Identifying the Infection Control Areas Requiring Modifications in Thoracic Surgery Units: Results of a Two-Year Surveillance of Surgical Site Infections in Hospitals in Southern Poland

Abstract

Background:

Surgical site infection (SSI) is considered to be a priority in infection control. The objective of this study is the analysis of results of active targeted surveillance conducted over a two-year period in the Department of Thoracic Surgery at the Pulmonology and Thoracic Surgery Center in Bystra, in southern Poland.

Patients and Methods:

The retrospective analysis was carried out on the basis of results of active monitoring of SSI in the 45-bed Department of Thoracic Surgery at the Pulmonology and Thoracic Surgery Center in Bystra between April 1, 2014 and April 30, 2016. Surgical site infections were identified based on the definitions of the European Centre for Disease Prevention and Control (ECDC) taking into account the time of symptom onset, specifically, whether the symptoms occurred within 30 d after the surgical procedure. Detection of SSI relied on daily inspection of incisions by a trained nurse, analysis of medical and nursing entries in the computer system, and analysis of all results of microbiologic tests taken in the unit and in the operating room.

Results:

In the study period, data were collected regarding 1,387 treatment procedures meeting the registration criteria. Forty cases of SSI were detected yielding an incidence rate of 3%. Most cases (55%) were found in the course of hospitalization and 45% were detected after the patient's discharge. The SSIs were classified as follows: superficial, 37.5%; deep infections, 7.5%; and organ/space infection, 55%. Among patients who were diagnosed with SSI, most were male (77.5%). For patients with an American Society of Anesthesiologists (ASA) score I–II the incidence rate was 2%; ASA score III or more, 3.7%. The incidence rate varied from 0.3% in clean surgical site to 6.5% in clean–contaminated site.

Conclusions:

The study validated the usefulness of targeted surveillance in monitoring SSIs in patients hospitalized in thoracic surgery departments. Surgical site infection surveillance identified areas of care requiring modifications, namely, organization of post-discharge and microbiologic diagnostics of infection cases.

Surgical procedures disturb the body's natural protective barriers and are associated with the risk for developing surgical site infection (SSI). This risk depends on a range of factors associated with the patient (age, gender, comorbidities, nutritional status), surgical procedure (type, extent, cleanliness of the surgical site), pre-operative (peri-operative prophylaxis, surgical site preparation), and post-operative (drainage, dressings, compliance with the aseptic technique protocols in surgical wound care) proceedings. Interest in the problem of surveillance of this form of infection has continued uninterrupted for many years, both globally [1] and in Poland [2].

In recent years, studies on the epidemiology and microbiology of SSIs in Poland were conducted in implant orthopedics, vascular surgery, obstetrics (cesarean deliveries), and with regard to certain procedures in general surgery [3,4]. These studies, based on the surveillance in one or two centers, indicated differences between morbidity rates in the examined populations of Polish patients and in studies from other countries. Additionally, ineffective surveillance of SSI was also found after the patient's discharge, especially after cesarean deliveries or in vascular surgery [3,4].

Surgical site infection epidemiology and microbiology, as well as the effectiveness of their treatment in Polish patients after thoracic surgery, have not been described separately in detail to date, despite the fact that surveillance of SSI is considered to be a priority in infection control. This is demonstrated by the fact that there is no Polish participation and data in European surveillance of SSI reports [5]. Detection, registration, and analysis of data and making data available as feedback to staff providing direct care for patients are essential elements to assess effectiveness of epidemiologic and microbiologic surveillance, and therefore, safety and optimization of treatment costs [6]. The objective of this study is the analysis of results of active targeted surveillance conducted over a two-year period in the Department of Thoracic Surgery at the Pulmonology and Thoracic Surgery Center in Bystra, Poland.

Patients and Methods

The retrospective analysis was carried out on the basis of results of prospective active targeted surveillance of SSI in the 45-bed Department of Thoracic Surgery at the Pulmonology and Thoracic Surgery Center in Bystra, Poland, between April 1, 2014 and April 30, 2016. Pneumonia and urinary tract infections were detected and monitored by passive surveillance. The number of admissions to the unit is approximately 1,800–1,900 annually; bed occupancy in the analyzed period was 65%. The staff employed in the unit comprises seven full-time physicians and two residents, 25 full-time nurses, and one part-time physiotherapist. The Infection Control Team consists of a physician who specializes in medical microbiology (1/5 full-time equivalent) and a full-time epidemiologic nurse.

