Clinical relevance and effect of surgical wound classification in appendicitis

j o u r n a l o f s u r g i c a l r e s e a r c h  j u l y 2 0 1 7 ( 2 1 5 ) 1 3 2 e1 3 9

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.JournalofSurgicalResearch.com

Clinical relevance and effect of surgical wound classification in appendicitis Retrospective evaluation of wound classification discrepancies between surgeons, Swissnoso-trained infection control nurse, and histology as well as surgical site infection rates by wound class Anastasija Wang-Chan,a,* Christian Gingert, MD,b,c Eliane Angst, PD,d,e and Franc Heinrich Hetzer, PhDa,f a

Department of Surgery and Orthopedics, Hospital Linth, Uznach, Switzerland Department of Visceral and Thoracic Surgery, Cantonal Hospital Winterthur, Winterthur, Switzerland c Faculty of Health, Department of Medicine, University of Witten/Herdecke, Herdecke, Germany d Department of Surgery and Orthopedics, Cantonal Hospital Schaffhausen, Schaffhausen, Switzerland e Department of Visceral Surgery and Medicine, Inselspital, University of Bern, Bern, Switzerland f Faculty of Medicine, University of Zurich, Zurich, Switzerland b

article info

abstract

Article history:

Background: Surgical wound classification (SWC) is used for risk stratification of surgical site

Received 7 September 2016

infection (SSI) and serves as the basis for measuring quality of care. The objective was to

Received in revised form

examine the accuracy and reliability of SWC. This study was purposed to evaluate the

14 March 2017

discrepancies in SWC as assessed by three groups: surgeons, an infection control nurse,

Accepted 24 March 2017

and histopathologic evaluation. The secondary aim was to compare the risk-stratified SSI

Available online 1 April 2017

rates using the different SWC methods for 30 d postoperatively. Methods: An analysis was performed of the appendectomies from January 2013 to June 2014 in the Cantonal Hospital of Schaffhausen. SWC was assigned by the operating surgeon at the end of the procedure and retrospectively reviewed by a Swissnoso-trained infection control nurse after reading the operative and pathology report. The level of agreement among the three different SWC assessment groups was determined using kappa statistic. SSI rates were analyzed using a chi-square test. Results: In 246 evaluated cases, the kappa scores for interrater reliability among the SWC assessments across the three groups ranged from 0.05 to 0.2 signifying slight agreement between the groups. SSIs were more frequently associated with trained infection control nurseeassigned SWC than with surgeons based SWC.

The study was approved by the Ethics Committee of Zurich, Switzerland, and informed consent was obtained from all patients. * Corresponding author. Department of Surgery and Orthopedics, Hospital Linth, Gasterstrasse 25, CH-8730 Uznach, Switzerland. Tel.: þ41552855359. E-mail address: [email protected] (A. Wang-Chan). 0022-4804/ª 2017 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). http://dx.doi.org/10.1016/j.jss.2017.03.034

wang-chan et al  perioperative characteristics of cases

133

Conclusions: Our study demonstrated a considerable discordance in the SWC assessments performed by the three groups. Unfortunately, the currently practiced SWC system suffers from ambiguity in definition and/or implementation of these definitions is not clearly stated. This lack of reliability is problematic and may lead to inappropriate comparisons within and between hospitals and surgeons. ª 2017 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Introduction Despite advances in surgical techniques and antibiotic prophylaxis, surgical site infections (SSIs) continue to be a major cause of patient morbidity.1,2 Since 1973, ongoing surveillance of postoperative wound infection rates has been advocated to decrease the incidence of SSIs.2,3 The surgical wound classification (SWC) system was first introduced in 1964 to describe the degree of microbial contamination present at the time of surgery, operative cases are divided into four categories: (1) clean, (2) clean/contaminated, (3) contaminated, and (4) dirty.4 In the following decades, many studies have established the predictive value of SWC in SSIs.2,5-7 In 1982, the SWC was modified and included in the Center for Disease Control and Prevention (CDC) guidelines for prevention of SSIs as mandatory when dealing with SSI risk stratification.8 Haley et al.5 demonstrated that effective programs for infection surveillance can substantially reduce the SSIs rate by 35%. These results lead to the formation of the National Nosocomial Infections Surveillance (NNIS) system which was developed and established throughout hospitals across the USA. Over the past 3 decades, the NNIS for SSIs control has since been established independently in various European countries including England, Germany, France, Belgium, and the Netherlands.9-12 In 2011, the Swissnoso consortium developed a new nationwide surveillance program for SSIs in Switzerland.13 In addition, there is an increase in public interest in the reporting of individual hospital data, especially SSI rates.14-16 To use infection rates as a basis for measuring quality of care in hospitals, these rates must be appropriate for comparison. Furthermore, it is important to ensure that quality measures and their risk stratification variables are valid and reliable.9,17 However, SWC is often applied subjectively, thereby biasing hospital quality comparability.18 We assessed a validation study to measure the level of agreement in wound classification between three groups composed of surgeons, a Swissnoso-trained infection clinical nurse (ICN), and histologic findings between surgical cases over the study period of one and a half years. The secondary aim of the study was to compare the risk-stratified SSI rates using the different SWC assessments across these three groups.

