higher American Society of Anesthesiologists score, posterior approaches, and blood transfusions) but these variables have not been combined to provide individual patient risks based on a composite of factors (e.g., risk stratification). Of the surgical adjuncts investigated, only irrigation with dilute betadine solution showed moderate support for reducing infection rates.

Conclusion. 

It is clear that the causes of postoperative spinal site infections are multifactorial and related to a complex interplay of patient and procedural influences. Because of these complexities, for any individual and surgical procedure, predictable infection rates likely exist that do not extrapolate to 0. Although we have identified factors associated with increased infection rates, further studies will be required to allow multifactorial risk stratification for individual patients and to further investigate the use of therapeutic adjuncts.

In Brief

It is clear that the causes of postoperative spinal infections are multifactorial and related to a complex interplay of patient and procedural influences. Because of these complexities, for any individual patient and surgical procedure, predictable infection rates likely exist that do not extrapolate to 0. Although we have identified factors associated with increased infection rates (older age, obesity, diabetes, malnutrition, higher ASA score, posterior approaches, and blood transfusions), further studies will be required to allow multifactorial risk stratification for individual patients and procedures. In addition, although numerous adjuncts have been proposed to reduce infection rates, additional safety and efficacy information will be required to make further recommendations.

Surgical wound infections continue to be one of the most common complications after spinal surgery. Surgical infections, in general, are a public focus because infection rates are used as a metric to compare hospitals, and insurance agencies have initiated policies to not reimburse expenses related to hospital-acquired infections. However, this is based, in part, on the assumption by monitoring or regulatory agencies and hospital administrators that surgical infections can be completely eliminated. This is not a realistic expectation for several reasons and reflects the common over-reliance on prophylactic antibiotics for preventing surgical infection.

It is widely accepted among leading infectious disease experts that risk factors for surgical infection are multifactorial. Contributory patient-related factors that are often difficult or impossible to modify include age, ASA score, obesity, diabetes, smoking, previous surgery in the area, previous infection, previous radiation, chronic skin conditions such as psoriasis, immunologic competency, nutritional status, and adherence to postoperative wound care instructions. Because these factors cannot be eliminated and can also compound in their effect on the patient through interaction, a zero rate outcome cannot be expected for the risk of hospital-acquired infection. Case-specific factors, such as posterior approach, instrumentation, bone graft harvest, blood loss or blood transfusion, and case length, possibly play an important role as well. Other risk factors to consider include possible breaches in sterile technique through prepping, draping, room traffic, contaminated instruments, and use of intraoperative fluoroscopy. Additionally, changes in the prevalence of more virulent infectious agents (e.g., methicillin-resistant Staphylococcus aureus, gram-negative rods) may also contribute. Our goal should be to optimize our perioperative approach to maintain infection rates at an acceptably low level. However, the complex nature of postoperative infections and individual variability make a risk calculation for any individual person very difficult, especially in the presence of multiple comorbidities. As we will discuss later, many operative and perioperative adjunct therapies have been proposed to help reduce infection rates. Although many have shown promise, the available studies lack sufficient evidence-based power to make strong recommendations, and extrapolation of data from other surgical subspecialties (such as laminar flow for total joints) may not be valid.

Watters et al¹ published an evidence-based clinical guideline evaluating the prophylactic use of antibiotics in spine surgery. This report was developed by the Antibiotic Prophylaxis Work Group of the North American Spine Society. The systematic review answered 11 questions related to prophylactic antibiotic therapy ranging from the efficacy of antibiotic prophylaxis with and without spine implants, recommended drugs, dosages, and time of administration, discontinuation, wound drains, differences by body habitus, and comorbidities. The guideline concluded that patients undergoing spine surgery should receive preoperative prophylactic antibiotics with a relatively strong recommendation.¹ The superiority of one agent or schedule over another could not be recommended. Wound drains were not recommended as a means to reduce infection. Obese patients were found to be at higher risk, and specific changes to antibiotic protocols based on comorbidities could not be recommended because of limited evidence. Given the work done on evaluating prophylactic antibiotics, the following questions determined the focus of this systematic review:

1. What preoperative and intraoperative risk factors increase postoperative surgical-site infection rates after major spine surgery?
2. Is there evidence to support the use of specific surgical adjuncts (in addition to perioperative antibiotics) that decrease postoperative surgical-site infections after spine surgery?

