SURGICAL INFECTIONS Volume 16, Number 1, 2015 ª Mary Ann Liebert, Inc. DOI: 10.1089/sur.2013.260
Performance Assessment of the Risk Index Category for Surgical Site Infection after Colorectal Surgery Masanori Watanabe,1 Hideyuki Suzuki,1 Satoshi Nomura,1 Hidetsugu Hanawa,1 Naoto Chihara,1 Satoshi Mizutani,1 Masanori Yoshino,1 and Eiji Uchida 2
Abstract
Background: The traditional National Healthcare Safety Network (previously National Nosocomial Infections Surveillance) risk index is used to predict the risk of surgical site infection across many operative procedures. However, this index may be too simple to predict risk in the various procedures performed in colorectal surgery. The aim of this study was to evaluate the usefulness of the risk index by analyzing the impact of the risk index factors on surgical site infection after abdominal colorectal surgery. Methods: Using our surgical site infection surveillance database, we analyzed retrospectively 538 consecutive patients who underwent abdominal colorectal surgery between 2005 and 2010. Correlations between surgical site infection and the following risk index factors were analyzed: length of operation, American Society of Anesthesiologists score, wound classification, and use of laparoscopy. The 75th percentile for length of operation was determined separately for open and laparoscopic surgery in the study model. Results: Univariate analyses showed that surgical site infection was more strongly associated with a > 75th percentile length of operation in the study model (odds ratio [OR], 2.07) than in the traditional risk index model (OR, 1.64). Multivariable analysis found that surgical site infection was independently associated with a > 75th percentile length of operation in the study model (OR, 2.75; 95% confidence interval [CI], 1.66–4.55), American Society of Anesthesiologists score ‡ 3 (OR, 2.22; 95% CI, 1.10–4.34), wound classification ‡ III (OR, 5.29; 95% CI, 2.62–10.69), and open surgery (OR, 2.21; 95% CI, 1.07–5.17). Performance of the risk index category was improved in the study model compared with the traditional model. Conclusions: The risk index category is sufficiently useful for predicting the risk of surgical site infection after abdominal colorectal surgery. However, the 75th percentile length of operation should be set separately for open and laparoscopic surgery.
T
The N-RIC is a widely used score that is operation-specific and is based on surveillance data collected prospectively [1– 3,8,9]. The index ranges from 0 to 3 points and is determined by three independent variables weighted equally. One point is scored for each of the following when present: (1) American Society of Anesthesiologists (ASA) physical status score > 2; (2) wound classified as contaminated or dirty–infected; and (3) operative time longer than the approximate 75th percentile for the specific operation being performed (3 h for colon surgery) [2,3]. In 2001, the N-RIC was modified to introduce laparoscopic surgery as an additional factor in a number of surgical procedures including cholecystectomy, colon surgery, gastric surgery, and appendectomy, to account for the
he National Healthcare Safety Network (NHSN), previously the National Nosocomial Infections Surveillance (NNIS), risk index category (N-RIC) was developed in the early 1990s and has been used to predict the risk of surgical site infection (SSI) across a broad range of operative procedures [1–4]. However, the N-RIC provides a generic score that has been criticized for being overly simplistic [5], and does not consider some aspects of colorectal surgery, such as stoma formation or the use of laparoscopy [6]. Gervaz et al. [7] reported that generic scores such as the N-RIC are not used by colorectal surgeons. We therefore considered it necessary to evaluate whether the N-RIC is useful for predicting SSI after colorectal surgery.
1 2
Institute of Gastroenterology, Nippon Medical School Musashikosugi Hospital, Kawasaki, Kanagawa, Japan. Department of Surgery, Nippon Medical School, Tokyo, Japan.
84
RISK INDEX CATEGORY FOR COLORECTAL SSI
85
lower risk of SSI after laparoscopic surgery compared with open surgery [4]. The modified N-RIC subtracts a point from the score of patients who undergo laparoscopic surgery, and the score for these procedures now ranges from - 1 to 2 points. For analyses of SSI rates after colonic operations, the NNIS combined categories - 1 and 0 into a single category because there is no significant difference in risk between these categories [10,11]. Prolonged operative time has traditionally been believed to be an important risk factor for SSI, and the length of operation is one of the factors used to determine the N-RIC [1]. However, recent studies did not identify prolonged operative time as a significant risk factor for SSI after colorectal surgery [8,12,13]. This may be related to the increase in laparoscopic surgery over recent years. Although laparoscopic surgery is associated with a longer operative time [14], it is also associated with a lower rate of SSI after colorectal surgery [13,15]. This trend differs from the previous evidence that prolonged operative time is associated with an increased risk of SSI. We therefore used our SSI surveillance data to analyze the relation between operative time and SSI by setting the 75th percentile of the length of operation separately for open and laparoscopic surgery. We also compared the usefulness of the SSI risk index category between the N-RIC model and the study model.
3 d after surgery. For class III (contaminated) and IV (dirty– infected) operations, various types and durations of third-and fourth-generation cephalosporin and carbapenem antibiotics were used, and de-escalation of antibiotic therapy was performed according to the results of gram staining and bacterial culture.
Patients and Methods
The independent variables used in the analyses were the four factors used to determine the N-RIC: (1) ASA score, (2) wound classification, (3) length of operation, and (4) use of a laparoscope. The anesthesiologists determined the ASA score during their pre-operative assessment. The surgical sites were classified as clean (class I), clean–contaminated (class II), contaminated (class III), or dirty–infected (class IV) according to the CDC definitions. All continuous and ordered variables were dichotomized by setting the following cutoff values: ASA score ( £ 2, ‡ 3), surgical wound class (II, III–IV), and length of operation ( £ 75th percentile, > 75th percentile). Correlations between SSI and the factors used to determine the N-RIC were determined by univariate and multivariable analyses. The 75th percentile of the length of operation was determined separately for open and laparoscopic surgery in the study model. The same analyses were performed in the N-RIC model and the study model.