Methods To evaluate the discrepancies in SWC assessment, we retrospectively analyzed the data in the Cantonal Hospital of Schaffhausen.

All patients undergoing urgent laparoscopic appendectomies between January 2013 and June 2014 were included in the study. Patients with primary open appendectomy were excluded. All patients provided informed written consent before participation. The study was approved by the Ethics Committee of Zurich, Switzerland. The following patient characteristics were collected: age, gender, American Society of Anesthesiologists (ASA) score, body mass index, experience level of the assessing surgeon in years, and the characteristics of the procedure, including SWC documented by each group (surgeons, a Swissnoso-trained ICN, and converted SWC from histologic assessment of the appendicitis), pathologic diagnosis and NNIS risk index category obtained from the Swissnoso database. At the end of each operation, the surgeon assessed the wound and assigned it to one of the traditional four SWCs (clean, clean/contaminated, contaminated, and dirty), as defined by CDC guidelines (1999).19 Wound classification guidelines19 were incorporated in the electronic documentation process to facilitate accurate assignment. To save the preliminary operative report, the SWC assessment was mandatory. A Swissnoso-trained ICN (trained quarterly on CDC guidelines) graded the SWCs retrospectively and independently after reading the description of the appendix in the operative and pathology reports. These SWCs were then reported to the central database and were used with other perioperative clinical variables for the calculation of the NNIS risk index. The NNIS index ranges from 0 to 3 with increasing risk and includes three equally weighted risk factors: (1) ASA classification of three or higher, (2) SWC contaminated or dirty, and (3) duration of operation >75th percentile for the specific procedure group. In addition, the pathology reports were converted into an SWC for the purposes of this study.19 According to CDC definitions, a normal appendix without inflammation was determined to have the classification clean-contaminated because the gastrointestinal tract is entered during the surgery. Diagnoses of acute, suppurative, or phlegmonous appendicitis were considered contaminated and perforated or evident necrotic/gangrenous appendicitis with or without an intraabdominal abscess were considered dirty.17 The SWC of the ICN and pathologic report SWC were evaluated for concordance with the original surgical assignment. Disagreement was defined as any discrepancy in classification among the assessment groups. Findings were evaluated using the percent agreement between the classification methods and the level of agreement, respectively interrater reliability was also calculated using kappa statistic as described by Cohen.17,20-22 Kappa values of 0.01-0.20 were interpreted as slight agreement, and values of 0.41-0.60 were considered as acceptable or moderate agreement. Kappa values between 0.81-1 was characterized as

134

j o u r n a l o f s u r g i c a l r e s e a r c h  j u l y 2 0 1 7 ( 2 1 5 ) 1 3 2 e1 3 9

almost perfect agreement.21,22 Statistical analysis was performed using the R 3.1.0 software. SSIs within a 30-d follow-up period were the secondary outcome of the study. The SSI categories used by Swissnoso correspond with those defined by the CDC in 1992.13,19,23 All patients were prospectively monitored for SSIs by the ICN during a surveillance program period of 1 mo. Surgical wounds were observed by health care professionals (direct observation) during the time of hospital stay or after discharge. SSI diagnosis was based on patient self-reporting, usually through standardized telephone interviews.13 General practitioners were contacted for complementary information in case SSIs were suspected. SSI rates were analyzed among the four SWCs using the chi-square test. An alpha level of <0.05 was considered significant.