Materials and Methods

Electronic Literature Database

The literature search is outlined in detail elsewhere.1a Briefly, a systematic search was conducted in Medline, EMBASE, and the Cochrane Collaboration Library for literature published from January 1990 through December 2008. We limited our results to humans and to articles published in the English language. Reference lists of key articles were also systematically checked. For our first question, we identified all articles evaluating preoperative and intraoperative risk factors for early (<3 months from discharge) surgical-site infection occurring as a result of spine surgery. For our second question, we identified all articles that evaluated the efficacy of preventative measures other than antibiotics for preventing postoperative surgical-site infection after major spine surgery. Articles were excluded if they were pediatric studies (age <18 years), tumor surgery, revision surgery, treatment for osteomyelitis, other than spine surgery, or evaluating antibiotics. Other exclusions included reviews, editorials, studies with <10 subjects, case reports, non-English-written studies, and animal studies (Figure 1). Question no. 1 was evaluated as a prognostic question and question no, 2 as a therapeutic question.

Figure 1: Inclusion and exclusion criteria.

Data Extraction

Each retrieved citation was reviewed by 2 independently working reviewers (D.C.N. and J.R.D.). Most articles were excluded on the basis of information provided by the title or abstract. Citations that seemed to be appropriate or those that could not be excluded unequivocally from the title and abstract were identified, and the corresponding full-text reports were reviewed by the 2 reviewers. Any disagreement between them was resolved by consensus. From the included articles, the following data were extracted: patient demographics, diagnosis, spine surgical intervention, risk factors evaluated, wound care measures, infection rates, and results.

Study Quality

Level of evidence ratings were assigned to each article independently by 2 reviewers using criteria set by The Journal of Bone and Joint Surgery, American Volume (J Bone Joint Surg Am)² for prognostic studies and therapeutic studies and modified to delineate criteria associated with methodologic quality and described elsewhere (Supplemental Digital Content, individual study ratings, available at: https://links.lww.com/BRS/A422; Tables 1 and 2).

Table 1: Summary of Common Preoperative Risk Factors Associated With Postoperative Surgical-Site Infection

Table 1: Continued

Table 2: Summary of Common Intraoperative Risk Factors Associated With Postoperative Surgical-Site Infection

Analysis

Postoperative surgical-site infection rates were reported as the proportion of patients experiencing an event within 3 months of surgery for prognostic studies evaluating risk factors. We accepted longer follow-up periods for the therapeutic studies evaluating preventative measures. Odds ratios (ORs) were reported if reported by the individual authors. If they were not reported, unadjusted ORs with corresponding confidence intervals and P values were calculated using the author's raw data where possible using Stata 9.0 (Stata Corp., College Station, TX).³ Because the rates of postoperative surgical-site infection were low (<10%), the ORs approximated the relative risk. Data were not pooled and summarized in a meta-analysis because of significant study heterogeneity and the inability to control for confounding factors. Data were summarized in tables, and qualitative analysis⁴ was performed considering the following 3 domains: quality of studies (level of evidence), quantity of studies (the number of published studies similar in patient population, condition treated, and outcome assessed), and consistency of results across studies (whether the results of the different studies lead to a similar conclusion).⁵ We judged whether the body of literature represented a minimum standard for each of the 3 domains using the following criteria: for study quality, at least 80% of the studies reported needed to be rated as a level of evidence I or II; for study quantity, at least 3 published studies were needed, which were adequately powered to answer the study question; for study consistency, at least 70% of the studies had to have consistent results. The overall strength of the body of literature was expressed in terms of the impact that further research may have on the results. An overall strength of “high” means that further research is very unlikely to change our confidence in the estimate of effect. The overall strength of “moderate” is interpreted as further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. A grade of “low” means that further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate, while “very low” means that any estimate of effect is very uncertain.1a

