Background
Antimicrobial resistance (AMR) is a severe threat to public health in Europe and worldwide, leading to growing costs, treatment failure, and mortality [
1,
2]. AMR results in reduced efficacy of drugs, and limits the available treatment options. The magnitude of the problem worldwide, the impact of AMR on health and on costs for the healthcare sector, together with the societal impact, are still unknown [
2]. Thus, surveillance of AMR is considered an essential component of an effective response to this problem, and results produced constitute a fundamental source of information on the burden and trends of resistance [
1]. Unfortunately, high resistance rates (RRs) to antimicrobial agents have been observed among bacterial pathogens that cause healthcare-associated and community-acquired infections worldwide and significant gaps in surveillance, lack of standards for methodology, data sharing and coordination have been highlighted [
2]. Especially Gram-negative bacteria, including multidrug resistant
Acinetobacter baumannii and
Enterobacteriaceae producing Extended-spectrum beta-lactamases (ESBL) and carbapenemases, have been associated to severe healthcare-associated infections and their occurrence has increased in the last decades [
1,
3]. In Europe, the most recent report of the European Antimicrobial Resistance Surveillance Network (EARS-Net) describes a general increase of AMR in the Gram-negative pathogens including
Klebsiella pneumoniae and
A. baumanni [
1]. In Italy, the results of the Italian Nosocomial Infections Surveillance in Intensive Care Units (SPIN-UTI) network revealed that
K. pneumoniae and
A. baumanni are among the most commonly isolated microorganisms in intensive care unit (ICU)-acquired infections [
4‐
6].
The overuse of antimicrobials is one of the main factors responsible for the development and spread of AMR. Therefore, European countries increasingly implement actions to control AMR in the community and the hospital setting through rational use of antimicrobials. Information on antimicrobial consumption in Europe is a prerequisite for antibiotic stewardship and can be an important source for healthcare professionals and policy makers in order to monitoring progress towards a prudent use of antibiotics [
7]. However, other factors, such as intra- and inter-hospital spread of resistant microorganisms, community contribution to resistance, and infection control policies and practices, may also play a role in determining the burden of resistance in a hospital and should be considered [
8]. We have previous reported intra- and inter-hospital spread of
K. pneumoniae clones in two Sicilian ICUs [
9] and the emergence of a carbapenem-resistant
K. pneumoniae clone in one of those ICUs [
10]. Furthermore, recently, we have described the dissemination, in the two ICUs of clonally related isolates of carbapenem-resistant
A. baumannii with simultaneous resistance to colistin, hypothesizing that prior carbapenem and colistin consumptions may have acted as triggering factors [
11].
All these findings support the need of strict adherence to control measures to prevent the dissemination of multidrug-resistant microorganism, together with the monitoring of antibiotic consumption at local level to promote the judicious use of antibiotics. The objective of the present study is to report the trends of K. pneumoniae and A. baumannii resistance indicators in an Italian ICU during a six-year period, from 2008 to 2013.
Discussion
Italy is one of the European countries with increasing spread of antimicrobial-resistant microorganisms, often multidrug-resistant [
1] and with high antibiotic consumption in the hospital setting [
1,
7,
16]. Several factors, such as antimicrobial consumption, clonal spread of resistant microorganisms, resistance mechanisms that might differ by species, the human and environmental reservoir, and infection control strategies, including screening policies, may play a role in the prevalence of antimicrobial-resistant pathogens in the hospital setting [
8,
14]. This issue is of interest especially in Italian ICUs where the highest prevalence of patients on antibiotic treatment is observed [
17] and outbreaks due to multidrug-resistant
K. pneumoniae and
A. baumannii are frequently reported [
9‐
11,
18]. Comparison of our data with those of the EARS-Net should be undertaken with caution, since European results are based on invasive isolates from blood or cerebrospinal fluid, thus not representative of isolates from other sites [
1]. Our study confirms that carbapenem resistance is endemic for
A. baumannii isolates, with dramatically higher RRs (about 95 %) than those in the other European countries [
1]. Furthermore, incidence density of this multidrug-resistant pathogen increased significantly from 2008 to 2013 in our ICU, and consumption of carbapenems, after a significant decrease, increased in the last year of the study. Recently, it has been reported that the increasing use of carbapenems was associated with the increasing incidence of healthcare-associated infection due to imipenem-resistant
A. baumannii, suggesting that caution in antibiotic use would play an important role in managing high RRs [
19]. The results of our study, although not confirmatory, may also suggest that increasing carbapenems consumption contributed to increasing rate of drug-resistant organisms. However, in order to further explain high RRs of pathogens associated with infections in the ICUs, it has been suggested [
11,
20,
21] that high RRs are correlated with high rates of antibiotic use, but increased and/or high RRs may also be due to transmission between patients. In fact, in the same ICU, we have recently described the dissemination of clonally related multidrug-resistant
A. baumannii isolates [
11].