Surveillance methods and materials
Since 2005, we have performed SSI surveillance prospectively following the U.S. Centers for Disease Control and Prevention (CDC) guidelines [3] for all patients undergoing gastrointestinal surgery in the Institute of Gastroenterology at the Nippon Medical School Musashikosugi Hospital (a branch hospital of a university), which has a mixture of gastrointestinal surgeons, including colorectal surgeons. In the present retrospective study, the attending physicians or nurses observed all wounds at least once per day until discharge. Post-discharge SSI surveillance was performed in the outpatient clinic, and all patients were followed up for 30 d post-operatively. The diagnosis of SSI was made after discussion among the attending physicians and nurses, and a dedicated infection control doctor. Using our SSI surveillance database from July 2005 to May 2010, we analyzed retrospectively 538 consecutive patients who underwent abdominal colorectal surgery. All included patients underwent at least one of the following procedures: large bowel resection, large bowel anastomosis, colostomy formation or closure, or simple colorectal perforation closure. Patients who underwent trans-anal or trans-sacral operations, parastomal hernia repairs without ostomy revision, or ileostomy creation or reversal without manipulation of the colon/rectum were excluded. No patients received open wound management because primary skin closure was performed in all patients according to institutional policy. Use of antibiotics
Essentially, for class II (clean–contaminated) operations, prophylactic use of a second-generation cephalosporin was implemented. Administration was started 30 min before skin incision, repeated every 3 h during surgery, and stopped within
Dependent variables
The dependent variable in this study was SSI as defined by the CDC criteria. Specifically, SSI was defined as an infection that occurred at the surgical site within 30 d after surgery and was characterized by at least one of the following: Purulent drainage from the surgical site; an organism isolated from a culture of fluid from the surgical site; incisional pain, tenderness, localized swelling, redness, or heat and opening of the wound; or diagnosis of SSI by the surgeon or attending physician [3]. Surgical site infections were classified as either incisional or organ/space infection. Incisional SSIs were further divided into those involving only the skin and subcutaneous tissues (superficial incisional SSIs) and those involving deeper soft tissues of the incision (deep incisional SSIs). Organ/space SSIs could involve any part of the anatomy other than the incised body wall layers that were opened or manipulated during the operation [3]. Independent variables
Statistical analysis
All statistical analyses were performed using JMP version 10.0.0 statistical software (SAS Institute, Cary, NC). The univariate relation between each independent variable and SSI was evaluated using the Pearson w2 test for categorical variables. Independent variables with a value of p < 0.05 on univariate analysis were included in a multivariable logistic regression model to determine the independent risk factors for SSI. A value of p < 0.05 was considered to be statistically significant. The predictive performances of the N-RIC and study models were assessed by constructing receiver-operating characteristic (ROC) curves and calculating the area under the curve (AUC) using a logistic regression model. Results
A total of 547 consecutive patients who underwent abdominal colorectal surgery were enrolled in SSI surveillance
86
WATANABE ET AL.
during the study period, and 538 of these were included in the final analysis. Nine patients were excluded from the analysis because they died (n = 5) or underwent reoperation (n = 4) within 30 d after surgery, and did not develop SSI during the study period. The mean age of patients was 65.5 – 12.7 y (range, 20–98 y), and 57.0% were male. The preoperative diagnoses included colorectal cancer (87.6%), diverticular disease (2.4%), benign polyps (2.4%), and other conditions (Table 1). The most common type of operation performed was colorectal resection (n = 466, 86.7%), followed by colostomy creation or closure (n = 57, 10.6%), bypass (n = 11, 2.0%), and simple closure of a colon perforation (n = 4, 0.7%). All patients were followed up for 30 d after surgery, except for three patients who developed SSI and died within 30 d. Compliance with the protocol for prophylactic antibiotic administration was analyzed in patients who underwent elective surgery and had class II wounds. After exclusion of patients with class III or IV wounds and patients who underwent emergency surgery, the remaining 454 patients were included in this analysis. The compliance rate for administration of prophylactic antibiotic type was 93.6%, and the compliance rate for duration of prophylactic antibiotic administration was 83.3%. Administration of further doses or of different antibiotic types occurred in 45 of 454 patients (9.9%) because they developed SSI, 16 of 454 patients (3.5%) because they developed remote infection, 11/454 patients (2.4%) because they developed postoperative ileus with symptoms of infection, and 15 of 454 patients (3.3%) because of infection of unknown origin. One patient with penicillin allergy and one with cephem allergy received different antibiotic types. A total of 97 patients (18.0%) were diagnosed with SSI, including 77 (14.3%) with incisional SSI and 20 (3.7%) with
Table 1. Indications for Colorectal Surgery (n = 538) Condition Colorectal cancer Diverticular disease Benign polyps FAP HNPCC Appendicitis with tumor Crohn disease Colon perforation (except for cancer, diverticulum) Post-operative anastomotic leak Colon perforation during colonoscopy with bowel preparation Trans-anal ileus tube Endoscopic dilation Foreign body perforation Unknown cause Bowel obstruction (except for cancer) Colon volvulus Ileocecal invagination Stoma creation Others Appendiceal cancer Acute diverticular bleeding
n
%
471 13 13 3 1 4 2 18
87.6 2.4 2.4 0.6 0.2 0.7 0.4 3.3
(10) ( 4) ( ( ( (
1) 1) 1) 1) 11 ( 4) ( 2) ( 5) 2 ( 1) ( 1)
2.0
0.4
FAP = familial adenomatous polyposis; HNPCC = hereditary nonpolyposis colorectal cancer.
organ/space SSI. Bowel anastomosis was performed in 408 patients, and anastomotic leakage developed in 15 of these patients (3.7%). The overall (incisional) SSI rates were 18.0% (14.3%) in all patients who underwent colorectal surgery, 10.4% (7.8%) in patients who underwent elective colon surgery, 21.6% (15.6%) in patients who underwent elective rectal surgery, and 34.2% (32.9%) in patients who underwent emergency colorectal surgery. Using our SSI surveillance data, the 75th percentile of the length of operation was 238 min in patients who underwent open surgery and 276 min in patients who underwent laparoscopic surgery. By univariate analyses, SSI was associated with a > 75th percentile length of operation in the N-RIC model (OR, 1.64; 95% confidence interval [CI], 1.05–2.56), and this association was stronger in the study model (OR, 2.07; 95% CI, 1.29–3.30) where the 75th percentile was determined by our data (Table 2). Table 3 shows the results of multivariable analysis including the four factors that showed a significant association with SSI on univariate analysis. The length of operation in the study model (OR 2.75; 95% CI 1.66–4.55), ASA score (OR 2.22; 95% CI 1.10–4.34), wound classification (OR 5.29; 95% CI 2.62–10.69), and open surgery (OR 2.21; 95% CI 1.07–5.17) were found to be independent risk factors for SSI. All of the risk index factors in the study model were therefore found to be associated independently with SSI after colorectal surgery. In the N-RIC model, there was a significant difference in SSI risk between the low-risk group (score 0 or 1) and the high-risk group (score 2 or 3; odds ratio [OR], 5.20; 95% CI, 2.44–11.04). This difference in risk was greater in the study model than in the N-RIC model (OR, 8.38; 95% CI, 3.37– 20.83). The AUC for the risk index category was slightly higher in the study model than in the N-RIC model (0.6817 versus 0.6797; Table 4). Figure 1 shows the comparison of SSI rates and risk index score between the N-RIC model and the study model. Discussion
The rate of SSI after colorectal surgery remains high compared with other general surgical procedures [8,16]. The reported rate of SSI after colorectal surgery ranges from 5% to 30% [7,9,13,16–19]. The reasons for this variation in SSI rates are probably multifactorial, depending on patient demographics and institutional differences. It is therefore necessary to compare SSI rates according to the risks of patients treated at each institute. The NHSN (previously NNIS) adjusts the risks by the N-RIC such as length of operation, ASA score, wound classification, and use of laparoscopy to feed back the collected data based on the N-RIC [1,2,4,10,11]. In 2005, in its procedure-associated module, the NHSN inherited the NNIS program for surveillance of health careassociated infection in surgery patients and reorganized all surgical procedures [11]. The NNIS and NHSN criteria differ in terms of the methods used to classify operative procedures, but not in the basic methods of SSI surveillance. Similar to the NNIS, NHSN facilities report their health care-associated infection surveillance data voluntarily for aggregation into a single national database. The N-RIC is a general score used for predicting the risk of SSI across a broad range of operative procedures [4]. However, it may be too simple to predict risk in the various procedures performed in colorectal surgery
RISK INDEX CATEGORY FOR COLORECTAL SSI
87
Table 2. Univariate Analyses for Surgical Site Infection after Colorectal Surgery Risk factor Length of operation, N-RIC model (min) £ 180 > 180 Length of operation, study model (min) £ 238 (open), £ 276 (laparoscopic) > 238 (open), > 276 (laparoscopic) ASA score £2 ‡3 SSI classification £ II ‡ III Open surgery Yes No
n
SSI (%)
276 262
40 (14.5) 57 (21.8)
404 134
61 (15.1) 36 (26.9)
485 53
79 (16.3) 18 (34.0)
496 42
76 (15.3) 21 (50.0)
443 95
89 (20.1) 8 ( 8.4)
OR
95% CI
p
1.64
1.05– 2.56
0.03
2.07
1.29– 3.30
0.002
2.64
1.43– 4.90
0.002
5.53
2.88–10.61
< 0.0001
2.73
1.28– 5.85
0.007
SSI = surgical site infection; OR, odds ratio; CI, confidence interval; N-RIC = National Nosocomial Infections Surveillance risk index category; ASA = American Society of Anesthesiologists.