Results

Study Selection

We identified 127 articles or reports from our literature search reporting on risk factors for surgical-site infection or adjunct therapy for preventing surgical-site infection in spine surgery. From these potential articles or reports, we judged 34 to undergo full-text review. After full-text review, we excluded 15 for the following reasons: surgical-site infections were not the primary outcome of analysis (n = 4), spine patients were not analyzed or analyzed separately from other orthopedic patients (n = 4), risk factors were not analyzed or raw data were not provided to calculate risks (n = 6), and pediatric patients (n = 1; Figure 2). Tables 1 and 2 report ORs for preoperative and intraoperative risk factors, respectively (Supplemental Digital Content, data from the individual studies, available at: https://links.lww.com/BRS/A422; Table 3). Tables 3 to 5 summarize individual studies evaluating adjunct perioperative measures for preventing surgical-site infection. Surgical-site infection rates ranged from <1%⁶ to 10.9%.⁷ Thirteen of the 19 studies reviewed reported a surgical-site infection rate of <5%. The remaining 7 were between 5% and 11% (Supplemental Digital Content, available at: https://links.lww.com/BRS/A422; Tables 3 and 4).

Figure 2: Flow chart showing results of infection literature search.

Table 3: Iodine Solution as a Preventative Method for Postoperative Surgical-Site Infection

Table 4: Wound Drainage as a Preventative Method for Postoperative Surgical-Site Infection

Table 5: Dressings and Other Methods for Preventing Postoperative Surgical-Site Infection

What Preoperative and Intraoperative Risk Factors Increase Perioperative Infection Rates After Major Spine Surgery?

A wide range of risk factors for surgical-site infection in spine surgery have been evaluated and reported. We summarized preoperative and intraoperative risk factors that were reported in 3 or more studies from our study selection in Tables 1 and 2, respectively. Unless specified, statistically significant associations were the result of multivariate analyses controlling for other potential confounding variables.

Preoperative Risk Factors

Age.

Seven case-control studies evaluated age as a preoperative risk factor for postoperative surgical-site infection.^6,8–13 Older age was a risk factor for postoperative surgical-site infection in the majority of the studies. In the only study that reported an OR as an effect measure for the association between age and postoperative infection, patients aged >60 years were nearly 3 times more likely to develop a postoperative surgical-site infection compared with those 60 years and younger (OR = 2.7; 95% confidence interval [CI], 1.2–6.3; P = 0.03).⁹ In 3 additional studies,^8,10,12 the mean ages of the cases were appreciably greater than the controls in 4 studies (P = 0.06, 0.001, 0.04, respectively), with cases having a mean age near or above 60 years of age and controls with a mean age <60 years. In the remaining 3 studies, mean ages were similar between cases and controls.^6,11,13

Sex.

Six case-control studies evaluated sex as a preoperative risk factor for postoperative surgical-site infection.^6,9–13 Risk varied widely by gender, and no study found a statistically significant association between sex and postoperative surgical-site infection.

Steroid Therapy.

Four case-control studies evaluated steroid therapy as a preoperative risk factor for postoperative surgical-site infection.^8,9,11,13 Only 1 study found a statistically significant association between steroid therapy and postoperative surgical-site infection.¹¹ Thirty-one percent of cases and 9% of controls had undergone steroid therapy (OR = 4.9, 95% CI, 1.1–29.6; P = 0.017).

Diabetes.

Seven case-control studies and 1 retrospective cohort study evaluated diabetes as a preoperative risk factor for postoperative surgical-site infection.^6,8–12,14 Five of these studies reported a statistically significant association between diabetes and postoperative surgical-site infection (P = 0.005, 0.04, 0.02, <0.001, 0.02, respectively).^6,9,10,14,15 Corresponding ORs were infinite (95% CI, 1.7 to infinite), 4.2 (95% CI, 1.1–16.3), 4.0 (95% CI, 1.2–12.8), 6.3 (95% CI, 3.3–12.1), and 3.5 (95% CI, 1.2–10.0).

Glucose Level.

Five case-control studies evaluated blood glucose levels perioperatively as a risk factor for postoperative surgical-site infection.^6,8,12–14 High glucose levels were defined by different cutoff levels ranging from >126 to >200 mg/dL or top 25th percentile of all patients. Four of 5 studies reported a statistically significant association between high blood glucose levels and postoperative surgical-site infection. ORs ranged from 3.0 to 3.3. One study that did not find a statistically significant association reported 15% of cases and 2% of controls had a blood glucose level >200 mg/dL (OR = 8.4; 95% CI, 0.38–503.7; P = 0.05).

