Surgical site infection in spinal surgery: a bibliometric analysis

All data involved in our study were obtained from the Web of Science Core Collection database. Identified the keywords and synonyms (surgical site infection; surgical site infections; surgical site infection; SSI), limited the article type to "Article" or "Review," with a language filter for English and a time range spanning from 1900 to 2022. The specific query was as follows: ((((TI = (surgical site infection OR surgical site infections OR surgical site infection OR SSI)) AND TI = (spine OR spinal OR lumbar OR thoracic OR cervical OR sacral)) AND DT = (Article OR Review)) AND LA = (English)) AND PY = (1900–2022).

The search was conducted on November 8, 2022. A total of 310 publications were retrieved, of which 307 were selected for further discussion after screening. We extracted and recorded the title, abstract, authors, publication year, country, journal, institution, keywords, and citation frequency for further analysis.

The analysis of data was carried out using VOSviewer, Scimago Graphica, and Microsoft Excel 2016. VOSviewer and Scimago Graphica are capable of performing bibliometric data analysis and can provide complementary benefits when used together. We analyzed author, country, journal, institution, and keywords with the aforementioned software.

Two authors independently extracted bibliometric indicators according to the specified query and discussed the differences until a consensus was reached. We used VOSviewer and Microsoft Excel 2016 to extract and analyze the data, including author, journal, institution, country, citations, keywords, and research trends. VOSviewer and Scimago Graphica were applied for data visualization.

A total of 307 articles related to spinal SSI were identified. Overall, articles and citations have had a linear growth trend from 2008 to 2022 [R²(Publications) = 0.9666, R²(Citations) = 0.9908] (Fig. 1). Among all years, 2018, 2020, 2021, and 2022 were the most contributed years, with the annual publications has reached 35 or more. Judging from the frequency of citations, all publications were cited 7225 times, with an average of 23.53 citations per publication. All of these indicate that research on spinal SSI has been receiving increasing attention in recent years, as the number of related publications continue to rise.

All articles came from 37 different countries, of which the USA was the most contributed country with 138 articles, followed by Japan (n = 52), China (n = 52), and Canada (n = 16). Based on the advantage of the total number of articles, the USA was far ahead in total citation frequency with 4935 citations, compared to Japan (n = 791), China (n = 631), and Canada (n = 493). The co-authorship between countries was analyzed using VOSviewer and Scimago Graphica software. By drawing network visualization graphics, it was evident that most articles were published from North America and Asia. The USA stands as a prominent research hub for global spinal SSI, exhibiting a close-knit collaboration with Canada and the Netherlands, as indicated in Fig. 2.

A total of 505 institutions participated in the publication of SSI-related articles. The top 10 contributing institutions were Johns Hopkins University (US; n = 14), Mayo Clinic (US; n = 14), University of Tokyo (Japan; n = 11), Columbia University (US; n = 10), University of Washington (US; n = 9), Thomas Jefferson University (US; n = 8), University of British Columbia (Canada; n = 7), Japanese Red Cross Medical Center (Japan; n = 7), Musashino Red Cross Hospital (Japan; n = 7), and Saitama Medical University (Japan; n = 7) (Table 1). In terms of total citations, Johns Hopkins University ranked first (14 articles; 835 citations), followed by Washington University (five articles; 749 citations), Vanderbilt University (four articles; 526 citations), and Harvard University (five articles; 399 citations).

We employed VOSviewer to construct network visualization and overlay visualization to analyze the co-authorship between institutions, with a minimum article count of 4, and a total of 36 institutions were included in the analysis (Fig. 3A). The thickness of the lines reflected the collaborative strength between institutions. University of Tokyo had the highest total link strength (n = 57), followed by Japanese Red Cross Medical Center (n = 54) and Musashino Red Cross Hospital (n = 54). Johns Hopkins University, Mayo Clinic, Columbia University, and 14 other institutions formed the central cluster. Japanese Red Cross Medical Center, Musashino Red Cross Hospital, University of Tokyo, and eight other institutions constituted another primary cluster with close cooperation among institutions. Analysis of the overlay visualization revealed that Johns Hopkins University and Vanderbilt University were the main institutions studying spinal SSI in the early years (around 2011), while Saitama Rehabilitation Center, Gunma University, and Rothman Institute have become the main forces in recent years (around 2020) (Fig. 3B).

All articles were published in 83 academic journals. The top 10 journals accounted for 54.7% of the total publications and 65.8% of the total citations (Table 2). Among them, Spine contributed 47 articles (1715 citations), far exceeding European Spine Journal (27 articles; 753 citations) and Spine Journal (20 articles; 799 citations).

