Topical Vancomycin for Prevention of Surgical Site Infection in Cranial Surgeries: Results of an Updated Systematic Review, Meta-Analysis and Meta-Regression
Keywords: Meta-analysis, surgical site infection, systematic review, vancomycin
The World Health Organization (WHO) has recommended several steps to prevent surgical site infections (SSI) after cranial surgery, which include patient preparation before surgery, handwashing, and surgical site preparation. Various strategies are applied in cranial surgeries to minimize the same including preop chlorhexidine gluconate (CHG) head wash, preop CHG-alcohol skin preparation, perioperative antibiotic prophylaxis, and postoperative incision care. Despite these, the rate of SSI continues to be still high (1–9%) and it has become the most preventable healthcare-associated infection. A few studies have advised the application of systemic antibiotics for the prevention of SSI, especially in cases of shunts or drain, while other studies disagree stating it does not have an effect in reducing SSIs but also may be harmful as it enhances the possibility of growth of resistant microbes and clostridium difficile. In this respect, many antibiotics like bacitracin, neomycin/polymyxin, rifampin, gentamicin, and vancomycin have been used topically to prevent SSI.
Topical vancomycin has emerged as an effective approach and several studies have shown that it has good efficacy in preventing SSI following spinal and orthopedic surgical procedures., This is due to its broad spectrum of action which covers almost all gram-positive cocci (Methicillin resistant staphylococcus aureus (MRSA), Methicillin sensitive staphylococcus aureus (MSSA), and Streptococci) which are the usual causes for cranial wound infection. Also, it has limited systemic absorption, thus, maintaining good local wound concentration without the risk of systemic toxicity. Many studies have proven the safety of topical vancomycin use in pediatric as well as in other age groups.,,,,, Despite these factors, there has been no definite guideline on the use of topical vancomycin in preventing SSIs. Thus, we sought to comprehensively summarize the published literature on the role of topical vancomycin application on SSIs in non-spinal neurosurgical procedures. A recent meta-analysis also summarized the role of topical vancomycin in neurosurgical procedures. However, we have included four new studies with 813 additional patients and expect to have more statistical power to get more conclusive results. Moreover, we have done meta-regression to study the effect of the number of patients in the studies and duration of follow-up on the rate of SSI.
The PRISMA checklist was followed for our meta-analysis and systematic review. A systematic search of the MEDLINE database (1966 to July 2020), EMBASE (1975 to July 2020), and Web of Science (1900 to July 2020) was performed with the keywords: 'vancomycin and craniotomy,' 'vancomycin and cranioplasty,' 'vancomycin and Deep Brain Stimulation (DBS),' 'vancomycin and VP shunt,' 'vancomycin and neurosurgery,' 'topical vancomycin,' 'vancomycin and neurosurgery,' 'vancomycin and cranial surgery,' on July 1, 2020. All the results were screened by two authors (HD and KG) and the assessment was done independently. Any disagreements were resolved through discussion. The abstract of each citation was then screened for relevance to our review. Cross-references of the studies were also screened. Of all the articles identified by searching the database, a preliminary screening of abstracts was done. Full-text articles were extracted and a secondary screening was done by the authors. After independently reviewing the final set of articles, the following inclusion and exclusion criteria were applied.
All studies which compared the rate of SSI in cranial neurosurgical procedures with and without topical vancomycin were included. Studies in which topical vancomycin was used for cranial procedures not involving a craniotomy such as shunts, deep brain stimulation, and cranioplasty were also included. Both retrospective studies and prospective studies were included.
Those studies were excluded where vancomycin was not administered topically. Studies wherein vancomycin was only administered into the shunt system or the intrathecal space, without irrigation of the surgical site were also excluded. Our study was concerned with evaluating the role of topical vancomycin administered over the incision site before closure in preventing SSIs.