Malnutrition.

Two case-control^8,11 and 2 retrospective cohort^16,17 studies evaluated malnutrition as a preoperative risk factor for postoperative surgical-site infection. All 4 studies used albumin as a measure of malnourishment (<3 mg/dL was a common indicator). Two also considered a total lymphocyte count of 1500 to 2000 cells/mm³ as malnourishment. Two of the 4 studies found a statistically significant association between malnourishment before surgery and a postoperative surgical-site infection.^16,17 Although the association was not significant, another study found 23% of cases and 11% of controls (OR = 2.5, 95% CI, 0.33–15.4; P = 0.29) to be malnourished.⁸ The study by Klein et al reported an adjusted OR of 15.6 (95% CI, 6.5–37.4; P = 0.002).

Smoking.

Five case-control studies evaluated smoking as a preoperative risk factor for postoperative surgical-site infection.^8,9,11–13 Two studies reported a statistically significant or near significant association between smoking and postoperative surgical-site infection; however, Fang et al⁹ reported smoking as a risk factor (OR = 2.26, 95% CI, 0.98–5.3; P = 0.06) and Olsen et al¹³ found that smokers were at decreased risk of postoperative surgical-site infection (OR = 0.4; 95% CI, 0.2–0.8; P = 0.006). The study by Olsen et al reported univariate analyses (i.e., did not control for other risk factors through multivariate analysis). The other 3 studies did not find significant associations; in 1 study there were more cases who were smokers (49% vs. 29%)¹¹ and in the other 2 there were more smokers in the control groups.^8,12

Body Mass Index/Obesity.

Six case-control studies evaluated body mass index (BMI)/obesity as a preoperative risk factor for postoperative surgical-site infection.^{6,8,9,12–14} Obesity was defined by different cutoff levels for BMI ranging from >25.5 to >35 kg/m.² Five of the 6 studies reported a statistically significant association between obesity and postoperative surgical-site infection. ORs ranged from 2.2 to 7.1. One study that did not find a statistically significant association (P = 0.26) reported 69% of cases and 47% of controls had a BMI > 27.3 (OR = 2.6; 95% CI, 0.6–12.8; P = 0.15).

ASA Score.

Five case-control studies evaluated the preoperative ASA score as a risk factor for postoperative surgical-site infection.^6,8,12–14 A high ASA score was defined as a score of 3 or greater. Four of the 5 studies reported a statistically significant association between high ASA scores and postoperative surgical-site infection.^8,12–14 ORs ranged from 2.6 to 9.7. One study that did not find a statistically significant association (P = 0.28) reported a median preoperative score of 2 for both cases and controls.⁶

Previous Spine Surgery.

Five case-control studies evaluated previous spine surgery as a risk factor for postoperative surgical-site infection.^6,8–10,12 Only 1 study found a statistically significant association between previous spine surgery and postoperative surgical-site infection (OR = 2.8, 95% CI, 1.6–5.0; P < 0.001); however, this was a univariate association and did not control for other potential confounding variables.¹²

Intraoperative Risk Factors

Allograft Use.

Three case-control studies evaluated allograft use as a risk factor for postoperative surgical-site infection, none of which reported a statistically significant association.^8,9,11 In 2 of the 3 studies, a greater percentage of controls received an allograft.^8,9 Apisarnthanarak et al⁸ (N = 60) reported 54% of cases and 79% of controls received an allograft for laminectomy with spinal fusion demonstrating a decreased risk of postoperative surgical-site infection in those with an allograft (OR = 0.31, 95% CI, 0.08–1.15; P = 0.08); however, this association did not reach statistical significance probably because of the relatively small sample size. Klekamp et al¹¹ reported those receiving an allograft were at an increased risk of a postoperative surgical-site infection (OR = 3.4, 95% CI, 0.55–36.5; P = 0.13); however, this was a univariate association and was not statistically significant.

Instrumentation.

