Surgical Site Infections in an Italian Surgical Ward: A Prospective Study

Abstract

Background:

Surgical site infection (SSI) remains a major cause of morbidity and death. This study analyzed the results of surveillance to evaluate the incidence, risk factors, and characteristics of SSI in patients who underwent an operation in a typical Italian surgical ward.

Methods:

A group of 1,281 patients operated on from August 2005 to December 2007 underwent prospective and direct observation of incisions by a surgeon according to the U.S. Centers for Disease Control and Prevention (CDC) National Nosocomial Infections Surveillance (NNIS) method. The minimum follow-up was 30 days. A locally-modified risk index score (LRI) based on the NNIS was calculated for each patient, using as a cut point the 75^th percentile of the duration of surgery (in minutes) for that particular procedure.

Results:

Seventy-six patients were affected by incision site infection, and the SSI rate was 5.9%. Thirty-four (2.6% of the series) were superficial incisional, 32 (2.5%) deep incisional, and 10 (0.8%) organ/space SSIs. An increasing value of the LRI was significantly (p < 0.05) related to an increasing risk of infection. The SSI rates were 0.6%, 3.7%, 7.3%, and 26.8% for LRI value of M = − 1, 0, 1, and ≥2, respectively. Obesity (body mass index >30 kg/m²), diabetes mellitus, and emergency surgery were associated with a higher risk of infection by multivariable analysis independent of the LRI.

Conclusions:

The NNIS method can be useful for SSI surveillance and monitoring in single surgical wards. Longer operations, diabetes mellitus, and obesity increase the risk of SSI, as does performance of surgery in an emergency situation.

Surgical site infection (SSI) is an important cause of morbidity and death for hospital inpatients. According to the National Nosocomial Infections Surveillance system (NNIS), the U.S. Centers for Disease Control and Prevention (CDC) reported that SSIs are the third most common nosocomial infection, accounting for 14–16% of all nosocomial infections in the United States. In surgical patients, SSIs are the most common nosocomial infection, accounting for 38% of all such infections [1].

About four years ago, it was reported that in Italian public hospitals, the infrastructures for infection control were suboptimal as judged by the guidelines and surveys published in other countries [2]. The Italian National Infections Surveillance System (INF-OSS), meant to collect hospital data from many regions, is still in progress [3]. Reported rates of SSIs range from 2.2–5.2%, but accurate standardized methods of monitoring SSIs are rarely adopted by surgical wards [4 –6]. Nevertheless, in 2005, an SSI surveillance program was introduced into our surgical unit.

The aim of this study was to evaluate the incidence, risk factors, and characteristics of SSI in patients who underwent an operation in a typical Italian surgical ward according to the CDC NNIS method.

Patients and Methods

From August 2005–December 2007, prospective surveillance was carried out for SSI in 1,626 patients who underwent an operation in the Unit of General Surgery of Alto Garda and Ledro Hospital in the province of Trento. The hospital provides 150 beds and is one of the seven acute care hospitals of the public health service of the Province of Trento (APSS). Ninety-three patients (6%) with “open” surgical incisions and 252 (15%) who were lost after hospital discharge or had a follow-up period shorter than 30 days were excluded. Ultimately, 1,281 patients were included in the analysis (Table 1). The mean age of the population was 54 years (range 8–97) and 670 patients were male and 611 female (ratio 1.1:1).

Table 1.

Operative Procedure Categories of 1,281 Patients Who Underwent Surgical Site Infection Survey, August 2005 to December 2007

Category	N (%) of cases
Other integumentary system surgery	57 (4.4)
Lymphatic system surgery	14 (1.1)
Breast surgery	37 (2.9)
Vein stripping	227 (17.7)
Herniorrhaphy	298 (23.4)
Incisional herniorrhaphy	20 (1.6)
Other genitourinary surgery	24 (1.9)
Thyroid surgery	53 (4.1)
Appendectomy	149 (11.6)
Cholecystectomy	212 (16.5)
Gastric surgery	32 (2.5)
Small bowel surgery	36 (2.8)
Other digestive surgery	25 (1.9)
Colon surgery	73 (5.7)
Rectal surgery	24 (1.9)

All the data were collected according to the NNIS guidelines [1,7]. The patient and operative characteristics that may influence the risk of SSI were registered (Table 2). We specifically examined smokers (patients who habitually smoke at least five cigarettes a day), persons affected by diabetes mellitus (insulin- or non-insulin-dependent) at the time of admission, and obese patients (body max index [BMI] >30 kg/m²). The classes of the surgical procedures and the American Society of Anesthesiologists (ASA) scores were collected. Perioperative blood transfusions (at least one) and the number of people in the operating room during surgery also were registered; we considered eight or more people to be potentially important because seven was the mean number of staff members in the operating room.