Keywords were analyzed using the network visualization function of VOSviewer, with a minimum word frequency of 10. A total of 50 high-frequency keywords were identified. All keywords were divided into three groups, "prevention," "treatment," and "prognosis," and the same color represented the same research direction. Among the "prevention" group, the keywords with the highest frequency were "spine surgery," "prevention," and "postoperative infection." In the "treatment" group, the most normal keywords were "fusion," "wound-infection," and "surgery." In the "prognostic" group, "surgical site infections," "risk-factors," and "impact" were common (Fig. 4A).

To further explore the evolving trends of keywords over the years, we used overlay visualization to analyze keywords and signify the average year of their appearance through color variation. The shades of purple represent an earlier average year of appearance, around 2017, whereas the shades of yellow indicate a later average year, around 2019. During the initial phase of researching spinal SSI, keywords such as "adolescent idiopathic scoliosis," "care," and "wound-infection" appeared more frequently, while "vancomycin," "risk," and "experience" have emerged as the main focal points of research in recent years (Fig. 4B).

Analyzing the top 100 cited articles can be used to evaluate the research hotspots and trends in the field. All articles were sorted in descending order of citation frequency, and further analysis was performed on the top 100 articles (Table 3). The top 100 most cited articles were published between 2008 and 2020. Among them, 2018 was the most productive year with 15 articles, followed by 2015 (11 articles) and 2014 (10 articles) (Fig. 5).

The top 100 cited articles were authored by researchers from 21 different countries. The USA, being the most influential country in the field of spinal SSI, contributed 60 articles (4499 citations), which was significantly more than Japan (15 articles; 584 citations), Canada (10 articles; 482 citations), China (10 articles; 453 citations), and the Netherlands (9 articles; 538 citations) (Fig. 6).

In terms of institution contributions, Johns Hopkins University contributed the most (n = 9), followed by Thomas Jefferson University, University of Washington, and University of British Columbia, each with five articles (Table 4).

Regarding journal contributions, the 100 most cited articles were published in 30 different journals. The most productive journal was Spine, with a total number of 21 articles (1531 citations), followed by European Spine Journal (13 articles; 644 citations), Spine Journal (12 articles; 722 citations), Journal of Neurosurgery-Spine (eight articles; 590 citations), and Journal of Bone and Joint Surgery-American Volume (five articles; 795 citations) (Table 5).

In terms of authors, the most productive authors were Cohen, David B. and ter Gunne, Albert F. Pull, both of whom published six articles, followed by McGirt, Matthew J. and Vaccaro, Alexander R. (each four articles) (Table 6).

The most common research topic was Risk Factors for spinal SSI (n = 52), followed by Prevention (n = 26) and Incidence (n = 22) (Fig. 7). Notably, eight of the 100 articles were related to spinal SSI in pediatrics.

Over the past 15 years, publications related to spinal SSI have increased steadily. In particular, the annual publication rate has been consistently above 35 articles in the past three years, indicating a growing interest in research related to spinal SSI. This may be related to the increased morbidity caused by comorbid factors such as diabetes, antibiotic abuse, immune suppression, and spinal instrumentation, as well as the improvement of diagnostic sensitivity [10]. As a global research center for spinal SSI, the USA has contributed 37% of the total article count and 56% of the total citation count, demonstrating its dominant position in this field. Among all institutions, Johns Hopkins University has contributed the most, with a total of 14 articles and 835 citations. However, the institution with the highest average citation count was Washington University, highlighting its significance in spinal SSI research. The top five journals in terms of publication volume were Spine, European Spine Journal, Spine Journal, World Neurosurgery, and Journal of Neurosurgery-Spine. However, among journals with more than five publications, the one with the highest average citation frequency was Journal of Bone and Joint Surgery-American Volume (132.7 times). Clinicians and scholars interested in spinal SSI should pay more attention to the abovementioned journals. It is noteworthy that the institutions and journals mentioned above are all affiliated with the USA, which explains why it occupies a dominant position in this field.

Keywords represent the core of the article, and the evolution in keywords over time can reflect the research trends in that field. The network visualization analysis of the keywords demonstrated that "risk-factors," "prevention," and "fusion" were the centers of the keyword clusters. As a common cause of C (HAI) and death, SSI accounts for roughly 20% of all HAIs [11]. The incidence of SSI in spinal surgery is approximately 2%, and it has shown an upward trend in recent years due to the extensive use of implants and the increased complexity of surgical procedures [12]. Smith et al. [13] evaluated the risk factors relevant to spinal SSI and found that the incidence of wound infection was significantly higher in patients who underwent spinal fusion or instrumentation, which may be attributed to the increased risk and complexity of the procedures. Spinal SSI can lead to instrumentation failure, neurological dysfunction, paraplegia, and even death, causing catastrophic consequences for patients. Due to the heterogeneity of patients and the diversity of treatments, there is currently a lack of universally applicable management guidelines for spinal infections [14]. Therefore, the focus of SSI management is on multifactorial comprehensive management with an emphasis on prevention, including preoperative risk stratification and intraoperative measures [15].