Six case-control studies evaluated the use of instrumentation as a risk factor for postoperative surgical-site infection.^6,9–13 Only 2 of the 6 studies found a statistically significant association between instrumentation and postoperative surgical-site infection (OR = 3.4, 95% CI, 1.3–9.3; P = 0.012 and OR = 2.5, 95% CI, 1.1–6.0; P = 0.03, respectively).^10,12 Three of the other 4 studies that did not report a statistically significant association did report a greater percentage of case patients receiving instrumentation than controls (77% vs. 69%, 74% vs. 57%, and 95% vs. 80%, respectively)^9,11,13; therefore, we cannot rule out a possible association. The metallic composition of hardware used was not reported.

Duration of Surgery.

Seven case-control studies evaluated duration of surgery as a risk factor for postoperative surgical-site infection.^6,8–13 Only 2 case-control studies reported a statistically significant association between duration of surgery (long duration defined by >75th percentile of all subjects) and postoperative surgical-site infection (OR = 4.7, 95% CI, 1.6–14.0; P < 0.001 and OR = 2.4, 95% CI, 1.2–4.6; P = 0.012, univariate), respectively.^12,14 The article by Apisarnthanarak et al⁸ reported an OR of 3.4 (0.95–12.1; P = 0.08) for a duration of ≥3 hours. The case-control studies by Fang et al, Friedman et al, and Klekamp et al also reported a greater duration of surgery among the risk factors compared to the controls, but these differences were not statistically significant.^6,9,11 However, 6 of 7 studies demonstrated a greater rate of infection in longer cases than in controls; therefore, we cannot rule out a possible association.

Transfusion.

Three case-control studies evaluated intraoperative transfusion as a risk factor for postoperative surgical-site infection.^8,12,13 Two of the case-control studies reported statistically significant associations between transfusion and postoperative surgical-site infection (OR = 6.7, 95% CI, 3.6–13.0; P < 0.001 and OR = 6.3, 95% CI, 2.5–15.6; P < 0.006, respectively)^12,13; however, both of these were the result of a univariate analysis not controlling for other potential confounding variables.

Posterior Approach.

Three case-control studies evaluated a posterior approach as a risk factor for postoperative surgical-site infection, with all 3 reporting a statistically significant association between posterior approach and postoperative surgical-site infection.^12–14 ORs were 3.5 (95% CI, 1.2–9.7; P = 0.02), 8.2 (95% CI, 2.0–33.5; P = 0.003), and 3.4 (95% CI, 1.2–9.8; P = 0.02), respectively.

In summary, with respect to preoperative risk factors, age (>60 years), presence of diabetes, malnutrition, obesity, ASA score ≥3, and higher glucose levels were positively associated with surgical-site infections. Sex, steroid therapy, smoking, and previous spine surgery were not found to consistently be associated with surgical-site infection. With respect to intraoperative risk factors, transfusions and a posterior approach were consistently positively associated with postoperative surgical-site infection. Allograft use, use of instrumentation, and duration of surgery were not consistently significantly associated with infection; however, duration of surgery and instrumentation were positively associated in most studies, despite some not reaching statistical significance.

Is There Evidence to Support Adjunct Intraoperative Wound Care Measures to Decrease Postoperative Surgical-Site Infection After Major Spine Surgery?

Eight studies were identified evaluating wound care measures other than antibiotics as methods for preventing postoperative surgical-site infection in spine surgery (Tables 3–5). Two were randomized controlled trials comparing close wound suction drainage to no wound suction drainage^18,19; 2 were randomized controlled trials comparing wounds irrigated with povidone-iodine solution to wounds treated with saline solution^20,21; 1 was a retrospective cohort study comparing patients who received silver-impregnated dressings to patients who received iodine or alcohol-based swab and dry gauze dressings⁷; 1 was a retrospective cohort study comparing ultraclean air technology (UCAT) using laminar flow to conventional operating room methods²²; 1 was a randomized controlled trial comparing shaving versus no shaving of the surgical site²³; 1 was a retrospective cohort study comparing a new operating room protocol to standard of care.²⁴

The randomized trials by Brown and Brookfield¹⁸ and Payne et al¹⁹ compared closed wound suction drainage with no suction drainage in patients treated with or without instrumentations for a myriad of conditions, including herniated disc, stenosis, spondylolisthesis, and postlaminectomy syndrome (N = 83 treatment group; N = 41 control group) and patients with laminectomy for herniated disc or stenosis (N = 200 treatment group; N = 97 control group), respectively. A 0% and 1.9% infection rate was reported in the treatment group, respectively, and a 0% and 1% infection rate was reported in the control group, respectively. These differences were not statistically significant.