Table 2.

Main Characteristics of Study Population of 1,281 Patients Who Underwent Surgical Site Infection Survey, August 2005 to December 2007

Characteristic	N (%) of cases
Age over 65 years	454 (35.4)
Male sex	670 (52.3)
Emergency surgery	268 (20.9)
Contaminated (III)-dirty/infected (IV) incision	84 (6.6)
Smoker	81 (6.3)
Diabetes mellitus	32 (2.5)
Obesity (body mass index >30 kg/m²)	73 (5.7)
Blood transfusion	91 (7.1)
American Society of Anesthesiologists score ≥3	157 (12.3)
≥8 people in operating room	310 (24.2)

The SSIs were divided into three categories: Superficial incisional, deep incisional, and organ/space. Incisions were classified from I to IV: Clean, clean-contaminated, contaminated, and dirty-infected, respectively. A locally modified NNIS risk index score (LRI) was calculated for each patient using as the duration cut point the 75^th percentile of duration of surgery (in minutes) for that particular operative procedure in our ward (Table 3). If the procedure was conducted laparoscopically, one was subtracted from the numbers of risk factors (ASA score ≥3; duration of surgery >75^th percentile; or contaminated-III or dirty-IV incision class) in the LRI, and “M” indicates = 1 in the modified risk category.

Table 3.

Patient Counts (%) for Components of Modified National Nosocomial Infections Surveillance Risk Index by Type of Surgical Procedure

Operative procedure category	N (%) of cases	Duration cut point^a	ASA score ≥3^b	Incision class III or IV^c
Other integumentary system surgery	57 (4.4)	37	4 (7)	18 (31.6)
Lymphatic system surgery	14 (1.1)	46	2 (14.3)	0
Breast surgery	37 (2.9)	110	8 (21.6)	0
Vein stripping	227 (17.7)	47	13 (5.7)	0
Herniorrhaphy	298 (23.4)	60	35 (11.7)	0
Incisional herniorrhaphy	20 (1.6)	139	1 (0.5)	0
Other genitourinary surgery	24 (1.9)	41	2 (8.3)	0
Thyroid surgery	53 (4.1)	185	4 (7.5)	0
Appendectomy	149 (11.6)	55	4 (2.7)	8 (5.4)
Cholecystectomy	212 (16.5)	87	29 (13.7)	11 (5.2)
Gastric surgery	32 (2.5)	282	9 (28.1)	9 (28.1)
Small bowel surgery	36 (2.8)	159	9 (25)	7 (19.4)
Other digestive surgery	25 (1.9)	176	7 (28)	1 (3.6)
Colon surgery	73 (5.7)	250	21 (28.8)	25 (34.2)
Rectal surgery	24 (1.9)	314	7 (29.2)	5 (20.8)

Duration cut point is the 75^th percentile in minutes for that procedure.

American Society of Anaesthesiologists physical status classification.

Incisions of class III and IV are contaminated and dirty, respectively.

The presence of fever (>38°C) and the duration in days of each SSI were registered. We considered the first day of the SSI to be the day of clinical diagnosis according to the NNIS criteria and the last day of infection to be the day of its clinical resolution.

Antimicrobial prophylaxis and therapy

We have always used an intravenous antibiotic for prevention of postoperative infections. Table 4 shows the antimicrobial agents of choice for each surgical procedure, as established by the APSS Infection Control Committee (CIPASS). The 1,197 patients with clean or clean-contaminated operations received only antimicrobial prophylaxis, which was administered by the anesthesiologist within 30 to 60 min before the skin was incised. Eighty-four patients who underwent contaminated (n = 70) or dirty-infected (n = 14) surgery received antibiotic therapy rather than prophylaxis.

Table 4.

