Numerous studies on the prevalence of HAI and AMU have been conducted worldwide over the past 10 years [3,4,5,6,7,8,9,10,11, 17,18,19,20,21]. However, there are still very few studies that define the study populations by recently having a surgical procedure, and not according to the ward they are hospitalized at on the study day [22]. This is the fourth nationwide prevalence survey of HAI and antibiotic consumption, conducted within the second European PPS, aimed to compare surgical patients with non-surgical patients in Serbian acute-care hospitals. We found that 29% of all adult patients had undergone a recent surgical procedure and the prevalence of HAI in that group of patients was almost two and half times higher than in non-operated patients. The most frequent infections were UTIs and SSIs. At least one antibiotic was taken by 42% of surveyed patients, mainly for the treatment of community-acquired infections (40%) and for surgical prophylaxis (29%).
Firstly, this study finds that according to the ULRA surgery was a significant RF for HAI (p < 0.001, OR: 2.38, 95% CI: 2.16–3.08), but it was not a significant independent RF according to MLRA (OR = 1.01 95% CI 0.829–1.238). The first Singapore PPS showed that surgery since admission detected in 23.7% patients in acute care hospitals and was independent RF for HAI [6], while multicenter PPS, conducted in Switzerland during 2016 showed that NHSN surgery was not independent RF for HAI (ULRA-OR: 1.77, 95% CI 0.89–3.54, p 0.106 and MLRA-OR 1.41, 95% CI 0.75–2.67, p 0.288) [19].
Secondly, in Serbian acute adult patients, the prevalence of HAI was 7.2% and 2.9% in surgical and non-surgical patients respectively. National PPS of HAI and antimicrobial prescribing done in Scotland noted similar results (4.0% vs. 6.5%) [20]. This characteristic observed across all four hospital size categories (Fig. 1) is not a new phenomenon and has been reported before [22]. The reasons for the difference are likely to be multifactorial. Sax et al. showed, importantly, that patients admitted to larger hospitals had a greater number of comorbidities. Also, prolonged hospital and ICU stay, medical devices and drugs use were more frequent in larger hospitals [23]. One of the reasons for the higher HAI prevalence in surgical patients would be the occurrence of SSI, the second most frequent HAI in Serbian PPS, with a prevalence of 3.4%. Furthermore, the most frequent HAI in Serbian patients were urinary tract infections (UTI) (130/548; 23.7%), SSIs (125/548; 22.8%) and pneumonia/lower respiratory tract infections (113/548; 20.6%). On the contrary, infections not associated with devices or operative procedures—including C. difficile infections and other gastrointestinal infections, and non-ventilator-associated pneumonia—accounted for approximately half of all HAI in prevalence surveys conducted in US hospitals in 2015 [3]. This distribution of HAI, different from the one established by a survey conducted in 2011, was explained by Magill et al. with experience which had shown that HAI, specifically UTI and SSI, could be prevented by national public-health focus and evidence-based interventions [3, 24, 25]. The significant differences in the prevalence of UC, CVCs, and the use of MV between both groups of our patients are shown in Table 1. MLRA revealed that the presence of invasive medical devices (UC, CVC, and MV) was independent RF for HAI in Serbian acute adult patients. To stop the spread of HAIs, more attention needs to be paid to the role of invasive medical devices. Implementing a quality-management system seemed like the most effective way to prevent a significant number of these infections. WHO encouraged countries with a weak quality-management system to hekp healthcare professionals use invasive medical devices in a manner that is fair, consistent, and effective [26].
Thirdly, the highest prevalence of all HAIs was noted in patients who had kidney transplantation (36.4%). This is an expected result because immunosuppression is one of the most important RF for infections, including HAI in this group of patients. Monlezun et al. [27] reported that 48.7% of renal transplant patients experienced at least one post-transplant infection. Similar findings were reported by U.S. Renal Data System which showed that UTIs were the most common bacterial infections requiring hospitalization in kidney transplant recipients, followed by pneumonia, SSI, and bloodstream infections [28]. Moreover, SSIs were the most prevalent among patients who had peripheral vascular bypass surgery (20.0%). This is not a surprising result since, in a few prospective analytic and experimental studies, SSI was identified as one of the most common postoperative complications after vascular reconstruction, particularly lower extremity bypass procedures [29, 30]. The published incidence of SSI after lower extremity bypass procedure varied from 4.8 [31] to 22.8% [32].
In Ghana's PPS for only 10% of SSI, the microorganism was reported with a dominance of gram-negative bacteria, while in Singapore’s PPS the most frequently identified bacteria in SSI were S. aureus (17.5%), followed by P. aeruginosa (14.6%), Escherichia coli (9.5%) and Acinetobacter spp. (4.4%) [6]. We registered microorganisms in 84.8% of SSI. Also, we observed different distribution of bacteria causing SSI—Acinetobacter spp. (14.7%), Klebsiella spp. (11.1%), S. aureus (10.6%) and P. aeruginosa (9.6%).