We utilized the overlay visualization function to analyze the trends in research hotspots over time. "Fusion" was the center of early keywords and was closely linked with other keywords such as "Instrumentation" and "Scoliosis" during the same period, indicating that the etiology of spinal SSI was the focus of early research. Compared to other orthopedic cleaning surgeries, spinal instrumentation surgery has a higher infection rate [16]. Based on a study of 108,419 cases by Justin S. Smith et al., patients with spinal scoliosis had a higher infection rate (3.7%) than those with degenerative spinal diseases (1.4%), spondylolisthesis (2.1%), and fractures (2.0%). In addition, patients who underwent spinal fusion had a 33% higher infection rate than those who did not undergo fusion (2.4% vs. 1.8%, P < 0.001), and patients with implants had a 28% higher infection rate than those without implants (2.3% vs. 1.8%, P < 0.001), which is likely related to the higher complexity and risk of instrumentation surgery [13]. Notably, pediatric spinal SSI has attracted the attention of scholars in the early stage of research (Avg. pub. Year: 2017.00). As mentioned above, eight of the top 100 cited articles were related to pediatrics, indicating that pediatric SSI may have its specificity. Cahill et al. [17] found that, unlike adults, the infection risk after spinal surgery in pediatric patients was inversely proportional to age, which meant younger patients have a higher infection rate. Furthermore, diagnosis and treatment in pediatrics are more challenging due to factors such as poor expression ability, low positive blood cultures, lack of specific early symptoms, and imaging features [18]. Therefore, laboratory and imaging examinations are the mainstays of early diagnosis in pediatrics spinal SSI. C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and absolute neutrophil count (ANC) are commonly used laboratory indicators. However, confounding factors after spinal surgery often affect their sensitivity and specificity. In recent years, studies have suggested that markers such as c (SAA), procalcitonin (PCT), interleukin-6 (IL-6), and leukocyte esterase can serve as crucial auxiliary diagnosis modalities for spinal SSI, with better predictive effect than CRP and leukocyte levels [15, 19].

In recent years, the high-frequency keywords have shifted to such as "risk," "vancomycin," and "experience," indicating a shift in the research focus of spinal SSI from etiology to prevention. According to the estimation by the Society for Healthcare Epidemiology of America (SHEA), up to 60% of SSIs can be prevented by following evidence-based guidelines [20]. The prevention of spinal SSI focuses on the risk factors, and numerous scholars have conducted detailed and comprehensive research on related factors. Currently, well-established factors that are significantly associated with SSI include diabetes, obesity, smoking, previous history of SSI, increased intraoperative blood loss, and prolonged operation time, among others [5, 21]. Preoperative, intraoperative, and postoperative prevention measures are another crucial part of reducing the incidence of SSI. Typical prevention measures include intrawound vancomycin powder, prophylactic antibiotics, closed-suction drainage, povidone–iodine irrigation, and incision closure with 2-octyl-cyanoacrylate [22]. Among them, intrawound vancomycin powder is the most extensively studied prevention measure. As previously mentioned, among the top 100 cited articles, 26 focused on prevention, including 12 related to vancomycin (Fig. 7). Since its first application as a local preventive measure for spinal SSI in 2011, vancomycin has been demonstrated to significantly reduce the incidence of SSI with a favorable safety profile [23]. However, the optimal dose and application site of vancomycin powder are still controversial and need further experimental research [24]. Notably, vancomycin powder has also been confirmed to be safe in children with spinal injuries [25].

The frequency of citations represents the degree of recognition of an article in the field, which can roughly reflect the quality and influence of the article.

The most cited article in the field of spinal SSI is "Risk factors for surgical site infection following orthopaedic spinal operations," published by Olsen et al. [12] in Journal of Bone and Joint Surgery-American Volume in 2008, with a total of 507 citations. This retrospective case–control study analyzed 2316 postoperative spinal patients, among whom 46 were diagnosed with SSI, and 227 patients without SSI were selected as the control group. A comprehensive and in-depth analysis of the independent risk factors for SSI after spinal surgery was performed using univariate and multivariate logistic regression analyses. The overall incidence of SSI was 2.0%, with a superficial SSI rate of 0.8% (18 cases), a deep SSI rate of 0.9% (20 cases), and an organ space SSI rate of 0.3% (8 cases). The study identified diabetes, poor timing of prophylactic antibiotics, hyperglycemia, obesity, and the involvement of two or more residents as independent risk factors for SSI in laminectomy, discectomy, and spinal fusion. In contrast, cervical spine surgery was independently associated with a significantly lower risk of SSI. Among all independent risk factors, diabetes had the strongest correlation with spinal SSI, followed by poor timing of prophylactic antibiotics and hyperglycemia.