Results During the 18-mo period, we identified 248 patients with the diagnosis of acute appendicitis. Two hundred forty-six patients underwent laparoscopic appendectomy and met the study’s inclusion criteria. To limit bias, two cases of primary open appendectomies were excluded.19 Conversion to open appendectomy was necessary four times; in these cases, no apparent complications were observed throughout the followup period. These cases were included in the final analysis (intention-to-treat analysis). Perioperative characteristics of the study population are demonstrated in Table 1. SWC as measured by surgeons, ICN, and histology are shown in Figure 1. In general, the ICN more frequently classified cases as contaminated or dirty/infected compared with the surgeons. The surgeons classified 53 cases (22%) as clean. Macroscopical appendicitis was confirmed in 212 cases (86.2%), whereas a minimal microscopical appendicitis was reported in 9 cases (3.6%). The histologic SWC of these few cases were graded retrospectively as cleancontaminated, particularly with regard to normal looking smooth serosa and only by the presence of minimal and focal inflammation signs. Of the 246 patients, 25 (10.2%) were found to have a normal appendix on microscopical examination, two included other pathologies like carcinoid tumor of the appendix. Because of this, the surgeons misclassified 84 cases (34.1%) as clean contaminated; of 118 cases (48%) classified as clean-contaminated, only in 34 cases, the histologic findings correlated with clean-contaminated SWC (Fig. 1). The surgical- and pathologic-related SWCs were identical in 63/246 cases (25.6%). The ICN mostly admitted the dirty SWCs of surgeons as correct in 51/52 cases (98.1%). Of the 246 evaluated cases, the surgical- and ICN-based SWCs differed in 175 cases (71%; Fig. 2). The ICN- and histologic-based SWCs differed only in 80 cases (33%; Fig. 3). Due to the fact that the SWC component is equally weighed for the calculation of NNIS index either contaminated or dirty/infected, which was assigned for final determining the intrinsic risk of surgical infection, there was an overall agreement among the ICN- and histologic-related SWCs of over 90% (160 accurate classifications and additionally 58 cases with little difference in histology respective contaminated versus dirty SWC [Fig. 3]).

Table 1 e Patient demographics and perioperative characteristics of cases (n [ 246). Demographics/characteristics Age median (range) 47 (5-90)

N

Percentage

246

Age category <20

65

26.4

20-35

70

28.4

36-50

43

17.5

51-65

43

17.5

>65

25

10.2

Gender Female

110

44.7

Male

136

55.3

ASA 1, 2

230

93.5

ASA 3, 4

16

6.5

BMI < 29

212

86.2

BMI 30-34

26

10.6

BMI > 35

8

3.3

>10 y

68

27.6

10-5 y

148

60.2

<5y

30

12.2

ASA score

Body mass index

Surgeon’s level of experience*

NNIS risk indexy 0

9

1

133

54.1

3.65

2

95

38.6

3

9

3.65

BMI ¼ body mass index. * Level of experience of assigning surgeon. y NNIS risk index provided is based on the SWC assigned by the infect control nurse in the Swissnoso databank.

The kappa scores for interrater reliability of SWCs among the surgeons, ICN-, and histology-related SWCs ranged from 0.05 to 0.2. There was only slight agreement between surgeons and ICN as well as between surgeons and histology. For the 240 procedures reviewed by ICN, there was a moderate agreement with histology, kappa 0.41 (95% confidence interval: 0.31-0.51). Six of 246 patients could not be assessed by ICN due to changes in the data evaluation method. The cumulative SSI rate for the total study population was 7 (2.85%) with 1.6% as superficial incisional SSI, 0.8% as deep incisional SSI, and 0.4% as organ space SSI. When stratifying by ICN-assessed SWC, the rate of any SSI was highest among cases that had ICN SWC contaminated (3.3%) or dirty infected (2.9%) and lowest in cases that were clean or cleancontaminated (0%). SSIs were more frequently associated with ICN corrected SWC, instead of operative surgeon based SWC (Table 2). According to the histologic assessment, there was the highest rate of SSIs (3.2%) for contaminated SWC. Considering the additional information contained in the pathology reports, the corrected distribution of SWC by ICN tends to be more accurate than the SWC from the operative report in terms of predicted infectious complications.

wang-chan et al  perioperative characteristics of cases

135

Fig. 1 e Distribution of surgical wound class of all cases.

