In recent years, the number of instrumentation surgeries has increased, along with advances in spinal surgery techniques. In addition, improvements in perioperative management have enabled older and high-risk patients to undergo spinal instrumentation surgery. However, postoperative complications such as surgical site infections (SSIs) have been receiving increasing attention. SSIs in spinal instrumentation surgery are serious complications that not only deteriorate patient outcomes but also prolong hospital stays, increase medical costs, and sometimes have fatal outcomes [1,2]. The incidence of SSIs in spinal instrumentation surgery ranges from 2.6% to 11.9% [3,4]. Once an SSI is suspected, additional treatments may be required, including long-term antibiotic therapy, surgical debridement, and implant removal. The identification of risk factors for SSIs may allow for the proactive intervention and prevention of SSIs in patients at high risk. Early diagnosis and treatment of SSIs are crucial to prevent infection from causing serious conditions. Thus, prospective randomized controlled trials (RCT) are desired to identify risk factors for SSIs. However, because SSI is a postoperative complication, conducting RCTs is difficult. In addition, given the very low incidence of SSIs, recruiting a sufficient number of patients with SSIs is difficult even with large-scale retrospective studies. Even retrospective studies with a substantial number of patients focusing on spinal instrumentation surgery are scarce [5,6]. This may be the reason for the lack of consensus on the risk factors for SSIs after spinal instrumentation surgery. Although solving these problems is difficult, a meta-analysis or a systematic review integrating retrospective studies may help identify risk factors for SSIs. Therefore, this study aimed to retrospectively investigate SSI cases in spinal instrumentation surgery performed at Kyorin University School of Medicine and identify SSI-associated factors.

Of the 828 surgical patients, 15 (1.85%) resulted in SSIs. Of these, two had superficial SSIs, and 13 had deep SSIs. Two patients with deep SSIs resulted in disseminated intravascular coagulation (DIC) and died during hospitalization. The median postoperative time to SSIs was 24 days (interquartile range [IQR], 12.5–30.5 days), and the median time from onset of SSIs to surgical intervention was 10 days (IQR, 6–14.25 days) in patients who required revision surgery. Degenerative lumbar disease (n=410, 49.5%) was the most common, followed by adult spinal deformity (n=127, 15.3%) (Table 1). Posterior fusion was the most common surgery (n=612, 73.9%) (Table 2). Microbiological culture was performed in all 15 patients with SSIs, and 14 of them (93.3%) had positive cultures. Methicillin-resistant Staphylococcus epidermidis was the most common causative bacteria. Only one patient had multiple bacterial SSIs, and the remaining 13 patients had single bacterial SSI (Table 3). Twelve patients underwent revision surgery for SSIs, with debridement being the most common procedure, and implant removal was required in four patients (33.3%) (Table 4). Of the three patients treated with antibiotics alone, two developed DIC from sepsis caused by SSIs and died as their general condition deteriorated. Eleven patients were successfully resolved after a single surgery, and only one patient required multiple surgeries for infection control.

Univariate analysis between groups I and N showed no significant differences in patients’ baseline characteristics such as age, sex, BMI, and smoking history. As for comorbidities, only a history of steroid use showed a significant difference, whereas ASA-PS, diabetes, malignancy, and dialysis showed no significant differences. Laboratory factors showed significant differences in serum albumin and CRP, but not in hemoglobin and A/G ratio. As for surgical factors, significant differences were found in the number of fixed vertebrae and perioperative blood transfusion, whereas no significant differences were found in surgical time, blood loss, duration until drainage removal, and number of previous surgeries at the same site (Table 5).

To identify independent risk factors associated with SSIs in spinal instrumentation surgery, multivariate logistic analysis was performed using variables significant in the univariate analysis such as history of steroid use, serum albumin, CRP, number of fixed vertebrae, and perioperative blood transfusion. The result showed that a history of steroid use (odds ratio [OR], 5.38; 95% confidence interval [CI], 1.41–20.49; p=0.014), serum albumin (OR, 0.34; 95% CI, 0.13–0.84; p=0.020), and perioperative blood transfusion (OR, 5.85; 95% CI, 1.46–23.50; p=0.013) were the independent risk factors for SSIs (Table 6). The cutoff value of serum albumin was identified as 4.0 g/dL using the ROC curve (Fig. 1). The area under the curve was 0.74. The sensitivity and specificity were 0.67 and 0.75, respectively.

SSI is associated with poor surgical outcomes, extended hospitalization, increased medical costs, and compromised life prognosis in some cases. Therefore, identifying risk factors and appropriate countermeasures are crucial to prevent or reduce the incidence of SSIs. Although numerous studies have investigated risk factors for SSIs in spine surgery, research specific to spinal instrumentation surgery remains limited, and a consensus on the risk factors has yet to be established [8–11]. Therefore, this study was conducted to identify risk factors associated with SSIs after spinal instrumentation surgery at a single institution. Over the past 9 years, 828 spinal instrumentation surgeries were performed, and 15 (1.85%) resulted in SSIs. Multivariate logistic regression analysis identified a history of steroid use, serum albumin, and perioperative blood transfusion as independent risk factors.