The third Slovenian national HAI PPS detected C. difficile gastrointestinal infections in almost half of the identified gastrointestinal infections [21]. Altogether, our data show that gastrointestinal HAIs were in the fourth place in frequency (13.8% of all HAI) and that there was no difference in the prevalence of C. difficile gastrointestinal infections in patients who recently had a surgery compared with non-operated patients (OR = 1.55 95% CI = 0.84–2.8, p = 0.163). Some studies provided evidence that although surgical patients tend to suffer more severe CDIs than medical patients, overall they still do better than medical ones [33, 34].
The present study found that exposure to at least one antimicrobial agent was registered in 42.4% of hospitalized patients. That was lower than the antimicrobial use prevalence reported in China's prevalence survey (49.63%) [7] and higher than reported in Japan's prevalence survey (33.5%) [5]. Compared to other PPS in European countries, Serbian patients received at least one antimicrobial agent (42.4%), which is below average values for hospitalized patients in Bulgaria (45.2%), Cyprus (45.8%), Greece (55.6%), Italy (44.5%), and Spain (46.3%), but above the EU average (30.5%) [12]. Similar to the study performed by Sax et al. [22], in our study surgical patients received antimicrobial agents significantly more frequently than non-surgical patients. Moreover, non-surgical patients received antimicrobial agents significantly more frequently for community-acquired infections (64.3% vs. 12.1%). We can assume that a higher rate of antibiotics prescriptions for community-acquired infections is a consequence of the fact that significantly higher proportions of non-surgical patients (65%) were patients over the age of 60 with comorbidities and a clinical history of the infection treatment before the hospitalization. The second PPS of HAI and antimicrobial use in European acute-care hospitals showed that the most common indication for prescribing antimicrobials was the treatment of community-acquired infection in 69.8% [12].
Antibiotics prophylaxis is defined as the administration of a single dose of the effective antimicrobial agent prior to the exposure with possible contamination, i.e., surgery, to decrease the risk of postoperative infections [35]. A single dose of antibiotic is recommended to be given 120 to 60 min before surgical incision. Moreover, the recently published guidelines of the Centers for Disease Control and Prevention (CDC) restrict any additional prophylactic antibiotics after the completion of the clean and clean-contaminated surgical procedures [36]. In contrast to data from UK Scotland, Finland, Luxemburg, Germany, Austria [12], more than 70% of our surgical patients received surgical prophylaxis for longer than 1 day.
Regarding the antibiotics administration dosing regimen and duration of antibiotics prophylaxis, the appropriate antimicrobial selection also plays a role in minimizing possible risks of bacterial resistance and C. difficile gastrointestinal infections. It has a secondary favorable effect of potentially minimizing infection rates. Cefazolin, a first-generation cephalosporin, provides adequate coverage against most of the organisms causing postoperative infections in patients with no history of beta-lactam allergy, and a history of MRSA infection. It causes minimal allergic reactions and side effects, achieves adequate tissue levels, and is relatively inexpensive. The presence of these benefits makes cefazolin the most appropriate perioperative antibiotic prophylaxis agent for the majority of surgical procedures [35]. Overall, cefazolin was prescribed only in 18.4% of surgical patients as perioperative surgical prophylaxis. The most frequently prescribed antibiotics in both groups of our patients was ceftriaxone, a third-generation cephalosporin belonging to the “watch group” by WHO. Plachouras et al. [12] reported that ceftriaxone was in third place of the list of antimicrobial agents accounting for 75% of antimicrobial use in acute-care hospitals in the European Union/European Economic Area countries during 2016–2107, with the highest consumption registered in Bulgaria, Romania, and Serbia.
This survey has its limitations. Firstly, data collection was done by local infection-prevention and control professionals and not by the study team. However, all participating data collectors were trained by the PPS coordination committee of the Ministry of Health of Serbia before data collection and most of them had experience with performing local PPSs in the past. Secondly, we did not derive the local HAI incidence from our prevalence results as the routine applicability of the Rhame and Sudderth formulae [37]. The strength of our study is that the data are representative of Serbia as the data from 61 hospitals for adult acute-care were included. Additionally, the data on the use of antibiotics and HAI from our study contributed to the quality of the National guidelines for the rational use of antibiotics published in 2018 by the Ministry of Health of the Republic of Serbia. We hope that the implementation of the National guidelines will significantly improve antibiotics prescription politics and reduce antimicrobial resistance in the future.