Antimicrobial Agents Used for Prophylaxis or Therapy for each Operative Procedure

Operative procedure	Antimicrobial agent
Other integumentary system surgery	Cefazolin
Lymphatic system surgery	Cefazolin
Breast surgery	Cefazolin
Vein stripping	Cefazolin
Herniorrhaphy	Cefazolin
Incisional herniorrhaphy	Cefazolin
Other genitourinary surgery	Cefazolin
Thyroid surgery	Cefazolin
Appendectomy	Cefoxitin
Cholecystectomy	Cefazolin or piperacillin-tazobactam
Gastric surgery	Cefazolin
Small bowel surgery	Cefazolin
Other digestive surgery	Cefazolin or cefoxitin
Colon surgery	Cefoxitin
Rectal surgery	Cefoxitin

Incision follow-up

Before the start of the survey period, the surgeons and nurses on our ward attended a training meeting on the NNIS guidelines conducted by infection control personnel (LF, GP, and PT) of the CIPASS. This training was repeated about every six months. During these meetings, the results from the previous survey periods were reported to the personnel of the surgical ward and forwarded to CIPASS.

Individual patient's verbal informed consent always was obtained. All 1,281 patients in analysis underwent direct observation of their incisions by a surgeon of our ward, and the minimum follow-up period was 30 days. Patients were evaluated daily during their hospital stays for onset of SSI. After discharge from the hospital, our periodic direct observation in the outpatient department was the only monitoring method adopted. Incisions were monitored by at least three reviews within the first 15 days, with one being carried out 30 days after the surgical procedure. Results were recorded initially on a data sheet and later in an electronic database.

Statistical analysis

Data were analyzed using the Epi-Info package software program version 3.3 (U.S. Centers for Disease Control and Prevention, Atlanta, GA). A 2 × 2 table was constructed, and bivariable comparisons were used to analyze the relations between SSI and the study variables reported in Table 2. This produced crude odds ratios (ORs) with their 95% confidence intervals (CIs).

After that, we carried out a logistic regression analysis to examine the influence of those variables that were significant in the bivariable analysis, obtaining adjusted ORs. The LRI was entered in the analysis after exclusion of the two variables (ASA score and incision class) already included in this index. The LRI categories 0, 1, and ≥2 were considered.

A p value <0.05 was considered statistically significant in both bivariable and multivariable analyses. The chi-square test was used to compare the differences between the rates of SSIs observed in the first and second years of the surveillance (2006 and 2007).

Results

The number of SSIs for the 1,281 patients was 76, with a cumulative rate of 5.9%. No patients died of SSI. Patients with class III and class IV incisions showed the highest percentage of SSIs (20%), whereas the lowest was registered for class I and class II cases (5%). Thirty-four patients (2.6%) had a superficial incisional SSI, 32 (2.5%) a deep incisional SSI, and 10 (0.8%) an organ/space SSI. Seven (70%) of the patients with organ/space SSI had an incisional SSI also. Table 5 shows the SSI rate according to patient characteristics.

Table 5.

Surgical Site Infection Probability According to Patient Characteristics

Patient characteristics	SSI (%)
Age over 65 years	38/454 (8.4)
Male sex	42/670 (6.3)
Emergency surgery	31/268 (11.6)
Contaminated (III)-dirty/infected (IV)	17/ 84 (20.2)
Smoker	7/ 81 (8.6)
Diabetes mellitus	12/ 32 (37.5)
Obesity (body mass index >30 kg/m²)	18/ 73 (24.7)
Blood transfusion	21/ 91 (23.1)
American Society of Anesthesiologists score ≥3	27/157 (17.2)
≥8 people in operating room	25/310 (8.1)

The LRI correlated well with the rate of infection. The SSI rates were 0.6% (1/168), 3.7% (28/743), 7.3% (22/303) and 26.8% (18/67) for risk index values of M, 0, 1, and ≥2, respectively.

The mean duration of superficial incisional SSIs was 7.3 days (range 3–18 days) with a median value of 6 days, whereas for the deep incisional SSIs, the mean and median durations were 10 days (range 5–32 days) and 7 days, respectively. Three patients (30%) with organ/space SSIs needed surgical treatment (re-operation). For the other seven, the mean duration of the SSI was 14.1 days (range 6–27 days) with a median of 10 days.