The randomized trials by Chang et al²⁰ and Cheng et al²¹ compared wounds irrigated with 0.35% povidone-iodine solution (3.5% betadine) to wounds irrigated with saline solution in patients treated for degenerative spinal disorders with segmental instability (N = 244 treatment group; N = 124 control group) and patients with a number of diverse of spine conditions, including stenosis, scoliosis, trauma, and spinal metastatic lesions (N = 414 treatment group; N = 206 control group), respectively. The authors did not report whether the betadine used was sterile or nonsterile. Chang et al²⁰ does not comment specifically on the preparation method. Cheng et al²¹ reports diluting 5 mL of commercially available 10% povidone-iodine solution to a final concentration of 0.35%. Both studies reported a 0% infection rate in the betadine group. In the no-solution groups, the study by Chang et al reported an overall infection rate of 4.8% (P = 0.029) and Cheng et al reported a 0.5% and 2.9% superficial and deep infection rate (P = 0.015), respectively.

In another randomized trial, Celik and Kara²³ compared shaving (N = 371) versus no shaving (N = 418) of the spine surgical site. Patients were being surgically treated for disc herniation, stenosis, lateral recess syndrome, and spinal tumors. Patients who were shaved experience a 1.07% (n = 4) infection rate, and those who were not shaved a 0.23% (n = 1) infection rate (P < 0.01).

In a retrospective cohort study of patients undergoing lumbar laminectomy with instrumented fusion, Epstein⁷ compared patients treated with silver-impregnated dressings to patients with iodine or alcohol-based swab and dry gauze dressings. The overall infection rate in those treated with the silver-impregnated dressings was 0%, and 10.9% (deep infection = 2.3%; superficial infection = 8.6%) in those treated with the dry gauze (no analytical statistics performed comparing groups).

In a retrospective cohort study of patients undergoing posterior spinal fusion with instrumentation for various spinal conditions including tumor, trauma, deformity, and degenerative conditions, Gruenberg et al²² compared UCAT (laminar flow) to conventional operating room conditions. The potential for selection bias through selective use of operating room protocols, standard versus laminar flow and by use of standard surgical gowns versus total body exhaust suits should be considered in this evaluation. The infection rate in the conventional group was 10.1% and 0% in the UCAT group (P < 0.017). There were no significant differences at baseline between groups with respect to age, diabetes, steroid therapy, previous surgeries, tumor cases, or previous radiotherapy. The number of levels (P < 0.001), utilization of allograft (P > 0.001), and operative time (P < 0.001) were significantly higher in the UCAT group.

Christodoulou et al²⁴ used a before/after design (6-month period each) to evaluate the “Nine Ps Protocol” in patients with various spine surgeries for scoliosis, disc herniation, stenosis, spondylolisthesis, tumor, trauma, and implant removal. The Nine Ps consisted of patient-related factors, personnel, place, preoperative length of stay, procedure, prosthetics, prophylaxis, packed red blood cells, and pus cultures. This was implemented after a 6-month period “outbreak” where the infection rate was 16.7%. After 6 months of implementing the protocol, the infection rate dropped to 3.8% (P = 0.032). Before the outbreak the authors reported an infection rate of 0%. There were more fusions during the outbreak period than the other 2 periods (P = 0.034).

Evidence Summary

The overall strength of the evidence defining preoperative and intraoperative risk factors for postoperative surgical-site infection is “moderate,” that is, further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate (Table 6). There is little research published on adjunct measures for preventing postoperative surgical-site infection in spine surgery. Most of this work has been done evaluating the safety and efficacy of antibiotics; however, 2 randomized controlled trials demonstrated that Betadine was superior to saline solution providing “moderate” evidence to support its use. There is “low” evidence, that is, further research is very likely to have an important impact of our confidence in the estimate of effect and is likely to change the estimate, indicating little to no difference in postoperative surgical-site infection in those treated with and without closed suction wound drainage, “low” evidence to support that no shaving is safer than shaving, “very low” evidence, that is, any estimate of effect is very uncertain, to support the relative safety of silver-impregnated dressings, “very low” evidence to support the relative safety of UCAT, and “very low” evidence to support the implementation of the Nine Ps protocol.