In summary, the surgeons tended to underclassify the SWC compared with the noted clinical diagnoses in the operative report. The operative diagnoses themselves without regard to SWC were confirmed in 95% on histopathologic examination.

Discussion The currently practiced SWC system is based on data that are decades old, offering a wide range of possible interpretations. Our study shows the potential danger of SWC as an unreliable variable to bias risk adjustment measures for SSI, affecting the comparability between surgeons and hospitals, by demonstrating a poor agreement between surgeons and either ICNor histology-translated SWC. Our data suggest that the postoperative SWC recorded by surgeons is frequently discordant with Swissnoso reference. Several studies have also demonstrated considerable

differences in the SWC of appendectomies. In 2013, three similar published articles show significant variation in the assessment of SWC.17,20,21 Snyder et al.20 demonstrated that the circulating nurse (CN) underclassified surgical wounds more frequently compared with the surgeon and to the independent clinical reviewer (trained in American College of Surgeons National Surgical Quality Improvement Program definitions and methodology). Moreover, Levy et al.17 described the SWCs determined by the CN, and the surgeon was concordant in only 8% of 312 evaluated pediatric appendicitis cases, kappa 0.01. They used diagnosis-related SWC from the operative report as reference for their decision of accuracy to determine SWC. These results highlight the subjectivity involved in traditional SWC systems. Further potential explanations for lack of reliability and accuracy might be a lack of understanding of definitions and inappropriate timing of SWC assignment. In other institutions, the CN is responsible for the graduation at the

Fig. 2 e Agreement between surgeons, Swissnoso-trained ICN (Swissnoso reviewer) and histologic assessment of SWC (%).

136

j o u r n a l o f s u r g i c a l r e s e a r c h  j u l y 2 0 1 7 ( 2 1 5 ) 1 3 2 e1 3 9

Fig. 3 e Agreement between the SWC by Swissnoso-trained ICN and histologic findings.

beginning of the case.17,20,22,24 Given that surgeons have the most knowledge of the intraoperative findings, we strongly believe that the postoperative SWC assessment’s quality will improve when conducted at the end of the operation and conducted by the responsible surgeon, in accordance with the CDC/Swissnoso guidelines.13,19 However, our results show that the surgeons did not understand the classification system, although the description of the appendix in the operative note and the diagnosis were correct. For quality improvement, it requires further regular training in definitions of Swissnoso. Furthermore, the CN often is the only one to assess SWC at many other institutions, particularly in the USA, giving communication between surgeon and CN a crucial significance in ensuring SWC accuracy.17,20,22,24-26 Zens et al.27 show that surgeon-directed SWC improves accuracy of documentation and decreases the degree of error in incorrectly identified cases. Based on our findings, we stress that reliance on the surgeon’s evaluation alone is no solution for reliability improvement. When taking the different levels of training in classification systems into account, the results of our study confirm that there are substantial differences in the application of the SWC definitions.

Dodds et al.22 underlined “that the system is not sufficiently descriptive” showing poor interrater concordance of SWC among a CN and four independent reviewers, with the kappa score of 0.1028-0.1597, concluding that changes in the definitions of SWC could improve the system. From our results, we derived that further precision of definitions or development of a new classification schema, which requires the histologic examination for a proper classification in appendectomies, would be beneficial. SWC is a complex variable. The misclassification and subsequent unreliability of SWC systems could be explained by the difficulty in application of guidelines to all operations. For example, in 2009, Vu et al.28 demonstrated that little is known about how to apply the SWC system to neonatal operations. Their findings show a poor concordance among 144 surgeons in SWC of 22 operations (kappa ¼ 0.30). The differences in assessment among surgeons were also shown for dermatologic procedures.29 Mioton et al.30 expressed their doubts concerning the proper application of the traditional SWC scheme in plastic surgery practice. Ju et al.18 regard nephrectomy as one procedure that can be classified differently and even a clean SWC status might be appropriate under some circumstances. Levy et al.26 published one multicenter

Table 2 e Surgical site infection rates stratified by surgical wound classification among the three different classification groups. Method to determine SWC Surgeon

Wound classification Clean

Histopathology

Total SSIs (n ¼ 7)

SSIs in %

Chi-square (c2)

P value

2.4319

0.49

0.4695

0.79

0.3016

0.86

53

3

5.7

118

2

1.7

Contaminated

23

1

4.3

Dirty

52

1

1.9

Clean-contaminated

14

0

0.0

Contaminated

123

4

3.3

Dirty

103

3

2.9

34

1

2.9

Clean-contaminated

ICN*

Patients (n ¼ 246)