Interesting, from the clinical point of view, was the behavior of incisional infections in reference to the time of the surgical operation and the relation between fever (>38°C) and SSI. Thirty-two (94%) of the superficial incisional SSIs began before the tenth postoperative day: 20 (59%) within five days and 12 (35%) between the sixth and tenth days. Twenty-nine (91%) of the deep incisional SSIs began within ten days: 13 (41%) within five days and 16 (50%) between the sixth and tenth days. All ten organ/space SSIs began within ten days: seven within five days and three between the sixth and tenth days. Seven patients (70%) with organ/space SSIs showed fever at the same time, whereas fewer than half of the patients with incisional SSIs did so: nine (26%) with superficial incisional SSIs and 15 (47%) with deep incisional SSIs.

In 37 of 66 incisional SSIs (56%) and whenever there was an organ/space SSI, microbiological samples were collected by swabbing of the incision, the drain exudates, or both. Among the patients suffering from superficial or deep incisional SSIs, the culture was positive in 32 cases (86%), and gram-positive cocci were the most common pathogens identified (78%): Nine Enterococcus faecalis, seven Staphylococcus aureus, three S. epidermidis, two S. hominis, two Streptococcus pyogenes, one S. haemolyticus, and one S. auricularis. Seven Pseudomonas aeruginosa (22%), six Enterobacteriaceae (19%) (three Enterobacter cloacae, one Klebsiella pneumoniae, one Proteus penneri, one Morganella morganii), two Bacterioides fragilis (6%), and two Candida albicans (6%) also were isolated. In 10 patients (31%), two of the previously listed microorganisms were involved. There were no methicillin-resistant S. aureus (MRSA) or enterococci resistant to vancomycin or ampicillin. Multi-drug-resistant (MDR) P. aeruginosa was never found. Four of the six Enterobacteriaceae were resistant to ampicillin, whereas the P. penneri was ampicillin- and cefotaxime-resistant. The K. pneumoniae was an extended-spectrum beta-lactamase (ESBL) producer.

Cultures always were positive in patients affected by organ/space infections. The microorganisms isolated were seven gram-positive cocci (two each of S. aureus, S. haemolyticus, and E. faecalis, and one S. warneri), one Serratia marcescens, one B. fragilis, and two each of P. aeruginosa and Candida albicans. In three patients, the organ/space SSI yielded two of the previously listed microorganisms. There was one MRSA that also was resistant to gentamicin, ciprofloxacin, and erythromycin. The Serratia marcescens was resistant to ampicillin and cefotaxime. There were no vancomycin- or ampicillin-resistant enterococci, and neither of the two P. aeruginosa was MDR.

Age over 65 years; performance of emergency surgery; and the presence of obesity, diabetes mellitus, incision class III or IV, blood transfusion, or ASA ≥3 were the risk factors significantly associated with SSI in bivariable analysis (Table 6). Logistic regression multivariable analysis (Table 7) showed that the value of the LRI increased in tandem with the risk. Emergency surgery, diabetes mellitus, and obesity independently increased the risk of SSI (p < 0.05). The global rate of SSI declined from 6.6% (33/498) in 2006 to 5.8% (33/565) in 2007, but this reduction was not statistically significant.

Table 6.

Results of Bivariable Analysis

Patient characteristics	Odds ratio (95% confidence interval)	P value
Age over 65 years	1.72 (1.03, 2.88)	<0.05
Male sex	1.20 (0.72, 20.1)	NS
Emergency surgery	3.63 (1.94, 6.91)	<0.05
Obesity (body mass index >30 kg/m²)	6.75 (3.50, 12.90)	<0.05
Smoker	1.44 (0.54, 3.61)	NS
Diabetes mellitus	9.17 (3.85, 21.46)	<0.05
Incision class III or IV	4.15 (2.09, 8.14)	<0.05
Blood transfusion	4.58 (2.38, 8.70)	<0.05
American Society of Anesthesiologists ≥3	4.02 (2.28, 7.05)	<0.05
≥8 people in operating room	1.50 (0.86, 2.61)	NS

NS = not significant.

Table 7.