The registry of the performed treatments and surgical procedures was kept on the basis of anesthesia cards. The registry comprises all treatments carried out under general anesthesia or general anesthesia plus epidural. The registry did not contain information on minor procedures under local anesthesia (chest tube, lymph node biopsy, lung biopsy, samples from superficial lesions of the chest wall). The registry accumulated data such as the number of medical history; age and gender of the patient; date of admission; date, type, and duration of surgical procedure; peri-operative risk in the five-point scale of the American Society of Anesthesiologists (ASA); type of anesthesia; and the use of peri-operative prophylaxis.

Procedures in the clean–contaminated site were considered to be all operations during which the airway has been opened (trachea, bronchus). Operations in the contaminated site were considered surgeries in patients with lung abscess or a cancerous tumor with signs of disintegration, if during the operation there was a tumor/abscess perforation, regardless of whether microbiologic examination was performed and what its outcome was. Dirty site procedures were the ones in patients with infection (most often pleural empyema) and a positive result of microbiologic examination of the material collected intra-operatively from the surgical site. All patients underwent peri-operative prophylaxis consisting of intravenous administration of a single dose of 2 g cefazolin less than 30 min preceding the skin incision (the antibiotic is administered during surgical site skin disinfection). Patients who are allergic to cephalosporins or have a severe allergic reaction to β-lactams in the history were given a single dose of 600 mg clindamycin in a 30-min intravenous infusion, completed before the skin incision. The average duration of a procedure was calculated based on 30 randomly selected treatments of a given type. Bronchi were normally closed with the use of a stapler. Surgical site infections and other infections were identified based on the definitions of the European Centre for Disease Prevention and Control (ECDC) taking into account the time of symptom onset, namely, if the symptoms occurred in the 30 d after the surgical procedure [7].

Detection of SSI relied on daily inspection of wounds by a trained nurse, analysis of medical and nursing entries in the computer system, and analysis of all results of microbiologic tests taken in the unit and in the operating room. Medical records were also reviewed in the case of patients who visited the hospital outpatient thoracic surgery clinic in the 30-d period after surgery on the basis of the list of control visits prepared by the clinic staff. Only 30% of all patients were covered by the surveillance in outpatient local clinic (post-discharge); the remainder did not come to control visits in the hospital. Microbiologic tests were performed when ordered by the attending physician. These were swabs from the incision if it showed signs of superficial infection or fluid from the pleural cavity in the case of pleural empyema.

The decision whether to implement antibiotic treatment is made by the attending physician. In the hospital, there are guidelines regarding empiric antibiotic therapy prepared by the Infection Control Team. The guidelines are updated every other year based on the results of microbiologic tests carried out in the unit. In the analyzed period, the recommended empirical SSI treatment was combination therapy of ceftriaxone with clindamycin or metronidazole or monotherapy with piperacillin-tazobactam.

At the conclusion of the two-year monitoring period, validation was performed through analysis of medical records for all cases registered as SSI to detect incorrect or missing entries. In cases in which such data were unavailable, information was verified concerning further infection treatment and its outcome.

In statistical analysis, t-test was applied for two independent variables and χ² test with the use of commercially available calculators (www.socscistatistics.com). The study was approved by the Bioethics Committee of Jagiellonian University Medical College, decision no. 122.6120.118.2016 dated May 25, 2016.

Results

In the analyzed period, data were collected regarding 1,387 treatment procedures meeting the registration criteria described above. The majority of patients were male (838; 60.5% of all operated subjects). The average patient age was 60 y (19–86 y). The general status of patients was described according to ASA score as: I or II in 604 patients (43.6%); III or IV in 761 patients (56.4%); and V in one operated patient (< 0.1%). Procedures carried out in clean surgical sites comprised 55% of all treatments (745 operations); in clean-contaminated sites, 40% (567 treatments); in contaminated sites, 3% (44 procedures); and in a dirty site, 2% (31 operations). Detailed data are presented in Table 1. The highest SSI incidence was found after pneumonectomy: 27.1%; nine infections were pleural empyema, four of these patients had bronchopleural fistula.