[6,7]. Napolitano [20] reported that the N-RIC warrants reevaluation in view of new risk factors for SSI in her Surgical Infection Society Presidential Address published in 2010. In 2011, the NHSN added factors that improved the assessment of SSI risk [21]. However, the N-RIC is still used widely in colorectal surgery because of its simple calculation using the length of operation, ASA score, wound classification, and use of laparoscopy [8,9]. We therefore evaluated the validity of the factors used to determine the N-RIC using the surveillance data from colorectal operations performed at our institute. A recent randomized trial reported that laparoscopic colorectal surgery was associated with a longer operative time than open colorectal surgery [14]. However, laparoscopic colorectal surgery is associated with a lower rate of SSI than open colorectal surgery [13,15]. We therefore considered that the 75th percentile of the length of operation should be determined separately for open and laparoscopic surgery. Univariate analyses showed that SSI was more strongly associated with a > 75th percentile length of operation in the
Table 3. Multivariable Analysis for Surgical Site Infection after Colorectal Surgery Risk index factor N-RIC model Length of operation > 75th percentile ASA score ‡ 3 SSI classification ‡ III Open surgery Study model Length of operation > 75th percentile ASA score ‡ 3 SSI classification ‡ III Open surgery
OR
95% CI
p
2.51
1.54– 4.18
0.0002
2.22 5.52 2.68
1.10– 4.34 2.70–11.34 1.29– 6.32
0.03 < 0.0001 0.01
2.75
1.66– 4.55
0.0001
2.22 5.29 2.21
1.10– 4.34 2.62–10.69 1.07– 5.17
0.03 < 0.0001 0.03
SSI = surgical site infection; OR < odds ratio; CI = confidence interval; N-RIC = National Nosocomial Infections Surveillance risk index category; ASA = American Society of Anesthesiologists.
study model (OR, 2.07) than in the N-RIC model (OR, 1.64). Moreover, the four factors used to determine the N-RIC were all significantly associated with SSI on univariate analyses, and were found to be independent risk factors for SSI on multivariable analysis. Similarly, Kurmann et al. [22] reported that the length of operation was a significant risk factor for SSI after laparoscopic sigmoid resection for benign disease on multivariate analysis of 2,571 patients, using the cutoff value of > 240 min that they set themselves. On the basis of these findings, we suggest reviewing the method of setting the 75th percentile of the length of operation, especially in colorectal surgery. The risk of SSI after colorectal surgery depends on factors related to the patient [23], the operation [18], and the surgeon’s compliance with basic principles such as antibiotic prophylaxis, skin preparation, and good surgical technique [24]. The patient factors that have traditionally been considered to have an important influence on SSI include age, diabetes mellitus, ASA score, smoking, steroid use, and obesity [3]. Among these, ASA score is probably one of the most important factors [1]. In this study, multivariable analysis found that ASA score ‡ 3 was an independent risk factor for SSI after colorectal surgery (OR 2.22; 95% CI 1.10–4.34). Some colorectal operative procedures are associated with high rates of SSI, including emergency operations [7,9,17]. It has been reported that the most important risk factor for SSI after emergency colorectal surgery is wound contamination [25]. It therefore appears that wound contamination may be a more significant risk factor for SSI in studies that include a higher proportion of emergency operations. In this study, 14.7% of patients underwent emergency operations, and wound contamination was an independent risk factor for SSI. Patients with a class IV wound may undergo delayed primary closure rather than primary closure, and would then be excluded from the SSI surveillance [3]. This suggests that the influence of wound contamination on SSI may not be always judged accurately. Similarly, Poon et al. [13] reported that the practice of leaving the skin layer of class IV wounds open for daily dressings may have prevented some cases of incisional SSI and reduced the influence of wound contamination on the risk of SSI. In this study, no
88
WATANABE ET AL.
Table 4. Performance Comparison of the Surgical Site Infection Risk Index Category between the National Nosocomial Infections Surveillance Risk Index Category Model and the Study Model after Colorectal Surgery N-RIC model
Risk index Low risk (0 or 1) High risk (2 or 3) Area under the ROC curve
n
SSI (%)
508 30
82 (16.1) 15 (50.0)
OR 1 5.20a 0.6797
Study model 95% CI
n
SSI (%)
Reference 2.44–11.04
517 21
84 (16.3) 13 (61.9)
OR 1 8.38a 0.6817
95% CI Reference 3.37–20.83
a 2
w test, p < 0.05. SSI = surgical site infection; OR, odds ratio; CI, confidence interval; N-RIC = National Nosocomial Infections Surveillance risk index category; ROC = receiver operating characteristic.
patients received open wound management during the research period, and we therefore consider that our data reflect accurately the influence of wound contamination on the risk of SSI. Some studies reported that laparoscopic surgery was associated with a lower rate of SSI after colorectal surgery [11,13], and that open surgery was an independent risk factor for incisional SSI by multivariable analysis [11,23]. Our results also indicate that open surgery is an independent risk factor for SSI compared with laparoscopic surgery (OR 2.21; 95% CI 1.07–5.17). As early as 2001, Gaynes et al. [4] recommended incorporation of the surgical approach into the risk assessment to improve the predictive value of the NNIS risk index for cholecystectomy, appendectomy, and colectomy. At the time of the 2008 NHSN report, the laparoscopic approach had been shown to lower the risk of SSI, but there were insufficient data available to differentiate risk beyond the three basic factors already used to determine the N-RIC [26]. Considering the recent increase in laparoscopic colon surgery, the use of laparoscopy now appears to be as important for assessment of SSI risk as the other three factors. The NHSN and some other researchers have conducted studies to evaluate the accuracy of the N-RIC for predicting the risk of SSI [8,21,27]. According to their reports, the AUCs of the N-RIC model showed relatively low predictive
FIG. 1. Comparison of SSI rates and risk index score between the N-RIC model and the study model after colorectal surgery. SSI = surgical site infection; N-RIC = National Nosocomial Infections Surveillance risk index category.
