Surgical Site Infections in orthopaedic patients – a 10-year retrospective observational study in a Polish hospital

INTRODUCTION Surgical site infections (SSIs) are among the most common healthcare-associated infections. They are associated with longer post-operative hospital stays, additional surgical procedures, treatment in intensive care units and higher mortality. MATERIAL AND METHODS Surgical site infections (SSIs) were detected in patients hospitalized in a 40-bed orthopaedics ward via continuous surveillance in 2009– 2018. The total number of study patients was 15,678. The results were divided into two 5-year periods before and after the introduction of the SSI prevention plan. The study was conducted as part of a national healthcare-Associated Infections surveillance programme, following the methodology recommended by the HAI-Net, European Centre for Disease Prevention and Control Program (ECDC). RESULTS 168 SSIs were detected in total, including 163 deep SSIs (SSI-D). The total SSI incidence rate was 1.1%, but in hip prosthesis: 1.2%, in knee prosthesis: 1.3%, for open reduction of fracture (FX): 1.3%, for close reduction of fracture (CR): 1.5%, and 0.8% for other procedures. 64% of SSI-D cases required rehospitalisation. A significantly reduction in incidence was found only after fracture reductions: FX and CR, respectively 2.1% vs. 0.7% (OR 3.1 95%CI 1.4-6.6, p<0.01) and 2.1 vs. 0.8% (OR 2.4 95%CI 1.0-5.9, p<0.05). SSI-Ds were usually caused by Gram-positive cocci, specially Staphylococcus aureus , 74 (45.7%); Enterobacteriaceae bacillis accounted for 14.1% and Gram-negative non-fermenting rods for 8.5%. implemented demonstrated a significant decrease (about 2.5-3 times) in SSI-D incidence in fracture reductions. Depending on the epidemiological ward, worthwhile to of SSIs associated to different types of orthopaedic surgery to assess the risks and take preventive measures.


INTRODUCTION
In the European Union, there is a great diversity with regard to the practices of control and employment of staff for infection control teams. In many countries, infection surveillance programmes struggle with human resource shortages and strong local cultural conditions [1]. Poland is one of the Central and Eastern European countries and it was only after the political transformation in 1989 that it began to implement a system of surveillance of healthcare-associated infections (HAIs). Prior to that, either at the national or at the acute health care facility levels, there were no organizational structures or mechanisms allowing to detect, qualify or prevent HAIs; and any attempts to start debating this issue in the public sphere that were undertaken by a few NGOs ended in failure. Legal regulations providing the foundations for the emergence of an infection control system did not come into existence until 2001. Therefore, in Poland, the system of surveillance of HAIs is relatively young, and since there is no tradition of HAI control and a deeply rooted uncertainty regarding this field, the situation is not conducive to continuous, active surveillance or even to registration of HAIs. It is also substantiated by Allerberger et al. [2] who wrote that, in Central and Eastern European countries, there is often insufficient information on the surveillance of HAIs as well as publicly available information on epidemiological methods and indicators. This is corroborated by Ider et al. [3] in their study of how infection control systems function in the countries of the post-Soviet bloc. They found weak commitment, lack of resources, poor 4 specialist knowledge and insufficient reporting or publishing of information on HAI epidemiology. Additionally, the countries of Central and Eastern Europe exhibit enormous differences with regard to legislation, structural elements and indicators of the methods for infection control and prevention [2]. Also, according to a report on HAIs by WHO (for 1995WHO (for -2010, in low-and middle-income countries, regular HAI surveillance can prove difficult. In the report, our region, i.e. Central and Eastern Europe, is represented only by Lithuania and Latvia, as well as Serbia (which is not part of the EU) [4]. In Poland, there are no clear or straightforward rules of conduct in surveillance and there is no obligation to provide information on infections to the public, hence, the data concerning Polish hospitals are extremely scarce, which may be a confirmation of the thesis put forward by the authors of the WHO report on difficulties in implementing HAI surveillance, especially active and targeted surveillance.
Zingg et al. have identified ten key elements essential for effective HAI control in the day-to-day practice of every hospital. They are, among others, organization of infection control structures at the hospital level; staff: nurses' workload and forms of employment; correct application of guidelines; education and practice; multimodal and multidisciplinary prevention programmes, positive organizational culture and audit and feedback [5]. Hence, a meaningful impulse for the adoption of activities associated with infection control can arise in the form of the hospital's efforts to carry out accreditation, done by an external entity. In Poland, accreditation and the entire process of preparation for it, consisting in the implementation of standards of service quality and safety of care, is voluntary and free of charge and carried out by the Quality Monitoring Centre in healthcare, a unit of the Ministry of Health.