Data extraction and quality assessment
Data regarding the characteristics of the study (such as publication year; country; study size, design, duration; Newcastle-Ottawa score), and type of surgery, use of implants were collected using a predesigned data extraction tool in Microsoft Excel 2016 (Microsoft Corp, Redmond, WA). We noted the dose of vancomycin, the technique of administration in each study; data regarding the outcomes: rate of SSI and interval between surgery and SSI; possible complications related to antibiotic use (i.e., rash, anaphylaxis). Data were recorded independently by two authors (HD and KG) from the included studies. For quality assessment, unblinded reviewers used the Newcastle-Ottawa scale for cohort studies [Table 2].
The statistical analysis of the pooled data was performed using the R software (R Foundation for Statistical Computing, Vienna, Austria) employing the 'meta,' 'metafor,' and 'dmetar' package. The results are reported as Mantel-Haenszel odds ratios (MH OR) along with their 95% CI for dichotomous data. The heterogeneity of the studies was evaluated by I2. The values of I2 <40% were considered nonsignificant, 40–60% were considered to represent moderate heterogeneity, and values higher than 60% were reported as high heterogeneity. The random-effects model was used for data analysis when heterogeneity was more than 40% while a fixed-effects model was used if heterogeneity was less than 40%. A P value of <0.05 was considered to be statistically significant.
Subgroup analysis was done to study the difference between the infection rates following the use of topical vancomycin in different types of surgical procedures. Meta-regression was done to assess the effect of the number of cases, duration of follow-up, and the year of publication of the study using the 'metafor' package in R.
The initial database query identified 2,066 citations (744 from EMBASE, 882 from PubMed/MEDLINE, and 440 from the Web of Science) [Figure 1]; 44 articles were selected for full-text review after the primary screening. After independently reviewing the selected articles, we were left with 14 articles that were chosen for the systematic review and meta-analyses. No further eligible studies were identified from the reference lists of the 14 full-text articles. Twelve articles had quantitative data and were included in the meta-analysis while the other two articles could not be included in the systematic review.,,,,,,,,,,,,, [Table 1] summarizes the various studies included in this review.
A total of 3,446 patients were included in the meta-analysis. SSI developed in 1.6% (29 out of 1,817) of the patients in the vancomycin group as compared to 5.28% (86 out of 1,629) in the control group. The pooled risk ratio was 0.24 with 95% CI: 0.12–0.51 (P-value: <0.00001) [Figure 2]. The difference between the subgroups was significant (P-value: <0.00001). The number needed to treat (NNT) was 27.2. The studies showed low heterogeneity with an I2 of 24%. The Egger's test revealed with X-axis intercept at ̶1.289 with P value being 0.007, suggesting the presence of publication bias [Figure 3].
Subgroup analysis was done after dividing the studies based on the type of surgical procedure or age group involved. There were three studies in the Ventriculoperitoneal shunt (VP) shunt subgroup involving 900 patients. The pooled risk ratio of infection in the vancomycin group was 0.27 (95% CI 0.03–2.8, P = 0.002) with low study heterogeneity with an I2 of 23%. This suggests a significant decrease in the risk ratio of developing SSI in the vancomycin group. The NNT was 19. There were three studies in the DBS/ Implantable pulse generators (IPG) subgroup with a total of 835 patients. The pooled risk ratio of infection in the vancomycin group was 0.24 (with 95% CI 0.09–0.60, P = 0.002) with no study heterogeneity with an I2 of 0%. The NNT was 18.9. There were three studies in the adult craniotomy subgroup with a total of 855 patients. The various indications of craniotomy were tumors and vascular cases. The pooled risk ratio of infection in the vancomycin group was 0.12 (with 95% CI 0.02–0.58, P = 0.003) with no study heterogeneity (I2 of 0%). The NNT was 26.8. There were two studies in the pediatric craniotomy subgroup with a total of 598 patients. The pooled risk ratio of infection in the vancomycin group was 0.02 (with 95% CI 0.00–2.92, P = 0.02) with no study heterogeneity with an I2 of 0%. The NNT was 39.5. There was only one study in the cranioplasty subgroup with a total of 258 patients. The pooled risk ratio of infection in the vancomycin group was 1.2 (with 95% CI 0.44–3.27, P = 0.72). Vancomycin decreased the incidence of SSI in all the subgroups significantly except for the cranioplasty subgroup which has only one study.