ability for colorectal surgery (0.59) [8] and digestive surgery (0.63) [27]. We speculate that this result reflects the elective nature of the surgery in the subjects studied, who therefore had a relatively low rate of wound contamination. The AUCs for the N-RIC model published by the NHSN in 2011 are 0.56 for colon surgery and 0.62 for rectal surgery [21]. In the 2011 report, the NHSN changed the N-RIC model to the improved procedure-specific model, resulting in significant improvements in the AUCs to 0.59 for colon surgery and 0.72 for rectal surgery [21]. The AUC for the risk index category was slightly higher in the present study model than in the N-RIC model (0.6817 versus 0.6797). Although the AUC for the study model was lower than that of the new NHSN procedure-specific model for rectal surgery, it was good in comparison with other models. The greatest merit of the N-RIC model is its simplicity, which enables easy calculation of risk with relatively good performance. The AUC in our study model (0.68) was better than the AUC in a previous study using the same N-RIC model [8]. We speculate that this is because our study population included many patients who underwent emergency operations, which is associated with a high risk of wound contamination. Our multivariable analysis showed ORs for SSI of 2.21 to 2.25 for length of operation, ASA score, and open surgery, with the highest OR (5.29) associated with wound classification. These results lead us to speculate that the N-RIC model may be more useful for populations including high levels of wound contamination than for populations who mainly undergo elective surgery. We assume that the N-RIC model will continue to be useful at facilities such as academic medical centers and large community hospitals, which perform relatively high proportions of emergency surgery. We therefore conclude that modifications such as that in our study model are useful. There are several limitations to this study. In 2011, the NHSN introduced a modified risk-adjusted method of predicting the risk of SSI. The new model incorporates the traditional factors used to determine the N-RIC as well as additional data elements currently collected by the NHSN [21]. The new risk index includes variables that were not evaluated in this study, including age, general anesthesia, medical school affiliation, number of hospital beds for colon surgery, and gender for rectal surgery. Other known risk factors that may increase the risk of SSI were also not evaluated, including obesity, smoking, preoperative corticosteroids, prolonged preoperative hospitalization, perioperative
RISK INDEX CATEGORY FOR COLORECTAL SSI
transfusion, intraoperative hypothermia, use of wound protectors, and postoperative glucose control. Conclusions
The N-RIC is sufficiently useful for predicting the risk of SSI after abdominal colorectal surgery. However, it is suggested that the 75th percentile of the length of operation should be set separately for open and laparoscopic surgery. Author Disclosure Statement
No conflicting financial interests exist. References
1. Culver DH, Horan TC, Gaynes RP, et al. Surgical wound infection rates by wound class, operative procedure, and patient risk index. National Nosocomial Infections Surveillance System. Am J Med 1991;91:152S–157S. 2. Horan TC, Culver DH, Gaynes RP, et al. Nosocomial infections in surgical patients in the United States, January 1986–June 1992. National Nosocomial Infections Surveillance (NNIS) System. Infect Control Hosp Epidemiol 1993; 14:73–80. 3. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol 1999;20:250–278. 4. Gaynes RP, Culver DH, Horan TC, et al. Surgical site infection (SSI) rates in the United States, 1992–1998: the National Nosocomial Infections Surveillance System basic SSI risk index. Clin Infect Dis 2001;33:S69–S77. 5. Vandenbroucke-Grauls C, Schultsz C. Surveillance in infection control: Are we making progress? Curr Opin Infect Dis 2002;15:415–419. 6. Schwenk W, Haase O, Neudecker J, et al. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005;3:CD003145. 7. Gervaz P, Bandiera-Clerc C, Buchs NC, et al. Scoring system to predict the risk of surgical-site infection after colorectal resection. Br J Surg 2012;99:589–595. 8. Pastor C, Baek JH, Varma MG, et al. Validation of the risk index category as a predictor of surgical site infection in elective colorectal surgery. Dis Colon Rectum 2010;53:721– 727. 9. Ho VP, Stein SL, Trencheva K, et al. Differing risk factors for incisional and organ/space surgical site infections following abdominal colorectal surgery. Dis Colon Rectum 2011;54:818–825. 10. National Nosocomial Infections Surveillance System. National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004;32:470–485. 11. Edwards JR, Peterson KD, Andrus ML, et al. National Healthcare Safety Network (NHSN) Report, data summary for 2006 through 2007, issued November 2008. Am J Infect Control 2008;36:609–626. 12. Konishi T, Watanabe T, Kishimoto J, et al. Elective colon and rectal surgery differ in risk factors for wound infection: Results of prospective surveillance. Ann Surg 2006;244:758–763. 13. Poon JT, Law WL, Wong IW, et al. Impact of laparoscopic colorectal resection on surgical site infection. Ann Surg 2009;249:77–81.
89
14. Gervaz P, Inan I, Perneger T, et al. A prospective, randomized, single-blind comparison of laparoscopic versus open sigmoid colectomy for diverticulitis. Ann Surg 2010; 252:3–8. 15. Kiran RP, El-Gazzaz GH, Vogel JD, et al. Laparoscopic approach significantly reduces surgical site infections after colorectal surgery: Data from national surgical quality improvement program. J Am Coll Surg 2010;211:232–238. 16. Kwaan MR, Sirany AM, Rothenberger DA, et al. Abdominal wall thickness: Is it associated with superficial and deep incisional surgical site infection after colorectal surgery. Surg Infect 2013;14:363–368. 17. Blumetti J, Luu M, Sarosi G, et al. Surgical site infections after colorectal surgery: Do risk factors vary depending on the type of infection considered? Surgery 2007;142:704–711. 18. Tang R, Chen HH, Wang YL, et al. Risk factors for surgical site infection after elective resection of the colon and rectum: A single-center prospective study of 2,809 consecutive patients. Ann Surg 2001;234:181–189. 19. Acı´n-Ga´ndara D, Rodrı´guez-Caravaca G, Dura´n-Poveda M, et al. Incidence of surgical site infection in colon surgery: Comparison with regional, national Spanish, and United States standards. Surg Infect 2013;14:339–344. 20. Napolitano LM. Perspectives in surgical infections: What dose the future hold? Surg Infect 2010;11:111–123. 21. Mu Y, Edwards JR, Horan TC, et al. Improving risk-adjusted measures of surgical site infection for the national healthcare safety network. Infect Control Hosp Epidemiol 2011;32:970–986. 22. Kurmann A, Vorburger SA, Candinas D, et al. Operation time and body mass index are significant risk factors for surgical site infection in laparoscopic sigmoid resection: a multicenter study. Surg Endosc 2011;25:3531–3534. 23. Sehgal R, Berg A, Figueroa R, et al. Risk factors for surgical site infections after colorectal resection in diabetic patients. J Am Coll Surg 2011;212:29–34. 24. Murray BW, Huerta S, Dineen S, et al. Surgical site infection in colorectal surgery: A review of the nonpharmacologic tools of prevention. J Am Coll Surg 2010;211:812–822. 25. Watanabe M, Suzuki H, Nomura S, et al. Risk factors for incisional surgical site infection in emergency colorectal surgery: A retrospective analysis. Surg Infect 2014;15:256– 261. 26. Nakamura T, Mitomi H, Ihara A, et al. Risk factors for wound infection after surgery for colorectal cancer. World J Surg 2008;32:1138–1141. 27. Prospero E, Barbadoro P, Annino I, et al. Surgical site infections: Might a longer locally defined T time affect the benchmarking? Am J Infect Control 2007;35:582–584.