Results of Multiple Logistic Regression Analysis of Variables Associated with High Risk of Surgical Site Infection by Bivariable Analysis

Patient characteristic	Odds ratio (95% confidence interval)	P value
Age over 65 years	1.06 (0.63, 1.78)	0.8352
Emergency surgery	2.44 (1.41, 4.22)	0.0015
Obesity (body mass index >30 kg/m²)	13.82 (4.69, 40.76)	0.0000
Diabetes mellitus	21.48 (4.26, 108.34)	0.0002
Blood transfusion	2.61 (0.93, 7.30)	0.0676
Locally modified risk index
0	3.06 (1.37, 6.83)	0.0063
1	4.97 (2.03, 12.15)	0.0004
≥2	75.27 (19.67, 287.78)	0.0000

Discussion

Postoperative SSIs are still an important source of morbidity and death in surgical patients [8]. These infections result in higher costs because of longer hospitalization, more nursing care, additional dressings, potential readmission to the hospital, and further surgical procedures. It was recently estimated that the costs of managing an SSI increase with the depth of infection, ranging from less than $400 per case for superficial SSIs to more than $30,000 per case for serious organ or space infections [9].

The development of postoperative incisional infections is multifactorial and arises from the complex interaction of host and environmental factors. These include the type of surgery, its duration, the operative technique, the extent of contamination at the operative site, and any patient co-morbidities. The incidence of SSI depends on the hospital, the surgeon, the surgical procedure, and the patient. Nevertheless, SSI rates are an established measure of the quality of clinical care [10], and infection surveillance with feedback to the staff helps reduce infection rates [11,12]. Similarly, in 1991, the Italian National Health Institute published a study that showed a reduction in incision infections in response to active infection surveillance in some surgical wards [13].

In a nosocomial infection surveillance system, the methods, the criteria used, the procedures, and the infection variables should be standardized. This is important to guarantee the quality and reliability of surveillance data, which are requirements for comparison of the infection rate within and between hospitals [14]. In practice, these factors are not always so easy for all hospitals in a national survey program to achieve; and, probably together with different regional infection control policies, they helped to stop full implementation of the Italian national program, INF-OSS [2,3].

In our surgical ward, the surveillance program was adopted in 2005 as part of the national prospective multicenter program. The global rate of SSI was 5.9%. This rate is similar to 3.3%–5.2% rates reported by three recent multicenter Italian surveys that used the NNIS methodology [4,6,15], but lower than the rate of 10.6% another center found [16]. These differences probably are attributable to the different distribution of surgical procedures analyzed, the different states of health of the patients treated, and the method of incision follow up adopted after hospital discharge. Telephone interviews with the patient, patient self-reporting in settings that differ from the surgical ward where the procedure was performed, postal questionnaires to patients or their general practitioners, and even the direct observation of incisions by untrained medical or nursing personnel could reduce the accuracy of SSI detection and lead to an incorrect classification [17]. Nevertheless, at the moment, there is no universally shared strategy for the post-discharge monitoring of incisional infections [14,18,19].

Others authors have conducted analyses in Italy for periods of a few months [5,15]. We performed an analysis for the first 29 months of a prospective and continuous survey that is still on course where the surgical procedures and the survey were done by the same surgeons. In this way, we not only have guaranteed early recognition and treatment of SSIs but also avoided the punitive atmosphere that surveillance performed by external personnel can create. Lacking an independent trained observer, our survey program was always monitored by the CIPASS, which gathers all the data of the seven acute care hospitals in the province of Trento.

In our experience, 94% of the superficial incisional SSIs, 91% of the deep incisional SSIs, and all the organ/space SSIs began within ten postoperative days. If these results are confirmed in our future surveillance, they could facilitate surveillance and make it more efficient, particularly after hospital discharge. Furthermore, the organ/space SSIs have the highest mean durations: 14.1 days vs. 6 days for superficial and 10 days for deep incisional infections. Considering that 30% of the patients in the first group need re-operation and that we have had no deaths caused by infections, these values can be an indirect measure of the time lost to SSIs in terms of mean time to cure.

Despite the common use of antibiotic as prophylaxis, we registered low rates of antimicrobial-resistant pathogens. Methicillin-resistant S. aureus accounted for only 11% of all S. aureus isolated, and 14% of all Enterobacteriaceae were ESBL producers. Our rate of MRSA isolation was lower than that in other Italian hospitals studied [6,20].

In the NNIS risk index, the duration cut point “T” is defined as the 75^th percentile of the distribution of procedure duration, rounded to the nearest hour, and is calculated for each category of surgical procedure. This cut-off was chosen arbitrarily by Culver et al. to provide the index with additional discriminatory power when applied to specific operative procedures, whose extremely long duration may serve as a marker for the length of exposure to potential contamination and for the complexity of the procedure and the surgical technique [21].