Clean-contaminated Contaminated Dirty

156

5

3.2

56

1

1.8

The chi-square test was used to evaluate results for significant variation in SSI distribution by SWC. * Infection control nurse quarterly trained by Swissnoso.

wang-chan et al  perioperative characteristics of cases

137

Fig. 4 e SWC algorithm. (Color version of figure is available online.)

study which highlights that SWC is not uniformly applied to some types of operations (nonspecific to institutions). The previously cited articles show that misclassification of SWC seems to be a universal problem. To raise the bar to more

than a moderate agreement (which means a kappa value of more than 0.60), we suggest to improve the existing system by taking steps in multiple areas to improve the general applicability and accuracy.

138

j o u r n a l o f s u r g i c a l r e s e a r c h  j u l y 2 0 1 7 ( 2 1 5 ) 1 3 2 e1 3 9

Education should be a first step toward system improvement. In 1993, Cardo et al.25 demonstrated that a high level of accuracy can be achieved by training and reeducating surgeons and nurses on the classification system or simplifying the classification system into two categories. Devaney and Rowell report that 19% of cases were incorrectly classified, and after 3 mo of further training, the misclassification rate decreased slightly to 14%.24 It is crucial to mention that education is only the first step to improvement and, alone, might only have limited influence on classification errors. Although Shiloach et al.31 had shown positive effects of training and audits, the SWC remained a problematic variable, requiring additional efforts. More consistently applied measures might be beneficial by using more available information, such as the pathology reports to verify the classification appropriately. Perhaps, these requirements might be best met by an independent reviewer (infect control team), who could retrospectively assess the SWC once informed of the intraoperative findings and histologic report. On the other hand, it could generally be simply a problem of not having sufficient time to reiterate the differences between the classes. A similar situation arose at the Cone Health in North Carolina. An algorithm was created by nurses for a quality improvement project to lead a discussion about SWC.32 The development of an algorithm is the right step toward building a common understanding of the SWC. To simplify the decision-making process, we propose the implementation of a decision chart, which, while simple, is based on enough key questions to provide a useful basis for the proper wound classification. By answering five short yes or no questions, it becomes more easy to find the proper wound class while at the same time making the relevant parameters more memorable (Fig. 4). An awareness of public’s claim for increased transparency of adverse events, the SSI rates should be adjusted for case mix and must be determined in a reliable way. According to the literature, NNIS index is considered as a predictive factor for the development of SSI.6,33,34 Cohen et al.35 described the ASA score to be a subjective variable sensitive to manipulation. Speicher et al.36 demonstrated that seemingly minor changes in SWC documentation can have significant influence on quality performance and perceived outcomes. They stress “the need for accurate and consistent descriptions in the surgeon’s operative note” and “the need to standardize the reporting of quality metrics nationwide,” which we support. The inconsistency in assessment underlines the question of whether any improvement in accuracy when using a system like the NNIS index, which includes much more “subjective” parameters, is beneficial. There are data suggesting that the NNIS index needs to be revised and new procedure-specific models predicting SSI occurrence should be developed.16,37,38 This was a single institution study, and it is possible that there might be a fundamentally different understanding of definitions in other institutions, therefore limiting the overall application of our results to a larger audience. Different interpretations of the guidelines may have a major impact on the classification results. For example, there are different opinions on how to classify a gangrenous appendicitis. Both contaminated and dirty categories may be acceptable.17 A further limitation is the disagreement as to whether clinical or

histologic findings are more important, particularly when the surgeon diagnosed a perforated appendicitis and histologic examination revealed no perforation signs. One limitation of the use of pathology reports for the SWC assignment is the inclusion of certain intraoperative events in the definition of a contaminated wound that would not be captured by the histologic examination. The decision of SWC for the Swissnoso database is in the responsibility of the ICN.22,31 Despite standardized and regular trainings, a discordance may exist among reviewers. All these factors together highlight the limitations of this study and the lack of a “gold standard” for determining SWC.