Table 1.

Characteristics of Patients with and without Infections

	Patients
Characteristics	With infections n = 40	Without infections n = 1347	OR (95%CI)	p
Age [years], median (quartiles: 1Q; 3Q)	65 (58.5; 67.3)	62 (55.0;68.0)	NA	0.17
Average length of hospitalization [days]	22.6	10.6	NA	< 0.01
Gender
Female	9	540	0.4 (0.20–0.92)	0.02
Male	31	807	0.4 (0.20–0.92)	0.02
Wound microbiology contamination
Clean/clean–contaminated	33	1,273	0.3 (0.12–0.64)	0.01
Contaminated/dirty	7	74	0.3 (0.12–0.64)	0.01
ASA score
ASA I +ASA II	12	594	0.5 (0.27–1.08)	0.07
ASA III	26	680	1.8 (0.94–3.50)
ASA IV + ASA V	2	73	0.9 (0.22–3.88)
Pre-operative stay
0–2 d	2	571	0.01 (0.02–0.39)
3 d or more	28	746	0.01 (0.02–0.39)

OR = odds ratio; CI = confidence interval; NA = not available; ASA = American Society of Anesthesiologists.

The time from the patient's admission to surgery ranged from two to 15 d (average of four days). The period from administering the antibiotic to skin incision was 5–30 min (10 min on average) and was calculated based on 30 randomly selected surgical procedures. After analyses performed periodically by the Infection Control Team, proper compliance with peri-operative prevention procedures (understood as the administration of appropriate antibiotics, in the correct dose and at appropriate times) is 98.3%. The duration of the hospital stay was 6–17 d (10.6 d on average) and was calculated for 30 randomly selected patients undergoing elective surgery who did not demonstrate infectious complications in the post-operative period.

As a result of conducting active surveillance, 40 cases of SSI were demonstrated, which is 3% of the recorded surgical procedures. Twenty-two infections (55%) were found in the course of hospitalization and 18 infections (45%) were detected after the patient's discharge from hospital. Thirteen patients from this group required readmission. Fifteen cases (37.5%) were classified as superficial infections, three cases (7.5%) as deep infection, and 22 cases (55%) as body cavity infection (pleural empyema). Among patients who were diagnosed with SSI, most were male (31; 77.5%); average patient age was 63 y. Infection was determined in 12 patients with ASA score I–II (incidence, 2%) and in 28 patients with ASA score III or more (incidence, 3.7%). The incidence rate was 0.3% in clean surgical site and 6.5% in clean–contaminated site.

Eighteen patients (45%) who were diagnosed with infection had information in their records concerning the occurrence of prolonged air leak and/or the (suspected) presence of bronchopleural fistula. A prolonged air leak is one that requires pleural drainage lasting more than five days after the surgical procedure [8]. Table 2 provides data regarding duration of surgery in patients with SSI.

Table 2.

Duration of Surgical Procedures in Infection versus Non-Infection Cases

		With infections			Without infections
		Duration of the operation			Duration of the operation
Type of procedure	No.	Mean [min]	Range [min]	No.	Mean [min]	Range [min]
Wedge resection	4	125	60–160	470	60	20–190
Lobectomy	19	100	65–250	418	105	40–265
Pneumonectomy	13	110	65–250	35	130	70–305
Mediastinal tumor	1	60	NA	55	80	25–200
Chest wall tumor	0	n/a	NA	24	70	15–145
Exploratory thoracotomy	0	n/a	NA	35	60	25–200
Decortication	1	70	NA	43	100	35–180
Mediastinoscopy	0	n/a	NA	47	45	15–95
Diagnostic videothoracoscopy	1	25	NA	118	30	15–90
Other	1			102
Total	40			1347

NA = not available.

Infections registered in the studied period based on passive reporting, results of microbiologic testing, and audits in the ward were: six respiratory tract infections, including only two that met the definition of pneumonia according to ECDC; three Clostridium difficile infections; four urinary tract infections confirmed microbiologically; and three blood stream infections (including one associated with central line).