The second most cited article is "Incidence, Prevalence, and Analysis of Risk Factors for Surgical Site Infection Following Adult Spinal Surgery" by ter Gunne, Albert et al. [6], published in Spine in 2009, with a total of 315 citations. The purpose of this study was to compare infected patients with uninfected patients through a retrospective cohort study, to calculate the incidence of SSI and identify the risk factors for postoperative wound infection after spinal surgery. A total of 3174 patients were enrolled in the study. There were 132 cases of SSI (4.2%), including 70 cases of superficial SSI (2.2%) and 84 cases of deep SSI (2.6%). Multivariate logistic regression analysis revealed that estimated blood loss (EBL) exceeding 1L, previous history of SSI, and diabetes were independent risk factors for spinal SSI. Analysis of superficial SSI showed that obesity, hypertension, multilevel fusion, surgical approach, and operative time longer than 2 h were associated with an increased risk of superficial SSI. However, only obesity significantly increased the risk of superficial infection, while the anterior surgical approach significantly reduced the risk of SSI. In deep infection, diabetes, obesity, previous history of SSI, surgery for spinal deformity, multilevel fusion, surgical approach, spinal uninstrumented fusion, osteotomies, and operative time longer than 2 h were associated with a higher infection rate. In comparison, discectomy alone and the anterior surgical approach had a lower infection rate. Among all risk factors, diabetes, obesity, previous history of SSI, and operative time longer than 2 h were independent risk factors for deep SSI.

"Reduced surgical site infections in patients undergoing posterior spinal stabilization of traumatic injuries using vancomycin powder," published in The Spine Journal by Devin, Clinton J. et al. in 2011, ranked third with 201 citations [26]. In this retrospective cohort study, 110 patients who underwent posterior spinal stabilization due to trauma were analyzed to evaluate the clinical efficacy of topical intrawound vancomycin powder to prevent SSI. All patients accepted standard intravenous antibiotic prophylaxis, on which cases were divided into a treatment group (with topical vancomycin; n = 56) and a control group (without topical vancomycin; n = 54). There were no statistical differences in age, BMI, surgical levels, or other patient parameters between the two groups. The study found that none of the patients with topical vancomycin prophylaxis suffered SSI and had no adverse reactions, while the control group had seven SSIs (13%), including two superficial and five deep. The study concluded that topical vancomycin powder could significantly reduce the incidence of SSI in patients with spinal trauma, which was consistent with previous studies [27]. The limitation of this study was the potential mismatch between the groups. The operation time in the control group was significantly longer than that in the treatment group (P = 0.01), and it was unclear whether it affected the incidence of SSI in the control group.

Among the top 100 cited articles, the latest article is "Incidence of Surgical Site Infection After Spine Surgery A Systematic Review and Meta-analysis" by Zhou et al., [5] published in Spine in 2020. This study conducted a meta-analysis of 27 studies related to spinal SSI, with subgroup analyses for SSI type, age, BMI, diagnosis, surgical site, approach, procedure, minimally invasive or not, operative time, blood loss, topical vancomycin powder, and bacterial culture results. The total incidence of SSI was 3.1% (603 of 22,475 cases), with a superficial SSI of 1.4% and a deep SSI of 1.7%. Among all primary diseases, patients with neuromuscular scoliosis had the highest incidence of SSI (13.0%), while patients with idiopathic scoliosis had the lowest infection rate (2.6%). Subgroup analysis indicated that the incidence of SSI after thoracic surgery (3.7%) was slightly higher than that of cervical (3.4%) and lumbar (2.7%) surgery. Besides, posterior surgery, instrumented surgery, traditional open spinal surgery, age over 60 years, prolonged operation time (≥ 3 h), and increased intraoperative blood loss (> 500 ml) all increase the risk of SSI. Topical vancomycin powder significantly reduced the risk and incidence of SSI (1.9% vs. 4.8%). Nevertheless, further research is needed to confirm the specific efficacy and potential side effects of topical vancomycin. Microbiological cultures suggested that the detection rate of Staphylococcus was the highest (50.2%), with Staphylococcus aureus and Staphylococcus epidermidis being the most common pathogenic bacteria.

This article has several limitations. Firstly, we retrieved from the database in English, and influential articles in other databases or non-English languages may be omitted. Secondly, like other bibliometric studies, our study will inevitably have a "cumulative effect" and may omit recent influential articles. Finally, citation frequency is not the only criterion to evaluate the quality and influence of articles. In future studies, other factors should be considered comprehensively.