Conclusion SWC is a complex and subjective variable that is subject to a wide range of interpretation. Appendectomies remain a challenge to classify with the SWC. The lack of clear definitions for the application of the SWC can lead to biased reporting results. Given the unreliability of the SWC systems, it is inappropriate to use it for risk-adjusted assessments or for quality benchmarking. Before hospital infection rate data become public and used to judge hospital quality performance and before comparisons using the SWC are made between at the European and international level, we need further standardization in the SWC definitions and in the assessment of data. To improve the SWC accuracy, we recommend the use of pathologic reports to assign the proper classification in appendectomies. Perhaps, we can reach efforts in making sure we classify wounds correctly with the use of the simplified algorithm (Fig. 4). As long as the SWC system has not been clarified, it is essential that the process of SWC assignment be subject to external (surveillance audit) and/or internal control. The uniformity of SWC classification definitions and the application of SWC to treatment are of sufficient importance to merit a collaborative study of this problem.

Acknowledgment The authors gratefully acknowledge the contributions of the following individuals for their time and assistance with this project: Nadine Behrle Dr-Ing., infection control department, ICN/Swissnoso reviewer; Christian Conrad MPH, lecturer of health sciences; and Monika Kriner PhD, independently contracted data analyst. Authors’ contribution: A.W-C. helped in collecting, analyzing and interpreting the data, writing the manuscript, and approving the final version of the manuscript. C.G. helped in language editing, providing critical revisions that are important for the intellectual content, and approving the final version of the manuscript. E.A. hepled in proof reading, providing critical revisions that are important for the intellectual content, and approving the final version of the manuscript. F.H.H. helped in general supervision, concepting and designing of study, and approving the final version of the

wang-chan et al  perioperative characteristics of cases

manuscript. All authors have read and approved the submission of the manuscript to Journal of surgical Research (JSR).

Disclosure The authors reported no proprietary or commercial interest in any product mentioned or concept discussed in the article.

references

1. Anderson DJ, Podgomy K, Berrios-Torres SI, et al. Strategies to prevent surgical site infections in acute care hospitals: 2014 update. Infect Control Hosp Epidemiol. 2014;35:605e627. 2. Cruse PJE, Foord R. The epidemiology of wound infections: a 10-year prospective study of 62,939 wounds. Surg Clin North Am. 1980;60:27e39. 3. Olson M, O’Connor M, Schwartz ML. Surgical wound infections. A 5-year prospective study of 20,193 wounds at the Minneapolis VA Medical Center. Ann Surg. 1984;199:253. 4. Berard F, Gandon J. Postoperative wound infections: the influence of ultraviolet irradiation of the operating room and of various other factors. Ann Surg. 1964;160:1e192. 5. Haley RW, Culver DH, White JW, et al. The efficacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals. Am J Epidemiol. 1985;121:182e205. 6. 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;91:152e157. 7. Ortega G, Rhee DS, Papandria DJ, et al. An evaluation of surgical site infections by wound classification system using the ACS-NSQIP. J Surg Res. 2012;174:33e38. 8. Garner JS. CDC guideline for prevention of surgical wound infections, 1985. Supersedes guideline for prevention of surgical wound infections published in 1982. (Originally published in November 1985). Revised. Infect Control. 1986;7:193e200. 9. Gastmeier P, Sohr D, Schwab F, et al. Ten years of KISS: the most important requirements for success. J Hosp Infect. 2008;70:11e16. 10. Astagneau P, L’Heriteau F, Daniel F, et al. Reducing surgical site infection incidence through a network: results from the French ISO-RAISIN surveillance system. J Hosp Infect. 2009;72:127e134. 11. Coello R, Gastmeier P, De Boer A. Surveillance of hospital acquired infection in England, Germany, and the Netherlands: will international comparison of rates be possible? Infect Control Hosp Epidemiol. 2001;22:393e397. 12. Mertens R, Jans B, Kurz X. A computerized nationwide network for nosocomial infection surveillance in Belgium. Infect Control Hosp Epidemiol. 1994;15:171e179. 13. Staszewicz W, Eisenring MC, Bettschart V, Harbarth S, Troillet N. Thirteen years of surgical site infection surveillance in Swiss hospitals. J Hosp Infect. 2014;88:40e47. 14. Haustein T, Gastmeier P, Holmes A, et al. Use of benchmarking and public reporting for infection control in four high-income countries. Lancet Infect Dis. 2011;14:471e481. 15. Martin M, Zingg W, Hansen S, et al, PROHIBIT study group. Public reporting of healthcare-associated infection data in Europe. What are the views of infection prevention opinion leaders? J Hosp Infect. 2013;14:94e98. 16. Biscoine F. Rates of surgical site infection as a performance measure: are we ready? World J Gastrointest Surg. 2009;1:11e15.