The NNIS risk index has been validated in populations other than those of the participating hospitals [22,23], whereas in other studies, this index is calculated taking into account the potential influence of local factors [24,25]. In fact, the length of the operation can be related, not only to factors intrinsic to the patient, but also to external circumstances, not only the complexity of the procedure, the skill of the surgeon, and the extent of tissue trauma, but also modern changes in operative technique and surgical tools. The latter are the reason that in this preliminary analysis, we decided to calculate the 75^th percentile in minutes for surgical procedures in our ward and to use this time cut-off to create a modified risk index (LRI). On the other hand, having the procedure durations in minutes, our choice does not exclude the possibility that it could be changed in further analyses as in the multicenter national program.

In the NNIS guidelines [1], the criteria for defining a deep incisional SSI include fever (>38°C). In our experience, only half (47%) of the patients with this kind of SSI have fever, whereas in organ/space SSI, fever was present in 70% of cases.

Although age over 65 years, emergency surgery, obesity, diabetes mellitus, incision class III or IV, blood transfusion, and ASA class ≥3 were risk factors for infection in bivariable analysis, only emergency surgery, diabetes mellitus, and obesity were independently associated with SSI in logistic regression multivariable analysis. The latter analysis showed the risk of SSI to rise significantly with the LRI values 0, 1, and ≥2. In other studies, the incidence of SSI was significantly higher in patients who underwent emergency surgery [15,26 –28]. Similarly, many other investigators have reported that diabetes [6,27,29,30] and obesity [6,30 –32] are two important risk factors for incisional infections.

Although we registered a reduction in the SSI rate with time, the difference between 2006 and 2007 was not statistically significant. After analyzing this reduction in relation to each independent risk factor—emergency surgery, diabetes mellitus, obesity and LRI values (data not shown)—we noticed that it was attributable to a reduction of the SSI rate in patients affected by obesity and with an LRI value ≤1, but the finding was not statistically significant. As noted by other authors, only a longer period of surveillance could confirm this trend [12,28].

The current study has the obvious limitation of being small. This is the reason it was not possible to evaluate the performance of the LRI in each of the operative categories. An expansion of our survey will be necessary to clarify their relation.

Conclusion

Our work shows how a single Italian surgical ward can adopt the methodology of the NNIS to survey its incisional infections as well as can multicenter or national programs. Using international standards, this activity is useful in the single center for early identification and treatment of SSI, but also to create a pool of data for use in the largest national analysis.

Footnotes

Acknowledgments

We thank Ms Ornella Rosà and Mrs. Adriana Filippi for their help in database construction and the nurses Sonia Dulcamara and Tiziana Zattera for their help in data collection.

All authors declare that they have no potential conflicts of interest relevant to this article.

Author Disclosure Statement

No conflicting financial interests exist.

References

Mangram

, Horan

, Pearson

et al. Guideline for prevention of surgical site infection, 1999. Infect Control Hosp Epidemiol, 1999; 20:247–278.

Moro

, Petrosillo

, Gandin

, Bella

. Infection control programs in Italian hospitals. Infect Control Hosp Epidemiol, 2004; 25:36–40.

Progetto

I-OSS

http://www.ccm.network.it/?q=node84. 2008 July 28.

Moro

, Morsillo

, Tangenti

et al. Rates of surgical-site infections: An international comparison. Infect Control Hosp Epidemiol, 2005; 26:442–448.

Busetti

, Giuliani

, Pisa

. Referenti Regionali Infezioni nelle Organizzazioni Sanitarie. La sorveglianza delle infezioni del sito chirurgico (ISC): l'Esperienza del Friuli Venezia-Giulia. Giornale Italiano Infezioni Ospedaliere, 2004; 11:136–147.

Fiorio

, Marvaso

, Viganò

, Marchetti

. Incidence of surgical site infections in general surgery in Italy. Infection, 2006; 34:310–314.

Horan

, Gaynes

, Martone

et al. CDC definitions of nosocomial surgical site infections, 1992: A modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol, 1992; 13:606–608.

Nichols

. Preventing surgical site infections: A surgeon's perspective. Emerging Infect Diseases, 2001; 7:220–224.

Urban

. Cost analysis of surgical site infections. Surg Infect, 2006; 7,Supp 1:19–22.

10.

Wenzel

. Nosocomial infections, Diagnosis-Related Groups, and study on efficacy of nosocomial infection control: Economic implications for hospitals under the prospective payment system. Am J Med, 1985; 78:3–7.