Address correspondence to: Dr. Masanori Watanabe Institute of Gastroenterology Nippon Medical School Musashikosugi Hospital 1-396 Kosugicho Nakahara, Kawasaki, Kanagawa 211-8533 Japan E-mail: [email protected]
Performance Assessment of the Risk Index Category for Surgical Site Infection after Colorectal Surgery Masanori Watanabe,1 Hideyuki Suzuki,1 Satoshi Nomura,1 Hidetsugu Hanawa,1 Naoto Chihara,1 Satoshi Mizutani,1 Masanori Yoshino,1 and Eiji Uchida 2
Abstract
Background: The traditional National Healthcare Safety Network (previously National Nosocomial Infections Surveillance) risk index is used to predict the risk of surgical site infection across many operative procedures. However, this index may be too simple to predict risk in the various procedures performed in colorectal surgery. The aim of this study was to evaluate the usefulness of the risk index by analyzing the impact of the risk index factors on surgical site infection after abdominal colorectal surgery. Methods: Using our surgical site infection surveillance database, we analyzed retrospectively 538 consecutive patients who underwent abdominal colorectal surgery between 2005 and 2010. Correlations between surgical site infection and the following risk index factors were analyzed: length of operation, American Society of Anesthesiologists score, wound classification, and use of laparoscopy. The 75th percentile for length of operation was determined separately for open and laparoscopic surgery in the study model. Results: Univariate analyses showed that surgical site infection was more strongly associated with a > 75th percentile length of operation in the study model (odds ratio [OR], 2.07) than in the traditional risk index model (OR, 1.64). Multivariable analysis found that surgical site infection was independently associated with a > 75th percentile length of operation in the study model (OR, 2.75; 95% confidence interval [CI], 1.66–4.55), American Society of Anesthesiologists score ‡ 3 (OR, 2.22; 95% CI, 1.10–4.34), wound classification ‡ III (OR, 5.29; 95% CI, 2.62–10.69), and open surgery (OR, 2.21; 95% CI, 1.07–5.17). Performance of the risk index category was improved in the study model compared with the traditional model. Conclusions: The risk index category is sufficiently useful for predicting the risk of surgical site infection after abdominal colorectal surgery. However, the 75th percentile length of operation should be set separately for open and laparoscopic surgery.
T
The N-RIC is a widely used score that is operation-specific and is based on surveillance data collected prospectively [1– 3,8,9]. The index ranges from 0 to 3 points and is determined by three independent variables weighted equally. One point is scored for each of the following when present: (1) American Society of Anesthesiologists (ASA) physical status score > 2; (2) wound classified as contaminated or dirty–infected; and (3) operative time longer than the approximate 75th percentile for the specific operation being performed (3 h for colon surgery) [2,3]. In 2001, the N-RIC was modified to introduce laparoscopic surgery as an additional factor in a number of surgical procedures including cholecystectomy, colon surgery, gastric surgery, and appendectomy, to account for the
he National Healthcare Safety Network (NHSN), previously the National Nosocomial Infections Surveillance (NNIS), risk index category (N-RIC) was developed in the early 1990s and has been used to predict the risk of surgical site infection (SSI) across a broad range of operative procedures [1–4]. However, the N-RIC provides a generic score that has been criticized for being overly simplistic [5], and does not consider some aspects of colorectal surgery, such as stoma formation or the use of laparoscopy [6]. Gervaz et al. [7] reported that generic scores such as the N-RIC are not used by colorectal surgeons. We therefore considered it necessary to evaluate whether the N-RIC is useful for predicting SSI after colorectal surgery.
1 2
Institute of Gastroenterology, Nippon Medical School Musashikosugi Hospital, Kawasaki, Kanagawa, Japan. Department of Surgery, Nippon Medical School, Tokyo, Japan.
84
RISK INDEX CATEGORY FOR COLORECTAL SSI
85
lower risk of SSI after laparoscopic surgery compared with open surgery [4]. The modified N-RIC subtracts a point from the score of patients who undergo laparoscopic surgery, and the score for these procedures now ranges from - 1 to 2 points. For analyses of SSI rates after colonic operations, the NNIS combined categories - 1 and 0 into a single category because there is no significant difference in risk between these categories [10,11]. Prolonged operative time has traditionally been believed to be an important risk factor for SSI, and the length of operation is one of the factors used to determine the N-RIC [1]. However, recent studies did not identify prolonged operative time as a significant risk factor for SSI after colorectal surgery [8,12,13]. This may be related to the increase in laparoscopic surgery over recent years. Although laparoscopic surgery is associated with a longer operative time [14], it is also associated with a lower rate of SSI after colorectal surgery [13,15]. This trend differs from the previous evidence that prolonged operative time is associated with an increased risk of SSI. We therefore used our SSI surveillance data to analyze the relation between operative time and SSI by setting the 75th percentile of the length of operation separately for open and laparoscopic surgery. We also compared the usefulness of the SSI risk index category between the N-RIC model and the study model.
3 d after surgery. For class III (contaminated) and IV (dirty– infected) operations, various types and durations of third-and fourth-generation cephalosporin and carbapenem antibiotics were used, and de-escalation of antibiotic therapy was performed according to the results of gram staining and bacterial culture.
Patients and Methods
The independent variables used in the analyses were the four factors used to determine the N-RIC: (1) ASA score, (2) wound classification, (3) length of operation, and (4) use of a laparoscope. The anesthesiologists determined the ASA score during their pre-operative assessment. The surgical sites were classified as clean (class I), clean–contaminated (class II), contaminated (class III), or dirty–infected (class IV) according to the CDC definitions. All continuous and ordered variables were dichotomized by setting the following cutoff values: ASA score ( £ 2, ‡ 3), surgical wound class (II, III–IV), and length of operation ( £ 75th percentile, > 75th percentile). Correlations between SSI and the factors used to determine the N-RIC were determined by univariate and multivariable analyses. The 75th percentile of the length of operation was determined separately for open and laparoscopic surgery in the study model. The same analyses were performed in the N-RIC model and the study model.
Surveillance methods and materials
Since 2005, we have performed SSI surveillance prospectively following the U.S. Centers for Disease Control and Prevention (CDC) guidelines [3] for all patients undergoing gastrointestinal surgery in the Institute of Gastroenterology at the Nippon Medical School Musashikosugi Hospital (a branch hospital of a university), which has a mixture of gastrointestinal surgeons, including colorectal surgeons. In the present retrospective study, the attending physicians or nurses observed all wounds at least once per day until discharge. Post-discharge SSI surveillance was performed in the outpatient clinic, and all patients were followed up for 30 d post-operatively. The diagnosis of SSI was made after discussion among the attending physicians and nurses, and a dedicated infection control doctor. Using our SSI surveillance database from July 2005 to May 2010, we analyzed retrospectively 538 consecutive patients who underwent abdominal colorectal surgery. All included patients underwent at least one of the following procedures: large bowel resection, large bowel anastomosis, colostomy formation or closure, or simple colorectal perforation closure. Patients who underwent trans-anal or trans-sacral operations, parastomal hernia repairs without ostomy revision, or ileostomy creation or reversal without manipulation of the colon/rectum were excluded. No patients received open wound management because primary skin closure was performed in all patients according to institutional policy. Use of antibiotics
Essentially, for class II (clean–contaminated) operations, prophylactic use of a second-generation cephalosporin was implemented. Administration was started 30 min before skin incision, repeated every 3 h during surgery, and stopped within
Dependent variables
The dependent variable in this study was SSI as defined by the CDC criteria. Specifically, SSI was defined as an infection that occurred at the surgical site within 30 d after surgery and was characterized by at least one of the following: Purulent drainage from the surgical site; an organism isolated from a culture of fluid from the surgical site; incisional pain, tenderness, localized swelling, redness, or heat and opening of the wound; or diagnosis of SSI by the surgeon or attending physician [3]. Surgical site infections were classified as either incisional or organ/space infection. Incisional SSIs were further divided into those involving only the skin and subcutaneous tissues (superficial incisional SSIs) and those involving deeper soft tissues of the incision (deep incisional SSIs). Organ/space SSIs could involve any part of the anatomy other than the incised body wall layers that were opened or manipulated during the operation [3]. Independent variables
Statistical analysis
All statistical analyses were performed using JMP version 10.0.0 statistical software (SAS Institute, Cary, NC). The univariate relation between each independent variable and SSI was evaluated using the Pearson w2 test for categorical variables. Independent variables with a value of p < 0.05 on univariate analysis were included in a multivariable logistic regression model to determine the independent risk factors for SSI. A value of p < 0.05 was considered to be statistically significant. The predictive performances of the N-RIC and study models were assessed by constructing receiver-operating characteristic (ROC) curves and calculating the area under the curve (AUC) using a logistic regression model. Results
A total of 547 consecutive patients who underwent abdominal colorectal surgery were enrolled in SSI surveillance
86
WATANABE ET AL.
