Surgical site infection following minimally invasive lobectomy: Is robotic surgery superior?

Abstract Background Surgical site infection (SSI) in thoracic surgery remains a significant cause of morbidity and prolonged hospitalization. Minimally invasive surgery (MIS) has significantly reduced the risk of SSI. We intended to compare whether there was difference between video‐assisted thoracic surgery (VATS) and robotic‐assisted thoracic surgery (RATS) in SSI and highlight possible factors influencing SSI in lobectomy. Methods This retrospective study analyzed patients who underwent minimally invasive lobectomy from January 2018 to December 2019. All patients' clinical characteristics and surgery‐related information which may be related to the likelihood of SSI were recorded. Results A total of 1231 patients' records were reviewed with 806 VATS and 425 RATS. SSI was classified as deep or superficial SSI. Eighty‐six (7.0%) patients were found to develop an SSI with 62 patients having deep infections and 24 had superficial infection. No statistical difference in the incidence rate and category of SSI was observed between patients undergoing VATS and RATS. Conclusions There was no difference in the incidence of SSI between VATS and RATS lobectomy. Male gender, heavy smoking, uncontrolled diabetes mellitus, body mass index (BMI) > 27.9, more blood loss, and the higher National Healthcare Safety Network (NHSN) risk index score (1 or 2) were the independent risk factors of SSI following minimally invasive lobectomy, while male gender, uncontrolled diabetes mellitus, BMI > 27.9, more blood loss and the higher NHSN risk index score (1 or 2) were the main predictors of deep SSI.


| INTRODUCTION
Infection is the most common postoperative complication. Among them, prevention of surgical site infection (SSI) has always been an important part of surgical technology. It is intuitively that SSI leads to prolonged postoperative hospital stay and an increase in related medical expense. 1 In severe cases, it can even be life-threatening. Since 1999, many countries have formulated and updated the guidelines for the prevention of SSI. The key basic measures include aseptic approaches, surgical skills, blood glucose control, maintaining normal body temperature, and appropriate perioperative antibiotics.
At present, surgery is still the main treatment for early-stage and selected locally advanced non-small cell lung cancer (NSCLC). With the increasingly complex types of surgery, combined with the long-term use of hormone, neoadjuvant therapy, previous history of thoracic surgery or trauma, the risk of pulmonary complications can be increased. Many studies have shown that compared with traditional open surgery, minimally invasive surgery (MIS) represented by laparoscopic or thoracoscopic, endoscopic, and robotic-assisted surgery has significantly reduced the risk of SSI. [2][3][4][5] However, there is little evidence for this in thoracic surgery. Over the last two decades, robotic-assisted thoracic surgery (RATS) is changing the concept and mode of MIS in general thoracic surgery practice, especially in pulmonary lobectomy for early-stage NSCLC. However, high medical costs and lack of haptic feedback are the major limiting factors for robotic-assisted surgery to become the standard technique of minimally invasive surgery all over the world. 6 Current studies have shown that RATS has similar perioperative outcomes and comparable oncological outcomes as video-assisted thoracic surgery (VATS), 7-10 while few studies have focused on the incidence of SSI in the two minimally invasive procedures. Based on 2 years of data, antibiotics use density (AUD) of VATS lobectomy was slightly lower than that of RATS lobectomy in our hospital ( Figure 1). AUD was expressed as daily dose of antibiotics per 100 patient days (occupied bed days). Therefore, we explored whether there was difference in SSI between RATS and VATS lobectomy, and aimed to determine the related risk factors of SSI following minimally invasive lobectomy.

| Study design
This retrospective study analyzed patients who underwent pulmonary lobectomy at Shanghai Chest Hospital from January 2018 to December 2019. All patients over 18 years old and less than 90 years old who undergone minimally invasive lobectomy by two certified lung surgeons with at least 5 years of experience in robotic technique and VATS were included. Patients with other resection such as wedge resection, segmentectomy, (bi)lobectomy, bronchial sleeve resection, or pneumonectomy were excluded. Patients undergoing other surgical procedure such as open technique were excluded. Patients who were lost to follow-up were also excluded. F I G U R E 1 Antibiotics use density, VATS, video-assisted thoracic surgery; RATS, robotic-assisted thoracic surgery