139

17. Levy SM, Holzmann-Pazgal G, Lally KP, et al. Quality check of a quality measure: surgical wound classification discrepancies impact risk-stratified surgical site infection rates in pediatric appendicitis. J Am Coll Surg. 2013;217:969e973. 18. Ju MH, Cohen ME, Bilimoria KY, et al. Effect of wound classification on risk adjustment in American College of Surgeons NSQIP. J Am Coll Surg. 2014;219:371e381. 19. Mangram AJ, Horan TC, Pearson ML, et al. Guideline for prevention of surgical site infection: 1999. Infect Control Hosp Epidemiol. 1999;20:247e277. 20. Snyder RA, Johnson L, Tice J, et al. Wound classification in pediatric general surgery: significant variation exists among providers. J Am Coll Surg. 2013;217:819e826. 21. Cohen J. A coefficient of agreement for nominal scales. Educ Psychol Meas. 1960;20:37e46. 22. Dodds PR, Meinke AK, Lincer RM, Fitzgerald EJ, Dodds JH. Interrater concordance of wound classifications in patients undergoing appendectomy. Surg Infect (Larchmt). 2013;14:445e450. 23. Horan TC, Gaynes RP, Martone WJ, et al. CDC definitions of nosocomial surgical site infections, 1992: A modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol. 1992;13:606e608. 24. Devaney L, Rowell KS. Improving surgical wound classification: why it matters. AORN J. 2004;80:208e223. 25. Cardo DM, Falk PS, Mayhel CG. Validation of surgical wound classification in the operating room. Infect Control Hosp Epidemiol. 1993;14:255e259. 26. Levy SM, Lally KP, Blakely ML, et al. Surgical wound misclassification: a multicenter evaluation. J Am Coll Surg. 2015;220:323e329. 27. Zens TJ, Rusy DA, Gosain A. Pediatric surgeon-directed wound classification improves accuracy. J Surg Res. 2016;202:432e439. 28. Vu LT, Nobuhara KL, Lee H, et al. Conflicts in wound classification of neonatal operations. J Pediatr Surg. 2009;44:1206e1211. 29. Hirschmann JV. Proper classification of surgical wounds. J Am Acad Dermatol. 2011;65:202e203. 30. Mioton LM, Jordan SW, Hanwright PJ, Bilimoria KY, Kim JY. The relationship between preoperative wound classification and postoperative infection: a multi-institutional analysis of 15,289 patients. Arch Plast Surg. 2013;40:522e529. 31. Shiloach M, Frencher SK, Steeger JE, et al. Toward robust information: data quality and inter-rater reliability in the American College of Surgeons National Surgical Quality Improvement Program. J Am Coll Surg. 2010;210:6e16. 32. Zinn J, Swofford V. Quality improvement initiative: classifying and documenting surgical wounds. Wound Care Advisor. 2014;3:32e38. 33. Brandt C, Hanson S, Sohr D, et al. Finding a method for optimizing risk adjustment when comparing surgical-site infection rates. Infect Control Hosp Epidemiol. 2004;25:313e318. 34. Friedman ND, Bull AL, Russo PL, Gurrin L, Richards M. Performance of the national nosocomial infections surveillance risk index in predicting surgical site infection in Australia. Infect Control Hosp Epidemiol. 2007;28:55e59. 35. Cohen ME, Bilimoria KY, Ko CY, et al. Effect of subjective preoperative variables on risk-adjusted assessment of hospital morbidity and mortality. Ann Surg. 2009;249:682e689. 36. Speicher PJ, Nussbaum DP, Scarborough JE, et al. Wound classification reporting in HPB surgery: can a single word change public perception of institutional performance? HPB (Oxford). 2014;16:1068e1073. 37. Geubbels EL, Grobbee DE, Vandenbroucke-Grauls CM, Wille JC, de Boer AS. Improved risk adjustment for comparison of surgical site infection rates. Infect Control Hosp Epidemiol. 2006;27:1330e1339. 38. Gaynes RP. Surgical-site infections and the NNIS SSI Risk Index: room for improvement. Infect Control Hosp Epidemiol. 2000;21:184e185.