during the study period, and 538 of these were included in the final analysis. Nine patients were excluded from the analysis because they died (n = 5) or underwent reoperation (n = 4) within 30 d after surgery, and did not develop SSI during the study period. The mean age of patients was 65.5 – 12.7 y (range, 20–98 y), and 57.0% were male. The preoperative diagnoses included colorectal cancer (87.6%), diverticular disease (2.4%), benign polyps (2.4%), and other conditions (Table 1). The most common type of operation performed was colorectal resection (n = 466, 86.7%), followed by colostomy creation or closure (n = 57, 10.6%), bypass (n = 11, 2.0%), and simple closure of a colon perforation (n = 4, 0.7%). All patients were followed up for 30 d after surgery, except for three patients who developed SSI and died within 30 d. Compliance with the protocol for prophylactic antibiotic administration was analyzed in patients who underwent elective surgery and had class II wounds. After exclusion of patients with class III or IV wounds and patients who underwent emergency surgery, the remaining 454 patients were included in this analysis. The compliance rate for administration of prophylactic antibiotic type was 93.6%, and the compliance rate for duration of prophylactic antibiotic administration was 83.3%. Administration of further doses or of different antibiotic types occurred in 45 of 454 patients (9.9%) because they developed SSI, 16 of 454 patients (3.5%) because they developed remote infection, 11/454 patients (2.4%) because they developed postoperative ileus with symptoms of infection, and 15 of 454 patients (3.3%) because of infection of unknown origin. One patient with penicillin allergy and one with cephem allergy received different antibiotic types. A total of 97 patients (18.0%) were diagnosed with SSI, including 77 (14.3%) with incisional SSI and 20 (3.7%) with
Table 1. Indications for Colorectal Surgery (n = 538) Condition Colorectal cancer Diverticular disease Benign polyps FAP HNPCC Appendicitis with tumor Crohn disease Colon perforation (except for cancer, diverticulum) Post-operative anastomotic leak Colon perforation during colonoscopy with bowel preparation Trans-anal ileus tube Endoscopic dilation Foreign body perforation Unknown cause Bowel obstruction (except for cancer) Colon volvulus Ileocecal invagination Stoma creation Others Appendiceal cancer Acute diverticular bleeding
n
%
471 13 13 3 1 4 2 18
87.6 2.4 2.4 0.6 0.2 0.7 0.4 3.3
(10) ( 4) ( ( ( (
1) 1) 1) 1) 11 ( 4) ( 2) ( 5) 2 ( 1) ( 1)
2.0
0.4
FAP = familial adenomatous polyposis; HNPCC = hereditary nonpolyposis colorectal cancer.
organ/space SSI. Bowel anastomosis was performed in 408 patients, and anastomotic leakage developed in 15 of these patients (3.7%). The overall (incisional) SSI rates were 18.0% (14.3%) in all patients who underwent colorectal surgery, 10.4% (7.8%) in patients who underwent elective colon surgery, 21.6% (15.6%) in patients who underwent elective rectal surgery, and 34.2% (32.9%) in patients who underwent emergency colorectal surgery. Using our SSI surveillance data, the 75th percentile of the length of operation was 238 min in patients who underwent open surgery and 276 min in patients who underwent laparoscopic surgery. By univariate analyses, SSI was associated with a > 75th percentile length of operation in the N-RIC model (OR, 1.64; 95% confidence interval [CI], 1.05–2.56), and this association was stronger in the study model (OR, 2.07; 95% CI, 1.29–3.30) where the 75th percentile was determined by our data (Table 2). Table 3 shows the results of multivariable analysis including the four factors that showed a significant association with SSI on univariate analysis. The length of operation in the study model (OR 2.75; 95% CI 1.66–4.55), ASA score (OR 2.22; 95% CI 1.10–4.34), wound classification (OR 5.29; 95% CI 2.62–10.69), and open surgery (OR 2.21; 95% CI 1.07–5.17) were found to be independent risk factors for SSI. All of the risk index factors in the study model were therefore found to be associated independently with SSI after colorectal surgery. In the N-RIC model, there was a significant difference in SSI risk between the low-risk group (score 0 or 1) and the high-risk group (score 2 or 3; odds ratio [OR], 5.20; 95% CI, 2.44–11.04). This difference in risk was greater in the study model than in the N-RIC model (OR, 8.38; 95% CI, 3.37– 20.83). The AUC for the risk index category was slightly higher in the study model than in the N-RIC model (0.6817 versus 0.6797; Table 4). Figure 1 shows the comparison of SSI rates and risk index score between the N-RIC model and the study model. Discussion
The rate of SSI after colorectal surgery remains high compared with other general surgical procedures [8,16]. The reported rate of SSI after colorectal surgery ranges from 5% to 30% [7,9,13,16–19]. The reasons for this variation in SSI rates are probably multifactorial, depending on patient demographics and institutional differences. It is therefore necessary to compare SSI rates according to the risks of patients treated at each institute. The NHSN (previously NNIS) adjusts the risks by the N-RIC such as length of operation, ASA score, wound classification, and use of laparoscopy to feed back the collected data based on the N-RIC [1,2,4,10,11]. In 2005, in its procedure-associated module, the NHSN inherited the NNIS program for surveillance of health careassociated infection in surgery patients and reorganized all surgical procedures [11]. The NNIS and NHSN criteria differ in terms of the methods used to classify operative procedures, but not in the basic methods of SSI surveillance. Similar to the NNIS, NHSN facilities report their health care-associated infection surveillance data voluntarily for aggregation into a single national database. The N-RIC is a general score used for predicting the risk of SSI across a broad range of operative procedures [4]. However, it may be too simple to predict risk in the various procedures performed in colorectal surgery
RISK INDEX CATEGORY FOR COLORECTAL SSI
87
Table 2. Univariate Analyses for Surgical Site Infection after Colorectal Surgery Risk factor Length of operation, N-RIC model (min) £ 180 > 180 Length of operation, study model (min) £ 238 (open), £ 276 (laparoscopic) > 238 (open), > 276 (laparoscopic) ASA score £2 ‡3 SSI classification £ II ‡ III Open surgery Yes No
n
SSI (%)
276 262
40 (14.5) 57 (21.8)
404 134
61 (15.1) 36 (26.9)
485 53
79 (16.3) 18 (34.0)
496 42
76 (15.3) 21 (50.0)
443 95
89 (20.1) 8 ( 8.4)
OR
95% CI
p
1.64
1.05– 2.56
0.03
2.07
1.29– 3.30
0.002
2.64
1.43– 4.90
0.002
5.53
2.88–10.61
< 0.0001
2.73
1.28– 5.85
0.007
SSI = surgical site infection; OR, odds ratio; CI, confidence interval; N-RIC = National Nosocomial Infections Surveillance risk index category; ASA = American Society of Anesthesiologists.