Table 6: Rating of Overall Strength of Evidence for Each Key Question

Discussion

The objectives of this systematic review were to determine the patient and perioperative risk factors that contribute to infections after spine surgery and to examine the level of evidence to support the use of therapeutic interventions to reduce infection rates (in addition to perioperative antibiotics).

We reported ORs for categorical variables and P values for continuous variables evaluating the association between preoperative and intraoperative risk factors for surgical-site infection, if the evaluation of such factors appeared in 3 or more studies. For preoperative risk factors, age (>60 years), presence of diabetes, malnutrition, obesity, ASA score ≥3, and higher glucose levels were all positively associated with surgical-site infection. Sex, steroid therapy, smoking, and previous spine surgery were not found to consistently be risk factors for infection. This does not suggest that they are not as some studies found them associated; however, we cannot make this claim based on the methods and findings from this systematic review. For intraoperative risk factors, transfusions and a posterior approach were consistently significant risk factors for postoperative surgical-site infection. Allograft use, use of instrumentation, and duration of surgery were not consistently associated with infection; however, the majority of the studies evaluating duration of surgery demonstrated a greater rate of infection in cases than in controls; therefore, we cannot rule out a possible association.

The overall strength of the evidence to support these findings is moderate because the majority of studies evaluating these risk factors were case-control studies. Case-control studies are automatically judged as level III evidence in the JBJS criteria; however, for studies identifying risk factors for rare events, it is the most efficient design. Further, these studies met many of the other important methodologic criteria. We identified a high quantity of articles and findings were generally consistent.

Because of study heterogeneity, it was not possible to pool data between studies and conduct a meta-analysis to arrive at single effect estimates for each risk factor. Nonetheless, the best evidence synthesis method of qualitatively evaluating individual study effect estimates is also a useful method for arriving at clinical recommendations if the findings are consistent among several studies.

There is little research published on adjunct measures for preventing postoperative surgical-site infection. Most of this work has been done evaluating the safety and efficacy of antibiotics.¹ The adjunct measure with the highest level of evidence (moderate) supporting its efficacy is diluted betadine. The evidence is moderate because only 2 studies were identified evaluating its efficacy though both were of high quality and were consistent. Both studies reported a 0% infection rate in the betadine group and 4.8% and 3.5% in the saline solution group, respectively.^20,21 However, the exact preparation methods for the betadine solutions including pharmacy involvement to ensure accurate and consistent dilutions are not discussed in these studies. Other antibiotic solutions have not been evaluated to be able to recommend other measures though some may exist. There is low evidence, indicating little to no difference in postoperative surgical-site infection in those treated with and without closed suction wound drainage, low evidence to support the relative safety of no shaving compared with shaving, though rates were lower in patients who did not get shaved, very low evidence to support the relative safety of silver-impregnated dressings, and very low evidence to support the relative safety of UCAT. In looking outside the spine literature, we identified other measures that have been evaluated with respect to their ability to prevent surgical-site infection. Many of these studies do not evaluate surgical-site infection but other surrogate outcomes such as bacterial growth, using various culture plating techniques. The following measures have been evaluated in the general orthopedic literature with respect to their association with surgical-site infections: ultraclean air,^25–29 occlusive garments,²⁵ double gloving,^30,31 betadine,^32,33 body exhaust suits,²⁹ ultraviolet lighting,³⁴ antiseptic scrubs,^35,36 and nasal swabs.^37,38 Such techniques as oxygenation or hyperoxygenation were not identified. Table 7 lists the various adjunct preventative measures reported in the spine and general orthopedic surgery literature. A systematic review of these studies and others outside of spine surgery is beyond the scope of this project. Such a review for the purpose of generalizing to spine surgery should be considered with caution regardless of the level of evidence given the complexity of spine surgery compared with general orthopedic surgery such as joint replacement. Further research is needed to explore the efficacy of these adjunct methods in spine surgery.