| Data collection
From patients included in the analysis, we collected clinical characteristics and surgery-related information which may be related to the likelihood of SSI. Patients' clinical characteristics included age, gender, body mass index (BMI), blood type, smoking history, diagnosis, preoperative concentrations of hemoglobin, albumin, diabetes mellitus, existence of hypertension, American Society of Anesthesiologist (ASA) score. Surgery-related information included surgery side, duration of operation, surgical procedure (VATS or RATS), and blood loss. Duration of operation was reported in minutes from the start of the surgery (incision) to the closing of the skin incision. National Healthcare Safety Network (NHSN) risk index was also recorded which ranged from 0 to 3. As an internationally recognized method for stratifying surgical risk, the NHSN risk index contains duration of operation, surgical wound class, and ASA index. Each variable's cutoff values were a contaminated or dirty surgical incision, an operative duration of 180 minutes and an ASA score of III, with 1 point assigned when each variable exceeded its respective cutoff value.
The main outcome measure was the incidence of SSI within 30 days after operation. According to the Centers for Disease Control and Prevention/National Nosocomial Infections Surveillance guidelines, SSI was defined as an infection that involves skin or subcutaneous tissue (superficial), fascia or muscle layers (deep), or any other anatomic components manipulated during surgery (organ space). In this study, deep and organ/space infection were considered as "deep SSI" because the two are more harmful to health than superficial SSI. Only the first episode of SSIs was included for patients who had more than one SSI during the study period.
All patients underwent perioperative prophylaxis consisting of intravenous administration of 2.0 g Cefazolin Sodium Pentahydrate for injection or 1.5 g cefuroxime within 30 to 60 minutes before the surgical incision, and repeat doses every 12 h for 48 h. Patients with documented β lactam allergy received clindamycin 0.6 g. If the operation time was more than 3 h, an additional dose of antibiotics was added. All patients had got the same immediate preoperative treatment according to approved skin preparation protocols and use of disinfection solutions.
The follow-up time was 30 days after the operation. If the patients were discharged, they were followed up by telephone, outpatient, or readmission.

| Statistical analysis
The two-tailed t test was used for continuous variables, unless the data were non-normally distributed. For such cases, we used the Mann-Whitney U test for comparison. The χ 2 test or Fisher exact test was used for categorical variables, which were summarized as frequencies and percentages. Independent prognostic factors for SSI were analyzed using univariate and multivariable logistic analysis. All statistical analyses were two-sided, and a p value of 0.05 was considered statistically significant. All statistical analyses were performed using R software (version 3.5.2).

| Operative techniques
Our center has more than 10 years of experience in MIS. Robotic technique was introduced in 2009. In all the patients under general anesthesia, double-lumen tracheal intubation, and isolated-lung ventilation were performed in the lateral decubitus position.
VATS was performed via three or four incisions. Via a 12-mm incision, a 30° thoracoscope was inserted in the 6th or 7th intercostal space on the anterior axillary line. A 40-mm utility incision was performed in the anterior axillary line in the 3th or 4th intercostal space. One 12 mm incision was performed at the 8th intercostal space posterior axillary line. If necessary, a fourth port for assistance was made at the ninth intercostal space on the posterior axillary line. VATS usually needs one surgeon and two assistants, one in charge of the thoracoscope and the other helping retract the lung and exposing the operating fields.
RATS was performed using the da Vinci® robot surgery type Si (Intiuitive Surgical, Sunnyvale, California), and using a three-port approach with a utility incision. First, a 12-mm incision for the camera port was made at the 7th or 8th intercostal space posterior axillary line. Second, two 8-mm port incisions were symmetrically performed at the 7th intercostal space mid-axillary line and the 9th intercostal space infrascapular line separately. A 40-mm utility incision was made at the 3th or 4th intercostal space on the anterior axillary line, which used by the bedside assistant for retracting the lung, exposing the operating fields, and stapling and specimen retrieval. Carbon dioxide was insufflated to a pressure of 8-10 mmHg.

| RESULTS
From January 2018 to December 2019, a total of 1231 cases who fulfilled the selection criteria were analyzed.

| Population characteristics
The clinical characteristics of all 1231 patients are exhibited in Table

| Perioperative results and surgical site infections
The surgery-related information and perioperative outcomes of the included patients by approach are listed in was no difference in the utilization rate of restricted and special antibiotics between the two groups. Table 3 highlight the demographic and operative information of the cohort by the presence or absence of SSIs. The logistic regression analysis showed that the risk factors significantly associated with the occurrence of SSI were male gender, elderly, heavy smoking, COPD, ASA III-IV, uncontrolled diabetes mellitus, BMI > 27.9, lower level of albumin, hypemricemia, the longer operation time, more blood loss, and the higher NHSN risk index score (1 0r 2). Furthermore, the multivariate analysis showed in Figure 2 revealed that the independent risk factors of SSI following minimally invasive lobectomy were male gender, heavy smoking, uncontrolled diabetes mellitus, BMI > 27.9, more blood loss, and the higher NHSN risk index score (1 0r 2).
In term of deep SSI, Table S1 revealed that male gender, elderly, heavy smoking, COPD, ASA III-IV, uncontrolled diabetes mellitus, BMI > 27.9, hypemricemia, the longer operation time, more blood loss, and the higher NHSN risk index score (1 0r 2) were the risk factors while the multivariate analysis showed in Figure S1 revealed male gender, uncontrolled diabetes, BMI > 27.9, more blood loss, and the higher NHSN risk index score (1 0r 2) were the main predictors of deep SSI.