[6,7]. Napolitano [20] reported that the N-RIC warrants reevaluation in view of new risk factors for SSI in her Surgical Infection Society Presidential Address published in 2010. In 2011, the NHSN added factors that improved the assessment of SSI risk [21]. However, the N-RIC is still used widely in colorectal surgery because of its simple calculation using the length of operation, ASA score, wound classification, and use of laparoscopy [8,9]. We therefore evaluated the validity of the factors used to determine the N-RIC using the surveillance data from colorectal operations performed at our institute. A recent randomized trial reported that laparoscopic colorectal surgery was associated with a longer operative time than open colorectal surgery [14]. However, laparoscopic colorectal surgery is associated with a lower rate of SSI than open colorectal surgery [13,15]. We therefore considered that the 75th percentile of the length of operation should be determined separately for open and laparoscopic surgery. Univariate analyses showed that SSI was more strongly associated with a > 75th percentile length of operation in the
Table 3. Multivariable Analysis for Surgical Site Infection after Colorectal Surgery Risk index factor N-RIC model Length of operation > 75th percentile ASA score ‡ 3 SSI classification ‡ III Open surgery Study model Length of operation > 75th percentile ASA score ‡ 3 SSI classification ‡ III Open surgery
OR
95% CI
p
2.51
1.54– 4.18
0.0002
2.22 5.52 2.68
1.10– 4.34 2.70–11.34 1.29– 6.32
0.03 < 0.0001 0.01
2.75
1.66– 4.55
0.0001
2.22 5.29 2.21
1.10– 4.34 2.62–10.69 1.07– 5.17
0.03 < 0.0001 0.03
SSI = surgical site infection; OR < odds ratio; CI = confidence interval; N-RIC = National Nosocomial Infections Surveillance risk index category; ASA = American Society of Anesthesiologists.
study model (OR, 2.07) than in the N-RIC model (OR, 1.64). Moreover, the four factors used to determine the N-RIC were all significantly associated with SSI on univariate analyses, and were found to be independent risk factors for SSI on multivariable analysis. Similarly, Kurmann et al. [22] reported that the length of operation was a significant risk factor for SSI after laparoscopic sigmoid resection for benign disease on multivariate analysis of 2,571 patients, using the cutoff value of > 240 min that they set themselves. On the basis of these findings, we suggest reviewing the method of setting the 75th percentile of the length of operation, especially in colorectal surgery. The risk of SSI after colorectal surgery depends on factors related to the patient [23], the operation [18], and the surgeon’s compliance with basic principles such as antibiotic prophylaxis, skin preparation, and good surgical technique [24]. The patient factors that have traditionally been considered to have an important influence on SSI include age, diabetes mellitus, ASA score, smoking, steroid use, and obesity [3]. Among these, ASA score is probably one of the most important factors [1]. In this study, multivariable analysis found that ASA score ‡ 3 was an independent risk factor for SSI after colorectal surgery (OR 2.22; 95% CI 1.10–4.34). Some colorectal operative procedures are associated with high rates of SSI, including emergency operations [7,9,17]. It has been reported that the most important risk factor for SSI after emergency colorectal surgery is wound contamination [25]. It therefore appears that wound contamination may be a more significant risk factor for SSI in studies that include a higher proportion of emergency operations. In this study, 14.7% of patients underwent emergency operations, and wound contamination was an independent risk factor for SSI. Patients with a class IV wound may undergo delayed primary closure rather than primary closure, and would then be excluded from the SSI surveillance [3]. This suggests that the influence of wound contamination on SSI may not be always judged accurately. Similarly, Poon et al. [13] reported that the practice of leaving the skin layer of class IV wounds open for daily dressings may have prevented some cases of incisional SSI and reduced the influence of wound contamination on the risk of SSI. In this study, no
88
WATANABE ET AL.
Table 4. Performance Comparison of the Surgical Site Infection Risk Index Category between the National Nosocomial Infections Surveillance Risk Index Category Model and the Study Model after Colorectal Surgery N-RIC model
Risk index Low risk (0 or 1) High risk (2 or 3) Area under the ROC curve
n
SSI (%)
508 30
82 (16.1) 15 (50.0)
OR 1 5.20a 0.6797
Study model 95% CI
n
SSI (%)
Reference 2.44–11.04
517 21
84 (16.3) 13 (61.9)
OR 1 8.38a 0.6817
95% CI Reference 3.37–20.83
a 2
w test, p < 0.05. SSI = surgical site infection; OR, odds ratio; CI, confidence interval; N-RIC = National Nosocomial Infections Surveillance risk index category; ROC = receiver operating characteristic.
patients received open wound management during the research period, and we therefore consider that our data reflect accurately the influence of wound contamination on the risk of SSI. Some studies reported that laparoscopic surgery was associated with a lower rate of SSI after colorectal surgery [11,13], and that open surgery was an independent risk factor for incisional SSI by multivariable analysis [11,23]. Our results also indicate that open surgery is an independent risk factor for SSI compared with laparoscopic surgery (OR 2.21; 95% CI 1.07–5.17). As early as 2001, Gaynes et al. [4] recommended incorporation of the surgical approach into the risk assessment to improve the predictive value of the NNIS risk index for cholecystectomy, appendectomy, and colectomy. At the time of the 2008 NHSN report, the laparoscopic approach had been shown to lower the risk of SSI, but there were insufficient data available to differentiate risk beyond the three basic factors already used to determine the N-RIC [26]. Considering the recent increase in laparoscopic colon surgery, the use of laparoscopy now appears to be as important for assessment of SSI risk as the other three factors. The NHSN and some other researchers have conducted studies to evaluate the accuracy of the N-RIC for predicting the risk of SSI [8,21,27]. According to their reports, the AUCs of the N-RIC model showed relatively low predictive
FIG. 1. Comparison of SSI rates and risk index score between the N-RIC model and the study model after colorectal surgery. SSI = surgical site infection; N-RIC = National Nosocomial Infections Surveillance risk index category.