Results of microbiologic testing were available for 32 patients. Gram-negative bacilli (Enterobacteriaceae) and anaerobic bacteria were isolated most often (19% for each group), followed by coagulase-negative staphylococci (16%) and enterococci (16%). Additionally, Streptococcus pneumoniae (6%), Moraxella catarrhalis (3%), and Staphylococcus aureus (3%) were isolated in SSI cases. In five cases, more than one bacteria species was isolated (16%). Among Enterobacteriaceae, the frequency of production of extended spectrum β-lactamases (ESBL) was 17%. All isolated strains of coagulase-negative staphylococci were methicillin-resistant. There was no vancomycin resistance among the Enterococcus strains isolated, and the only strain of Staphylococcus aureus isolated was sensitive to methicillin. One of the two Streptococcus pneumoniae strains isolated demonstrated reduced susceptibility to penicillin (minimum inhibitory concentration [MIC] for penicillin, 8; for ceftriaxone, 2). In one of the reported cases of superficial SSI caused by a typical respiratory pathogen, Streptococcus pneumoniae, the same microbe was cultured from intra-operative material (cancerous tumor with signs of disintegration) that confirms the significance of contamination of the incision in the development of infection. Two cases of patients with organ/space infection showed first cultures to be negative, but subsequent ones taken after several days demonstrated Enterococcus faecalis growth.

In different types of SSI, various microbial causes were found, however, the difference was not statistically significant. In superficial surgical site infections (SSI-S), gram-positive cocci were predominant (70% of cultures) followed by anaerobes (30% of cultures, always Brevibacterium sp.). In the remaining ones, gram-negative bacilli were predominant: in deep surgical site infections (SSI-D) only one positive culture was obtained (Brevibacterium sp.), and in organ/space surgical site infections (SSI-O) there were six gram-positive cocci per 12 cultures (including four Enterococcus sp. and one Streptococcus pneumoniae), four gram-negative bacilli, two anaerobes (Fusobacterium sp., Prevotella sp.), and one Moraxella catarrhalis.

We decided to exclude the data obtained on the applied antibiotic therapy from further analysis because the intention to include an antibiotic did not always result clearly from the entries in patient records. Antibiotic therapy was often begun even before diagnosing SSI. In many cases, the results of antibiotic treatment were not documented.

The duration of hospitalization in patients who were diagnosed with infection during their hospital stay was from 15 to 74 d (22.6 d on average) and was longer than the duration of hospitalization of patients who had no infectious complications. For patients who were diagnosed with infection after discharge, the time from discharge to recording the infection (date of visiting the clinic when the infection was diagnosed or the date of re-admission to hospital because of symptoms of infection) ranged from five to 59 d (20 d on average). In two cases, this time exceeded one month (37 and 59 days). However, the information obtained from patients showed that the first symptoms of infection such as fever or pus leaking from the incision developed within 30 d after surgery. Fifteen patients with diagnosed SSI (37.5%) required at least one additional surgical procedure (decortication, seton stitch, thoracostomy). Six patients were diagnosed eventually with chronic pleural empyema and underwent to thoracostomy. In-hospital mortality in patients diagnosed with SSI was 7.5%.

Based on the results obtained and the surveillance conducted, modifications were implemented in January 2017 regarding the work of several groups of staff, which aim at improving the sensitivity of infection supervision and selected elements of antibiotic stewardship, i.e., introduction of the Southampton Wound Scoring System to objectify the assessment of wound healing, and a trained nurse can now draw samples for microbiologic testing on his/her own without prior doctor's order. Antibiotic therapy is reviewed by the infection control doctor.

Discussion

The data available in the scientific literature on hospital-acquired infections in patients undergoing thoracic procedures are surprisingly scarce. In a single-center prospective study in a university hospital in Spain, the overall incidence of hospital-acquired infections after surgical lung resection was 25.8% of operated patients, and at the same time, SSIs were 21% of the registered cases. One-third of the SSIs were detected after the patient's discharge [9]. Another prospective study demonstrated the overall incidence of infections at the level of only 4.9%, and here, the incidence of SSIs was 1.7%. An insignificantly higher incidence of infections in thoracotomy was found in comparison with video-assisted thoracoscopic surgery, and the only important factor associated with the occurrence of SSI was the low value of forced expiratory volume [10]. In a study by Imperatori et al. [11], the incidence of surgical incision line infections was 3.2% and the incidence of pleural empyema was 2.4%. Increased risk of infection was associated with the duration of the surgical procedure longer than 180 min, age greater than 70 y, baseline albumin <3.5 g/dL, and the presence of any additional disorder. Low forced expiratory volume timed (FEV₁) concentrations predisposed to developing pneumonia but not SSI [11]. In research by Nadir et al. [12], frequency of wound dehiscence after thoracotomy was high (6.6%) and most of these patients were diagnosed with SSI. The microbe isolated most frequently was Staphylococcus epidermidis (54%) [12].