The incidence of SSIs after spinal instrumentation surgery ranged from 2.6% to 11.9% in previous studies [3,4]. In the present study, the rate was 1.85%, which was lower than that reported in previous studies. Several studies have indicated that percutaneous pedicle screws (PPS) are associated with significantly lower infection rates than conventional open surgery [12,13]. In this study, posterior fixation with PPS was performed in 9.4% (n=78), which may partially explain the low incidence of SSIs compared with previous studies.

Several studies have reported that a history of steroid use was a risk factor for SSIs after spinal instrumentation surgery. Ismael et al. [14] reported that chronic steroid use increased the incidence of SSIs by two to threefold and mortality by fourfold. Wicke et al. [15] reported that these findings are likely related to the suppressive effect of steroids on the immune system. Lieber et al. [16] reported that steroid-induced hyperglycemia may also increase the SSI risk by a mechanism similar to delayed wound healing in patients with diabetes. These findings indicate that steroid use may contribute to SSIs through immunosuppression and delay wound healing because of hyperglycemia.

Preoperative hypoalbuminemia has also been reported as a risk factor for SSI development in spinal instrumentation surgery. Dong et al. [17] reported that albumin <3.0 mg/dL was an independent risk factor for SSIs. Bohl et al. [18] and He et al. [19] found that hypoalbuminemia (albumin <3.5 mg/dL) was an independent risk factor for SSIs and was associated with poor surgical wound healing. Therefore, preoperative nutritional correction in patients with hypoalbuminemia may help prevent SSIs. According to Burden et al. [20], preoperative oral nutritional supplementation of 600 kcal/day for approximately 8 days in patients with colorectal cancer reduced the incidence of SSIs. Another study reported that SSIs could be reduced in patients with gastric cancer with low nutrition by providing ≥25 kcal/kg ideal body weight per day for approximately 10 days before surgery [21]. While preoperative nutritional intervention has been proven to be effective in reducing SSIs among patients with cancer, very few studies have examined preoperative nutritional intervention for malnourished patients undergoing spinal instrumentation surgery. Thus, prospective studies focusing on patients undergoing spinal instrumentation surgery are warranted.

Raghavan and Marik [22] identified perioperative blood transfusion as a risk factor for SSIs, partially due to transfusion-related immunomodulation, a phenomenon in which antigens in blood products induce T-cell unresponsiveness and subsequent immunosuppression. In addition, bacterial contamination of blood products during transfusion was reported to cause SSIs [23]. These findings demonstrate the importance of appropriate measures for controlling perioperative bleeding to minimize the need for perioperative transfusion. Zhang et al. [24] reported that the use of tranexamic acid (TXA) can effectively reduce intraoperative bleeding, thereby minimizing the need for perioperative blood transfusion in patients undergoing multilevel spine surgery. Winter et al. [25] reported that the local use of TXA in spine surgery similarly has hemostatic efficacy and potentially improves safety compared with intravenous TXA therapy. Therefore, perioperative TXA administration may be worth considering, particularly in patients for whom perioperative blood transfusions are anticipated preoperatively.

This study has several limitations. First, given the retrospective design, the accuracy and precision of data collection may be insufficient. In addition, the single-center setting might cause biases in disease variation and patient selection. However, a single-institution study can offer the advantage of standardized treatment approaches, including surgical procedures and operating room conditions. Second, the SSI group (n=15) has a small sample size. The power analysis indicates that 26 SSI cases are needed, and with an incidence of SSIs of 2%, a population of 1,300 cases would be required. This study is underpowered, highlighting the need for a multicenter study to examine a sufficient number of cases. Although many studies have reported SSI cases related to spine surgery including patients without implants, this study focused on spinal instrumentation surgery, which limited the total number of patients and consequently the number of patients with SSIs. Third, this study contained various approaches such as posterior, anterior, or combined anteroposterior with posterior being the most common. This diversity is also a major limitation, as the posterior approach generally has a higher incidence of SSIs than anterior approach. Therefore, the approach may have affected the study results, and future analyses with a unified approach would be more desirable. Fourth, the study included various diseases or etiologies. In particular, trauma, metastatic spinal tumors, and osteoporotic vertebral fractures tend to be associated with a higher incidence of SSIs, and their inclusion may have affected the results. Ideally, analyzing the risk factors of SSIs in each etiology would provide clearer results; however, obtaining a sufficient number of patients in each etiology might be difficult. Considering these limitations, future multicenter studies are desirable.

A history of steroid use, hypoalbuminemia, and perioperative blood transfusions were found to be independent risk factors for SSIs after spinal instrumentation surgery. Preoperative nutritional intervention, appropriate anemia management, and intraoperative and postoperative bleeding control may reduce the incidence of SSIs. However, this study has several limitations, and more studies addressing these limitations are desirable.