ability for colorectal surgery (0.59) [8] and digestive surgery (0.63) [27]. We speculate that this result reflects the elective nature of the surgery in the subjects studied, who therefore had a relatively low rate of wound contamination. The AUCs for the N-RIC model published by the NHSN in 2011 are 0.56 for colon surgery and 0.62 for rectal surgery [21]. In the 2011 report, the NHSN changed the N-RIC model to the improved procedure-specific model, resulting in significant improvements in the AUCs to 0.59 for colon surgery and 0.72 for rectal surgery [21]. The AUC for the risk index category was slightly higher in the present study model than in the N-RIC model (0.6817 versus 0.6797). Although the AUC for the study model was lower than that of the new NHSN procedure-specific model for rectal surgery, it was good in comparison with other models. The greatest merit of the N-RIC model is its simplicity, which enables easy calculation of risk with relatively good performance. The AUC in our study model (0.68) was better than the AUC in a previous study using the same N-RIC model [8]. We speculate that this is because our study population included many patients who underwent emergency operations, which is associated with a high risk of wound contamination. Our multivariable analysis showed ORs for SSI of 2.21 to 2.25 for length of operation, ASA score, and open surgery, with the highest OR (5.29) associated with wound classification. These results lead us to speculate that the N-RIC model may be more useful for populations including high levels of wound contamination than for populations who mainly undergo elective surgery. We assume that the N-RIC model will continue to be useful at facilities such as academic medical centers and large community hospitals, which perform relatively high proportions of emergency surgery. We therefore conclude that modifications such as that in our study model are useful. There are several limitations to this study. In 2011, the NHSN introduced a modified risk-adjusted method of predicting the risk of SSI. The new model incorporates the traditional factors used to determine the N-RIC as well as additional data elements currently collected by the NHSN [21]. The new risk index includes variables that were not evaluated in this study, including age, general anesthesia, medical school affiliation, number of hospital beds for colon surgery, and gender for rectal surgery. Other known risk factors that may increase the risk of SSI were also not evaluated, including obesity, smoking, preoperative corticosteroids, prolonged preoperative hospitalization, perioperative
RISK INDEX CATEGORY FOR COLORECTAL SSI
transfusion, intraoperative hypothermia, use of wound protectors, and postoperative glucose control. Conclusions
The N-RIC is sufficiently useful for predicting the risk of SSI after abdominal colorectal surgery. However, it is suggested that the 75th percentile of the length of operation should be set separately for open and laparoscopic surgery. Author Disclosure Statement
No conflicting financial interests exist. References
1. Culver DH, Horan TC, Gaynes RP, et al. Surgical wound infection rates by wound class, operative procedure, and patient risk index. National Nosocomial Infections Surveillance System. Am J Med 1991;91:152S–157S. 2. Horan TC, Culver DH, Gaynes RP, et al. Nosocomial infections in surgical patients in the United States, January 1986–June 1992. National Nosocomial Infections Surveillance (NNIS) System. Infect Control Hosp Epidemiol 1993; 14:73–80. 3. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol 1999;20:250–278. 4. Gaynes RP, Culver DH, Horan TC, et al. Surgical site infection (SSI) rates in the United States, 1992–1998: the National Nosocomial Infections Surveillance System basic SSI risk index. Clin Infect Dis 2001;33:S69–S77. 5. Vandenbroucke-Grauls C, Schultsz C. Surveillance in infection control: Are we making progress? Curr Opin Infect Dis 2002;15:415–419. 6. Schwenk W, Haase O, Neudecker J, et al. Short term benefits for laparoscopic colorectal resection. Cochrane Database Syst Rev 2005;3:CD003145. 7. Gervaz P, Bandiera-Clerc C, Buchs NC, et al. Scoring system to predict the risk of surgical-site infection after colorectal resection. Br J Surg 2012;99:589–595. 8. Pastor C, Baek JH, Varma MG, et al. Validation of the risk index category as a predictor of surgical site infection in elective colorectal surgery. Dis Colon Rectum 2010;53:721– 727. 9. Ho VP, Stein SL, Trencheva K, et al. Differing risk factors for incisional and organ/space surgical site infections following abdominal colorectal surgery. Dis Colon Rectum 2011;54:818–825. 10. National Nosocomial Infections Surveillance System. National Nosocomial Infections Surveillance (NNIS) System Report, data summary from January 1992 through June 2004, issued October 2004. Am J Infect Control 2004;32:470–485. 11. Edwards JR, Peterson KD, Andrus ML, et al. National Healthcare Safety Network (NHSN) Report, data summary for 2006 through 2007, issued November 2008. Am J Infect Control 2008;36:609–626. 12. Konishi T, Watanabe T, Kishimoto J, et al. Elective colon and rectal surgery differ in risk factors for wound infection: Results of prospective surveillance. Ann Surg 2006;244:758–763. 13. Poon JT, Law WL, Wong IW, et al. Impact of laparoscopic colorectal resection on surgical site infection. Ann Surg 2009;249:77–81.
89
14. Gervaz P, Inan I, Perneger T, et al. A prospective, randomized, single-blind comparison of laparoscopic versus open sigmoid colectomy for diverticulitis. Ann Surg 2010; 252:3–8. 15. Kiran RP, El-Gazzaz GH, Vogel JD, et al. Laparoscopic approach significantly reduces surgical site infections after colorectal surgery: Data from national surgical quality improvement program. J Am Coll Surg 2010;211:232–238. 16. Kwaan MR, Sirany AM, Rothenberger DA, et al. Abdominal wall thickness: Is it associated with superficial and deep incisional surgical site infection after colorectal surgery. Surg Infect 2013;14:363–368. 17. Blumetti J, Luu M, Sarosi G, et al. Surgical site infections after colorectal surgery: Do risk factors vary depending on the type of infection considered? Surgery 2007;142:704–711. 18. Tang R, Chen HH, Wang YL, et al. Risk factors for surgical site infection after elective resection of the colon and rectum: A single-center prospective study of 2,809 consecutive patients. Ann Surg 2001;234:181–189. 19. Acı´n-Ga´ndara D, Rodrı´guez-Caravaca G, Dura´n-Poveda M, et al. Incidence of surgical site infection in colon surgery: Comparison with regional, national Spanish, and United States standards. Surg Infect 2013;14:339–344. 20. Napolitano LM. Perspectives in surgical infections: What dose the future hold? Surg Infect 2010;11:111–123. 21. Mu Y, Edwards JR, Horan TC, et al. Improving risk-adjusted measures of surgical site infection for the national healthcare safety network. Infect Control Hosp Epidemiol 2011;32:970–986. 22. Kurmann A, Vorburger SA, Candinas D, et al. Operation time and body mass index are significant risk factors for surgical site infection in laparoscopic sigmoid resection: a multicenter study. Surg Endosc 2011;25:3531–3534. 23. Sehgal R, Berg A, Figueroa R, et al. Risk factors for surgical site infections after colorectal resection in diabetic patients. J Am Coll Surg 2011;212:29–34. 24. Murray BW, Huerta S, Dineen S, et al. Surgical site infection in colorectal surgery: A review of the nonpharmacologic tools of prevention. J Am Coll Surg 2010;211:812–822. 25. Watanabe M, Suzuki H, Nomura S, et al. Risk factors for incisional surgical site infection in emergency colorectal surgery: A retrospective analysis. Surg Infect 2014;15:256– 261. 26. Nakamura T, Mitomi H, Ihara A, et al. Risk factors for wound infection after surgery for colorectal cancer. World J Surg 2008;32:1138–1141. 27. Prospero E, Barbadoro P, Annino I, et al. Surgical site infections: Might a longer locally defined T time affect the benchmarking? Am J Infect Control 2007;35:582–584.
Address correspondence to: Dr. Masanori Watanabe Institute of Gastroenterology Nippon Medical School Musashikosugi Hospital 1-396 Kosugicho Nakahara, Kawasaki, Kanagawa 211-8533 Japan E-mail: [email protected]