5
The objective of this study was to analyse the impact of infection control and prevention activities in patients with locomotive organ diseases treated surgically and the implementation of "perioperative control card (perioperative checklist)" which was part of the hospital's preparation for accreditation. The authors and the Infection Control Team of the investigated hospital, in their previous analysis of research material from 2009-2013 [6], indicated a problem of high, higher than expected, SSI incidence rate in surgical patients and the urgent need for action concerning the prevention and control of SSIs. In 2014, the hospital was preparing, for the first time, to undergo the process of accreditation and the Infection Control    Table 1 The list of surgeries and ICD-9 codes.

Code
Operative  [8], which are in line with the ECDC definitions [7,9]. The SSIs were qualified as superficial / deep incisional or bathing the patient prior to surgery with antiseptic soap, changing bed linen and the patient's underwear immediately before surgery, operative field hair removal immediately before surgery using surgical clippers -without the use of blades, surgical hand hygiene according to the WHO recommendations, disposable surgical draping, the application of antiseptic to the edges of the wound before stitching the skin, giving systematic (every 6 months) feedback concerning infections to the department.
The ward was granted a positive Accreditation Certificate in 2014, thus, this year became the basis for the comparison of the results from the period before and after accreditation.
To compare both periods, the choice was made to examine the SSI incidence rate  rods of the family Enterobacteriaceae accounted for 14.1% and Gram-negative nonfermenting rods for 8.5% (Table 3). Table 3 Most On the other hand, the observed total incidence of 1% is a significantly good result considering previous Polish reports concerning SSI in orthopaedic surgery, e.g. in Sosnowiec, the SSI incidence was 6.6% [11], and, in Kraków, 2.6% [12]. However, multicentre data are needed to give a more complete picture of the situation and allow to draw comparisons, because, in view of the fact that data on SSI  [13].
The discrepancies are probably associated with the sensitivity of the method, as the presumed high SSI detection in Norway, but also with the organization of work and the whole system of health care organization as well as the flow of information between different participants of the surveillance systems, which is connected with infection detection in post-discharge care. At the same time, the risk of exposure to SSI following HPRO and KPRO observed in the examined ward for several years reflects the expected level of risk, i.e. it is comparable to the average obtained in the European HAI-Net programme. Also, the microbial aetiology does not differ from the reports of other authors [14].
Unfortunately, the data presented are not so optimistic. Our attention is drawn to the dominance of one of the forms of SSI, that is, the lack of superficial infections, despite the fact that they should make up -in the case of HPRO and KPRO -around 50-60% of cases [14]. At this point, two hypotheses can be made, one suggesting SSI-D overdetection, that is, the tendency to incorrectly classify cases, the other indicating too low detection of superficial SSIs. Therefore, it is very likely that the real SSI incidence is even 2 times higher. This retrospective study had some limitations. Firstly, the research involves only one centre. Secondly, in the period studied, despite participation in the multicenter programme, the surveillance of infection method was not validated, hence, its sensitivity is not known in this particular case.

Conclusion
The introduction of multimodal and multidisciplinary SSI prevention and epidemiology programmes in 2014 has resulted in lower SSI incidence rates in some types of orthopaedic operations. This trend was most strongly visible following FX and CR fracture reductions. After HPRO, KPRO and other procedures, a stable, expected level of SSI incidence was maintained. The results described confirm the possibility of implementing an infection surveillance system also throughout Poland.
Making this system stronger and encouraging its participants to make the results public can reinforce the regional and national surveillance systems. Especially

Declarations
Ethics approval and consent to participate The use of data was approved by the Bioethical Committee of the Jagiellonian University (No. KBET /122.6120.118.2016 from 25.05.2016). All the data entered into the electronic database and analysed in this study was previously anonymised.

Consent for publication
Not applicable Availability of data and materials The datasets generated or analysed during this study are available and can be accessed from Anna Rozanska (e-mail: a.rozanska@uj.edu.pl) on reasonable inquiry.