| DISCUSSION
Despite the recent popularization of minimally invasive approaches, a deeper understanding of the pathogenesis of infections, and routine use of preoperative antibiotics, SSI is still a major cause of morbidity and readmissions for patients undergoing thoracic surgery. SSI is considered as a priority in postoperative infection control that surgeons, nurses, and clinical pharmacists should pay attention to, especially in low-and middleincome countries. Indeed, this problem has not been well monitored, most of which are discovered, judged, and treated by clinical medical staff based on experience. In the field of thoracic surgery, due to the use of cutting stapler, the incision of lobectomy is clean/ contaminated without opening the airway. The rate of SSI about clean/contaminated incision is between 5% and 10%. Minimally invasive techniques, which limit size of incision and area of skin contamination, have been shown to reduce the incidence of SSI in thoracic surgery.
We conducted a comparative analysis of SSI between VATS and RATS for lobectomy at a single institution. To the best of our knowledge, this comprehensive observational study is the first to compare the incidence rate of SSI in patients undergoing robot-assisted and video-assisted lobectomy. Further, we investigated the risk factors of SSI following minimally invasive lobectomy.
Consistent with previous research results, 11 male gender was independent risk factors for SSI following minimally invasive lobectomy (p = 0.026), which was also risk factors for deep SSI (p = 0.012). The impact of gender differences on SSI also occurs in other types of surgical procedures. 12,13 Taking into account the influence of gender, some scholars believe that males are related to smoking, 14 and smoking is a risk factor for SSI. 15 However, this study did not show that smoking was an independent risk factor for SSI after minimally invasive lobectomy.
We also showed that blood loss exceeding 100 ml and the higher NHSN index (1 0r 2) were independent risk factors of SSI, which were consistent with previous studies. The NHSN index contains the wound class, the ASA index, and the operation time. In our cohort, all patients had clean-contaminated wounds, so the score of this item was consistent, and none of them scored. Patients with higher ASA score (level of III or V), or patients with longer operation time (up to 180 minutes) may be inclined to suffer more risks of SSI following lobectomy.
Obesity is a risk factor for SSI, which has been found in general surgery and orthopedic surgery. 3,16,17 In our study, BMI > 27.9 was an independent risk factor for SSI, but this conclusion was not reached in deep SSI. We speculate that obesity plays an important role in the occurrence of superficial SSI, which requires a larger cohort to study. Several studies have found diabetes mellitus is an important risk factor in general and orthopedics surgery. [18][19][20] However, there is no specific study to discuss the association between SSI and diabetes mellitus after lobectomy. In our study, according to the level of glycated albumin, glycosylated hemoglobin, and perioperative blood glucose, patients with diabetes mellitus were divided into diabetes mellitus controlled group and diabetes mellitus uncontrolled group. Intriguingly, we identified that the uncontrolled diabetes mellitus acted as an independent risk factor of SSI following minimally invasive lobectomy, while controlled diabetes mellitus was not. Chronic hyperglycemia can cause neutrophil dysfunction, resulting in a poor baseline inflammatory response. High levels of perioperative blood glucose may have a synergistic effect. Therefore, continuous monitoring and effective control of perioperative blood glucose levels in patients with diabetes mellitus could alter the risk of SSI.
As is well-known, VATS and robotic-assisted technique were becoming cumulatively common worldwide. The evolution of MIS has allowed for smaller incisions with less potential raw surface for exposure to pathogens and remarkably reduced site manipulation compared with open technique. This study demonstrated that the incidence rate of SSI following minimally invasive lobectomy was 7.0%, and there was no difference on SSI between the two surgical technique (VATS 7.1% vs. RATS 6.8%, P = 0.871).
Although RATS had a shorter operation time than VATS in our study, the shorter operation time did not bring the benefit of reducing the incidence rate of SSI. Thus, there is no evidence that robotic-assisted surgical technique can further reduce the incidence of SSI. It is still necessary to verify whether the high cost of robotic surgery justifies its use.

| CONCLUSIONS
There was no difference in the incidence of SSI between VATS and robotic-assisted lobectomy. Male gender, BMI > 27.9, uncontrolled diabetes mellitus, more blood loss, and the high NHSN index (1 0r 2) were independent predictors of SSI.

SUPPORTING INFORMATION
Additional supporting information may be found in the online version of the article at the publisher's website.