An active program of surveillance of SSI carried out by the authors allowed the incidence of these infections in our center to be determined as similar to that of other centers, but most (58%) were deep and organ/space infections, which require hospital treatment. The risk of SSI was associated primarily with the potential site contamination during the opening of the bronchi (procedures in the clean–contaminated site). It appears that body cavity infections are often secondary to the presence of a bronchial fistula or prolonged air leak. The risk of the occurrence of such a fistula is increased in patients who underwent right-sided pneumonectomy and in mechanically ventilated patients in the post-operative period [13]. The fistula incidence is also influenced by the choice of a surgical technique, location of tumor, and prior use of induction therapy [14,15]. Pneumonectomy appeared to be the most relevant risk factor, probably the result of a combination of factors including contamination from the airway (post-operative presence of a large pleural space that does not tolerate contamination well) and increased possibility of bronchopleural fistula compared with lobectomy. All pneumonectomy procedures in our study were performed because of cancer, which greatly increased the rate of bronchopleural fistula and empyema in the procedure.

In the analyzed group of patients, the statistically significant risk factors were being male and operation in a surgical site other than microbiologically clean. Reports concerning infections of the Polish population confirm the statistically higher risk of infection in males [16]. Being male can also mean an increased risk of diagnosing post-discharge infection (17).

Smoking is a reported SSI risk factor taken into account in the recommendations concerning the prevention of this type of infection. In the research by Kotseva et al. [18], it was demonstrated that males smoke more cigarettes and control their blood pressure less frequently, which could determine a lower tendency for self-prophylaxis and abnormal general health-oriented behaviors.

Almost half of all SSIs (45%) were detected after discharge, however, it is difficult to estimate precisely the actual incidence of post-discharge infections because only less than 30% of patients register at the local outpatient clinic in the post-operative period. The specificity of thoracic procedures, in particular the procedures involving lung tissue resection, results in the fact that pulmonary complications seem to have greater importance for the final outcome of therapy. Polish data from the years 2002–2006 regarding post-procedure pneumonia (PPP) after surgical procedures in thoracic surgery indicate the incidence of 2.7%, with mortality of 12%. A substantial proportion of these cases (75%) required hospitalization in intensive care units [19]. A much higher risk of PPP is associated with large resection procedures (lobectomy, pneumonectomy). In a French study, among 168 patients who underwent such procedures, the PPP incidence was 25% and constituted the major cause of death in the postoperative period. Chronic obstructive pulmonary disease (COPD), the scope of resection, respiratory tract colonization at the time of surgery, and being male were independent risk factors for the occurrence of PPP [20]. Of a variety of factors predisposing one to respiratory tract infection, colonization of the respiratory tract by numerous micro-organisms is cited, which is a common phenomenon among patients undergoing thoracic surgery [21 –23]. These findings were even the cause of debate about whether peri-operative prophylaxis in thoracic surgery should be aimed at preventing SSI or respiratory complications [24].

Unfortunately, there was low incidence of infections other than SSI in the examined material, however, this was probably because of low detection of such infections, especially pneumonia. The primary reason was passive supervision of such infections, but the specificity of the studied group of patients is also significant. Consequently, there are objective issues in the detection of such infections, such as low sensitivity of radiography in differentiating inflammatory and atelectatic changes, which are typical of these patients because they result from surgical damage to the lung tissue, pleural effusion after surgery, etc. Hence, ECDC criteria, which take into account radiography, may prove to be useless in this patient population.