Meta-regression revealed that the number of patients in the study was not significantly associated with the effect size differences (F1, 10 = 0.74, P value = 0.41) [Figure 4]a. Our regression model did not explain any variability in our effect size data (R2 0%). Meta-regression revealed that the follow-up duration of the individual studies was not significantly associated with the effect size differences (F1, 9 = 0.25, P value = 0.63) [Figure 4]b. Our regression model did not explain any variability in our effect size data (R2 0%). Meta-regression revealed that the year of the publication of the study was not significantly associated with the effect size differences (F1, 10 = 0.15, P value = 0.70) [Figure 4c]. Our regression model did not explain any variability in our effect size data (R2 0%) [Figure 4d]. The results of the meta-regression are summarized in [Table 3].
Risk of SSI in cranial surgeries
The United States Centers for Disease Control defines SSIs “as infection related to an operative procedure that occurs at or near the surgical incision within 30 or 90 days (if prosthetic material implanted) of the procedure, depending on the type of procedure performed.” The criteria for defining SSI include pus discharge from the incision site; positive culture of the wound exudate; and wound dehiscence with clinical signs of infection (redness, tenderness, swelling, warmth). The common skin commensals are the most common pathogens causing SSIs after clean procedures. These include S. aureus, coagulase-negative staphylococcus, streptococcus. Besides the significant morbidity, SSI accounts for a whopping 22% of all healthcare-related infections, resulting in an estimated loss of $1–$10 billion for spinal surgeries alone. The majority of these cases return late to work resulting in an indirect loss of productivity.
A recent study involving 5,463 patients observed that the rate of SSI after common neurosurgical interventions was 1.94%, with the highest rate after vascular (3.4%) and CSF diversion procedures (3%). They also divided the SSI into superficial (superficial to bone), deep (involving bone), and organ/space infection (involving epidural space or deeper layers).
The goal of any therapy should be to prevent these infections in clean surgeries like cranial neurosurgery. There is no replacement of preoperative preparation and the use of antibiotics before the surgery. Various strategies have been adopted worldwide to decrease SSIs. Some of these include the liberal use of a hand sanitizer in the postoperative period, improving skin preparation in neurosurgical patients, and proper antibiotic prophylaxis guidelines. However, the ever-increasing rate of antibiotic resistance is a matter of concern.
S. aureus has been reported to be present in as many as 48.6% SSI, out of which 56% are resistant to beta-lactams., MRSA has been emerging as an increasingly common pathogen in postoperative infections, and thus, prompts the use of more selective antibiotics for prophylaxis such as vancomycin. In non-spinal neurosurgery, several risk factors have been identified as independently responsible for the increased risk of infections: male gender, age more than 60 years, smoking, lengthier preoperative hospital stay, surgery duration of more than 3 h, emergency surgery or re-exploration surgery, contaminated surgery, the use of implants, and CSF leakage. Also, skin flora such as S. aureus, Cutibacterium acnes, and coagulase-negative staphylococcus further increase the applicability of vancomycin in-infection prevention bundles. These are highly successful in the prevention of infection with over 53% reduction seen. Hence, the use of vancomycin powder prevents increased morbidity and the cost of treatment in the cases as evidenced by our results.
Goal and findings of our review as compared to other studies
The goal of our review was to evaluate the role of topical vancomycin in the form of an intrawound powder as a safe and effective modality based on available literature. We have unequivocally proven that only 1.6% (29 out of 1817) of the patients in the vancomycin group got SSIs as compared to 5.28% (86 out of 1629) in the control group. Systemic vancomycin does not appear to be as beneficial; although studies are limited, available data suggest it to be no more beneficial compared to standard prophylaxis.,,,, This is in addition to its safety profile when administered locally as there have been low to non-existent levels of vancomycin detected in the serum. We found that topical vancomycin is useful in decreasing the risk of infections with a pooled risk ratio of 0.29 and the NNT was 26.73. It was found to be useful in all the subgroups except in cranioplasty which had only one study.