11.

Haley

, Culver

, White

et al. The efficacy of infection surveillance and control programs in preventing nosocomial infections in US hospitals. Am J Epidemiol, 1985; 121:182–205.

12.

Geubbels

, Nagelkerke

, Mintjes-de Groot

et al. Reduced risk of surgical site infections thorough surveillance in a network. Int J Qual Health Care, 2006; 18:127–133.

13.

Greco

, Moro

, Tozzi

, De Giacomi

. Effectiveness of an intervention program in reducing postoperative infections. Italian PRINOS Study Group. Am J Med, 1991; 91:164S–169S.

14.

Bruce

, Russel

, Mollison

, Krukowski

. The quality of measurement of surgical wound infection as the basis for monitoring: A systematic review. J Hosp Infect, 2001; 49:99–108.

15.

Petrosillo

, Drapeau

CMJ

, Nicastri

et al. Surgical site infections in Italian hospitals: A prospective multicenter study. BMC Infect Dis, 2008; 8:34–43.

16.

Prospero

, Cavicchi

, Bacelli

et al. Surveillance for surgical site infection after hospital discharge: A surgical procedure-specific perspective. Infect Control Hosp Epidemiol, 2006; 27:1313–1317.

17.

Gastmeier

. Postdischarge surveillance for surgical site infection: The continuing challenge. Infect Control Hosp Epidemiol, 2006; 27:1287–1290.

18.

Reilly

, Allardice

, Bruce

et al. Procedure-specific surgical site infection rates and postdischarge surveillance in Scotland. Infect Control Hosp Epidemiol, 2006; 27:1318–1323.

19.

Petherick

, Dalton

, Moore

, Cullum

. Methods for identifying surgical wound infection after discharge from hospital: A systematic review. BMC Infect Dis, 2006; 6:170–180.

20.

Lizioli

, Privitera

, Alliata

et al. Prevalence of nosocomial infections in Italy: Result from the Lombardy survey in 2000. J Hosp Infect, 2003; 54:141–148.

21.

Culver

, Horan

, Gaynes

et al. Surgical wound infection rates by wound class, operative procedure, and patient risk index. Am J Med, 1991; 91:152S–157S.

22.

Friedman

, Bull

, Russo

et al. Performance of the National Nosocomial Infections Surveillance risk index in predicting surgical site infection in Australia. Infect Control Hosp Epidemiol, 2007; 28:55–59.

23.

Brandt

, Hansen

, Sohr

et al. Finding a method for optimizing risk adjustment when comparing surgical-site infection rates. Infect Control Hosp Epidemiol, 2004; 25:313–318.

24.

Campos

, Cipriano

, Freitas

. Suitability of the NNIS index for estimating surgical-site infection risk at a small university hospital in Brazil. Infect Control Hosp Epidemiol, 2001; 22:268–272.

25.

Leong

, Wilson

, Charlett

. Duration of operation as a risk factor for surgical site infection: Comparison of English and US data. J Hosp Infect, 2006; 63:255–262.

26.

Watanabe

, Kohnoe

, Shimabukuro

et al. Risk factors associated with surgical site infection in upper and lower gastrointestinal surgery. Surg Today, 2008; 38:404–412.

27.

Neumayer

, Hosokawa

, Itani

et al. Multivariate predictors of postoperative surgical site infection after general and vascular surgery: Results from the patient safety in surgery study. J Am Coll Surg, 2007; 204:1178–1187.

28.

Rioux

, Grandbastien

, Astagneau

. Impact of a six-year control programme on surgical site infections in France: Results of the INCISO surveillance. J Hosp Infect, 2007; 66:217–223.

29.

Malone

, Genuit

, Tracy

et al. Surgical site infections: Reanalysis of risk factors. J Surg Res, 2002; 103:89–95.

30.

Kaya

, Yetim

, Dervisoglu

et al. Risk factors for and effect of one-year surveillance program on surgical site infection at a university hospital in Turkey. Surg Infect, 2006; 7:519–526.

31.

Barber

, Miransky

, Brown

et al. Direct observations of surgical wound infections at a comprehensive cancer center. Arch Surg, 1995; 130:1042–1047.

32.

Anaya

, Dellinger

. The obese surgical patient: A susceptible host for infection. Surg Infect, 2006; 7:473–480.