The presenting study is a unique study on SSIs in thoracic surgery with respect to the Polish patients. The results of the analysis allow us to define the strengths and weaknesses of infection surveillance and its priorities for the future. This analysis, although coming from a single center, is all the more important in that in recent ECDC reports (years 2010–2011 and 2014) concerning surveillance of SSI with respect to the recommended, frequently conducted surgical procedures such as coronary artery bypass grafting (CABG), knee and hip joint endoprosthesis, cesarean deliveries, intestine surgeries, cholecystectomy, and laminectomy, there is a lack of data from Poland [5]. In CABG procedures, the rate of cumulative incidence reported by ECDC was 3.5%. These were infections recorded in approximately 60% of cases during hospitalization, and their predominant etiologic agents were gram-positive cocci (61%) and bacilli from the family Enterobacteriaceae (22%) [5]. Among the other procedures covered by surveillance in the Healthcare-associated Infections Surveillance Network (HAI-Net) system, only intestinal procedures presented higher rates of morbidity (9.5%).

The data presented indicate that continuous targeted surveillance of infections is also possible in Polish hospitals, however, choosing the scope of surveillance should be approached carefully. Regarding patients undergoing thoracic surgery, it is advisable to extend the scope of monitoring by at least one other clinical form of infection, which is PPP. A negative trend is the observed tendency to contract services with external microbiologic laboratories, which may influence the effectiveness of surveillance through the deterioration of communication between the microbiologist and the department staff and impact the reduction in reliability of data concerning the etiology of infections. As a result of the analysis conducted, we decided to implement a combination of measures designed to expand surveillance of infections in our center with post-operative pneumonia, improve the quality of microbiologic diagnostics, and proper compliance with the principles of rational antibiotic therapy.

However, the present study has some limitations. The first is that it was conducted in a single center, therefore the relatively small number of registered infections and heterogeneity of the groups (type of surgery, type of infection). Another limitation may also be reporting that takes into account the specific closed set of risk factors, which stems from using the protocol according to the ECDC. Consequently, it is not possible to conduct a more detailed analysis taking into account risk factors such as malnutrition or diabetes mellitus. However, the universal ASA score was applied for the study as peri-operative risk assessment taking into account morbidities, including chronic diseases such as diabetes mellitus.

Prospective analysis was also not performed for the incidence of other types of infection in patients after thoracic surgery, particularly PPP.

Additionally, the data concerning the cause of SSIs must be considered as of little credibility, because most patients were receiving antibiotics even prior to microbiologic testing. It was usually included because of the fever observed or increases in inflammatory parameters signaling the occurrence of SSI. This explains why almost one fifth of the performed microbiologic tests did not give pathogenic flora growth. Moreover, in eight cases (20%) microbiologic examinations were not conducted, meaning that the use of microbiologic diagnostics is still not optimal. With body cavity infection (pleural empyema) secondary to bronchopleural fistula, the etiologic factor that is undoubtedly significant can be the flora colonizing the respiratory tract. The low proportion of Staphylococcus aureus as the cause of SSIs can be associated with the effectiveness of peri-operative prophylaxis with cefazolin of potent anti-staphylococcal activity.

Conclusions

The study validated the usefulness of targeted surveillance in monitoring surgical site infections in patients hospitalized in the cardiothoracic surgery department. Low effectiveness of post-discharge surveillance has been demonstrated, therefore, there is a need to modify SSI registration in outpatient care. High incidence of SSI after pneumonectomy should be further investigated. Active surveillance should also be considered for other forms of infection in this patient population, especially for pneumonia. However, ECDC criteria for pneumonia identification should be modified for this patient population.

The most relevant risk factor for developing SSI after thoracic surgery is surgical field cleanliness. Other risk factors identified are being male and high ASA score. A weak point in SSI surveillance was determined to be microbiologic diagnostics of these infections.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Culbertson

, Altemeier

, Gonzales

, et al. Studies on the epidemiology of postoperative infection of clean operative wounds. Ann Surg, 1961; 154:599–610.

Kryński

, Borowski

, Becla

, et al. Studies on the epidemiology of staphylococcal infections in surgical clinics. Role of the clinical personnel in spreading of intra-hospital infections. Przegl Epidemiol, 1961; 15:135–141.

Wójkowska-Mach

, Batycki

, Hulbój

, et al. Hospital-acquired infections after caesarean delivery in selected hospitals in the southern Poland. Ginekol Pol, 2008; 79:536–543.

Wójkowska-Mach

, Bulanda

, Cencora

et al. Surgical site infections following vascular burgery, Prz Epidemiol, 2007; 61:683–691.

http://ecdc.europa.eu/en/publications/Publications/SSI-in-europe-2010-2011.pdf (Last accessed July 14, 2017).