The applicability of topical vancomycin is especially enhanced due to its use in surgeries with deep brain stimulation as the implant infection mandates explanation of the implants. In DBS or IPG placement the pooled risk ratio was 0.24 and NNT was 19. This finding is understandable as intravenous levels of antibiotics given systematically have no bearing due to the avascular nature of the implants. Moreover, the use of topical vancomycin in pediatric cases is particularly beneficial as local administration reduces the systemic side effects. The pooled risk ratio of infection in the vancomycin group in pediatric patients was 0.09 with the NNT being 35.
Our meta-analysis includes all the studies describing the use of vancomycin powder in preventing SSI in non-spinal neurosurgery published until now. The previous meta-analysis neither included pediatric subgroup analysis nor had a complete review of all the studies with ventriculoperitoneal shunts. Also, there were some discrepancies in data extraction. Considering the number of studies included in our review (14 vs. 10) and comprehensive subgroup analysis, the current review represents the most updated version of the use of vancomycin powder in the prevention of non-spinal SSIs.
Advantages of vancomycin
Topical vancomycin antibiotics prophylaxis is being currently utilized in various surgical specialties. It is warranted for operational sites where SSIs carry significant morbidity and mortality concerns. Data from neurosurgery has consistently demonstrated that antibiotic prophylaxis with vancomycin in cases where there is MRSA colonization in the anterior nares reduces the risk of SSIs. S. aureus is a skin commensal and colonizer of anterior nares in more than a third of the hospitalized patients and colonization before surgery is associated with an increased risk of postop S. aureus infections. As the most common organism causing SSIs are S. aureus, the efficacy of vancomycin found in our systematic review is understandable.
In one study, only 2% of the cases with cranial neurosurgery who received topical vancomycin had detectable serum levels of vancomycin. Thus, the reported drug reactions to vancomycin such as red man syndrome, ototoxicity, nephrotoxicity, Stevens-Johnson syndrome, and vasculitis are extremely rare when it is applied topically. A higher Body mass index (BMI), longer hospital stays, and Asian descent have been associated with increased risk of preoperative colonization of S. aureus and subsequent SSIs., Perhaps, if need be, a selective use can be considered but no evidence regarding the same has been found.
Risks of using topical vancomycin
Although there are concerns regarding the breeding of antibiotic-resistant S. aureus with the widespread use of prophylactic topical vancomycin. One recent study showed that the application of topical vancomycin in spinal surgeries is not related to the growth of vancomycin-resistant bacteria. However, an increased incidence of infections caused by gram-negative bacteria was seen. More long-term follow-up is required to unequivocally rule out any complications resulting from the use of topical vancomycin in neurosurgery., The retrieved studies reported no complications commonly associated with vancomycin, such as renal failure, skin rash, red man syndrome, anaphylaxis, seromas, or other local or systematic side effects. Even when the studies directly compared the adverse effects among two randomized groups, there were no differences in the cerebrospinal fluid (CSF) leakage, seroma formation, pseudo-meningocele, bone flap resorption, or seizures.
The limitations of our study include limitations inherent to the included studies; low level of evidence, and scientific rigor. None of them were designed efficiently, randomized, placebo-controlled, or double-blind trials. Most of them were retrospective institution-based experiences with a variety of protocols to reduce SSIs. Although a systematic search of the database was done for available studies, it was limited to the English language and those indexed in the medical literature.
The topical application of vancomycin is safe and effective in preventing SSIs in non-spinal neurosurgery. The effectiveness holds for procedures with implantable pulse generators and ventriculoperitoneal shunts in both adults and pediatric cases. The minimal increased cost is offset by the cost needed to treat infections and the indirect morbidity is also reduced. The safety and extremely low rate of serum absorption further increase the applicability. Perhaps, a better body of evidence in the form of a large multi-institutional randomized controlled trial will result in concrete results in its favor.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3]