Table 7: List of Adjunct Measures for Preventing Postoperative Surgical-Site Infection Published in the Literature

Risk stratification methods have been developed in other surgical disciplines such as total joint surgery. However, this would be difficult with the available spine literature because the studies are heterogeneous and cannot be adapted well to a meta-analysis. Further, individual authors who have reported ORs for specific risk factors, have not attempted to combine these risk factors through such methods as prediction modeling or risk stratification. However, this is certainly an area for future research because it may allow attempts at defining “acceptable” infection risks and may reveal modifiable factors such as adjustment in antibiotic therapy and perioperative adjuncts to help reduce infection rates.

From this systematic review, it seems that older patients, with obesity and diabetes, who are sicker (i.e., higher ASA scores) are at a higher risk of a surgical-site infection. What we do not know is how each of these potential risk factors, when encountered in combination, adds to the overall risk of surgical-site infection. Authors did not report any type of risk stratification based on single or multiple risk factors. For example, it seems that patients with diabetes are at least twice as likely (and as high as 6 times) to develop a surgical-site infection. What we do not know is what the additional risk is if they are also obese and over the age of 60. In addition, we do not know the added risk of these factors coupled with a posterior approach. The ideal clinical tool would allow for risk stratification whereby the surgeon could calculate a risk of surgical-site infection based on a composite of risk factors before surgery that may be modified by intraoperative interventions such as the surgical approach, duration of surgery, or need for a blood transfusion. This would assist in surgical decision-making with respect to the most appropriate management. In a complex clinical environment, spine surgeons are often challenged in making treatment decisions because of the lack of prognostic information, and resources, pertaining to the severity of spine diseases or conditions. The lack of consideration of disease severity in clinical decision-making thus far, requires spine surgeons to make treatments decisions without objective disease severity measure guidelines. For example, patients with greater severity (i.e., more positive risk factors) should not be expected to have the same outcomes as those with a lesser disease severity and presumably a different treatment pathway is indicated. This lack of objective clinical guidelines combined with a dearth of information examining the severity of spine disease or deformity on treatment outcomes puts spine surgeons in a challenging position of trying to make optimal clinical decisions without much needed information relating disease severity to potential outcomes. Thus, addressing this gap in knowledge is paramount for ensuring that spine surgeons have the necessary information and resources to enable them to make the most-informed clinical decisions and provide the highest patient care. We recommend that further research be performed to develop a composite disease severity index that assists in clinical decision-making and patient prognosis. Such a system could also be tied to other clinical outcomes and patient health-related quality of life.

Clinical Recommendations

It is imperative to recognize that the causes of perioperative spinal infections are multifactorial, including patient- and procedure-specific risks. Even though we cannot provide true risk stratification, it is essential to counsel patients and document comorbidities and risk factors that may increase infection rates. Because some of the factors are not easily modified, as a general rule, strict adherence to maintenance of sterile technique and basic surgical principles for tissue handling is paramount. Currently, administration of preoperative antibiotics without specific recommendations for drug, redosing, and discontinuation, is the strongest supported recommendation for reducing perioperative spinal infections. Although there are numerous adjuncts that have shown promise in reducing infection, the available studies do not provide sufficient evidence to allow specific recommendations without additional safety and efficacy information.

Key Points

• Because the causes of perioperative spinal infections are multifactorial, for any individual patient and specific surgical procedure, predictable infection rates likely exist that do not extrapolate to 0.

• Individual patient, operative, and perioperative variables have been identified that are associated with increased infection rates (i.e., older age, obesity, diabetes, malnutrition, higher American Society of Anesthesiologists score, posterior approaches, and blood transfusions) but these variables cannot currently be combined to provide individual patient risks based on a composite of factors (e.g., risk stratification).
• We recommend further research and prospective data collection to develop a risk stratification system that takes into account multiple risk factors to allow calculation of expected infection rates based on comorbidity or surgical procedure profiles. Establishment of such a severity system would then allow a more systematic evaluation of measures to reduce infection, including perioperative adjuncts.
• There is currently insufficient evidence to strongly recommend specific adjuncts to reduce infection other than administration of preoperative antibiotics. However, in our review of adjunct measures, irrigation with dilute betadine solution provided moderate support for infection reduction.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.spinejournal.com).

Acknowledgments

The authors thank Ms. Nancy Holmes, RN, for her administrative assistance and Mr. Jeff Hermsmeyer, BS, for his assistance in searching the literature, abstracting data, and proofing.