Council recommendation of 9 June 2009 on patient safety, including the prevention and control of healthcare associated infections.. Official Journal of the European Union 2009/C 151/01.

European Centre for Disease Control and Protection.. Surveillance of surgical site infections in European hospitals—HAISSI protocol. Version 1.02. http://ecdc.europa.eu/en/publications/120215_ted_ssi_protocol.pdf (Last accessed July 14, 2017).

Burt

, Schrager

. Prevention and management of postoperative air leaks. Ann Cardiothorac Surg, 2014; 3:216–218.

Nan

, Fernandez-Ayala

, Farinas-Avarez

, et al. Nosocomial infection after lung surgery: Incidence and risk factors. Chest, 2005; 128:2647–2652.

10.

Rovera

, Imperatori

, Militello

, et al. Infections in 346 consecutive video-assisted thoracoscopic procedures. Surg Infect, 2003; 4:45–51.

11.

Imperatori

, Rovera

, Rotolo

, et al. Prospective study of infection risk factors in 988 lung resections. Surg Infect, 2006; 7(Suppl 2):57–60.

12.

Nadir

, Kaptanoglu

, Sahin

, et al. Post-thoracotomy wound separation (DEHISCENCE): A disturbing complication. Clinics, 2013; 68:1–4.

13.

Sirbu

, Busch

, Aleksic

, et al. Bronchopleural fistula in the surgery of non-small cell lung cancer: Incidence, risk factors and management. Ann Thorac Cardiovasc Surg, 2001; 7:330–336.

14.

Sonobe

, Nakagawa

, Ichinose

, et al. Analysis of risk factors in bronchopleural fistula after pulmonary resection for primary lung cancer. Eur J Cardiothorac, 2000; 18:519–523.

15.

Panagopoulos

, Apostolakis

, Koletsis

, et al. Low incidence of bronchopleural fistula after pneumonectomy for lung cancer. Interact Cardiovasc Thorac Surg, 2009; 9:571–575.

16.

Pobiega

, Myjak

, Pomorska-Wesolowska

, et al. Virulence potential of Staphylococcus aureus strains isolated from diabetic foot ulcers among patients from southern Poland. Curr Vasc Pharmacol, 2016; 14:547–551.

17.

Woz

, Mitchell

, Hesko

, et al. Gender as risk factor for 30 days post-discharge hospital utilisation: A secondary data analysis. BMJ Open, 2012; 2:e000428.

18.

Kotseva

, de Bacquer

, Jennings

, et al. Time trends in lifestyle, risk factor control, and use of evidence-based medications in patients with coronary heart disease in Europe: Results from 3 EUROASPIRE Surveys, 1999–2013. Global Heart 2016. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.gheart.2015.11.003 (Last accessed July 14, 2017).

19.

Wójkowska-Mach

, Różańska

, Bulanda

, et al. Nosocomial perioperative pneumonia. Analysis of the National Nosocomial Infections Surveillance System of the Polish Society of Nosocomial Infections. Anaesthesiol Intensive Ther, 2007; 1:26–31.

20.

Schussler

, Alifano

, Dermine

, et al. Postoperative Pneumonia after Major Lung Resection. Am J Respir Crit Care Med, 2006; 173:1161–1169.

21.

Kim

, Kim

, Kang

, et al. Prevalence of and risk factors for postoperative pulmonary complications after lung cancer surgery in patients with early-stage COPD. Int J Chron Obstruct Pulm Dis, 2016; 11:1317–1326.

22.

D'Journo

, Rolain

, Doddoli

, et al. Airways colonization in patients undergoing lung cancer surgery. Eur J Cardiothorac Surg, 2011; 40:309–321.

23.

Yamada

, Sekine

, Suzuki

, et al. Trends of bacterial colonization and the risk of postoperative pneumonia in lung cancer patients with chronic obstructive pulmonary disease. Eur J Cardiothorac Surg, 2010; 37:752–757.

24.

Schussler

, Dermine

, Alifano

, et al. Should we change antibiotic prophylaxis for lung surgery? Postoperative pneumonia is the critical issue. Ann Thorac Surg, 2008; 86:1727–1733.