In Europe, the percentage of carbapenem-resistant
K. pneumoniae is already high and increasing in some countries. The highest RRs were reported by Greece followed by Italy (28.8 %) with statistically significant increasing trends [
1]. Since carbapenem- resistant
K. pneumoniae isolates are frequently found to be carbapenemase-producing, these results highlight and confirm that surveillance, active screening of patients at high-risk, notification of health authorities, implementation of infection control measures and the prudent use of antimicrobials are key elements to contain the spread of carbapenem-resistant
K. pneumoniae isolates [
1,
22,
23]. In Italy since 2013 the Ministry of Health established a national surveillance of carbapenemase-producing
Enterobacteriaceae (particularly,
K. pneumoniae and
Escherichia coli) [
24]. In our ICU the percentage of carbapenem-resistant
K. pneumoniae was higher than that one documented in Europe with a significant increasing trend, although consumption of carbapenems decreased until 2012 and increased in 2013. As described for
A. baumanni, dissemination of clonally related carbapenem-resistant
K. pneumoniae isolates has been also previously documented in our unit, thus contributing to the increasing trend of carbapenems resistance [
9,
10]. Notably, results on high carbapenem RRs of
A. baumannii and
K. pneumoniae are worrisome because carbapenems are last-line antibiotics. In addition, as previously reported, when carbapenems resistance is due to the presence of carbapenemases, accumulation of other resistance determinants, makes these isolates extensively drug- or pandrug-resistant, leaving few or no effective treatment options [
22,
23,
25,
26].
In European countries, a large proportion of 3GC-resistant
K. pneumoniae isolates, ranging between 85 % and 100 %, was ascertained as ESBL-positive [
1]. In our study, frequencies of ESBL-producing
K. pneumoniae among 3GC-resistant isolates were very high, reaching 100 % in 2008, 2009 and 2010. Furthermore, the incidence density of ESBL producers increased, although not significantly, from 2010 to 2013. The high percentage of ESBL-producing
K. pneumoniae complicates the treatment of serious infections caused by these bacteria leading to a significant increase in carbapenems use, considered the first line class for ESBL producer, with a consequent impact on the emergence of resistance to these antibiotics [
1].
In our ICU the percentage of 3GC-resistant
K. pneumoniae isolates was higher than that one reported for Italy (47.7 %) [
1]. Furthermore, the incidence density of resistant isolates increased significantly from 2010 to 2012 together with the 3GC consumption. In 2013, the 3GC consumption decreased but the incidence density of resistant isolates increased with no evidence of clonal dissemination of 3GC-resistant
K. pneumoniae in the ICU. Interestingly, in the German ICUs the dramatic increase of 3GC-resistant
K. pneumoniae has been associated with an increased trend of antibiotic usage [
14].
Finally, in the present study no correlations between resistance data and the corresponding antibiotic consumption were observed, thus the multifaceted nature of the spread and emergence of resistance can, at least partly, explain this result. In fact, although it is clear that antibiotics may act as promoters of resistance inducing key processes for the emergence and spread of resistance, as mutagenesis, recombination and/or horizontal gene transfer [
8,
14,
19,
27], this occurrence is also influenced by other factors than by antibiotic consumption alone. Antibiotic-resistant bacteria, as well as resistance genes, can spread from person to person to the environment, and then back to humans thus, infection prevention and control activities to limit the spread of resistant bacteria are crucial [
28].
Furthermore, although DDD measurements are useful for comparison and benchmarking, they may not fully correlate with subsequent antibiotic resistance development due to the intrinsic biases [
29].
Our study has the limitation of the retrospective study design as well as the fact that is conducted in a single institution in Italy, thus the results cannot be representative of all ICUs of Southern Italy. Moreover, the ecological study design cannot prove a causative relationship between the two factors. Thus, a prospective multicentre patient-based study design is needed to confirm these findings. Furthermore, in our study design, data on AMR are laboratory-based, thus precluding a patient-based evaluation (i.e. date of admission to the ICU, clinical data confirming infection and other information) to differentiate whether microorganisms were associated with healthcare-associated or community-acquired infection or colonization. Since the emergence of drug-resistant microorganisms is commonly seen in nosocomial settings and invasive isolates could be of much more interest than microorganisms isolated from colonization episodes, microorganisms associated with infection episodes should be analyzed separately. Besides, no data about other factors such as patient’s characteristics and the amount of antibiotics prescribed in the outpatient setting and potential confounders, such as antimicrobial stewardship interventions, were taken into account. Finally, outbreaks not investigated might have influenced resistance data.
Competing interests
The Authors declare that they have no competing interests
Authors’ contributions
AA conceived the study design, coordinated the study, drafted, supervised and reviewed the manuscript; MB participated in study design, performed and interpreted the statistical analysis and drafted the manuscript; AQ performed and interpreted the statistical analysis and drafted the manuscript; AM performed and interpreted the statistical analysis and drafted the manuscript; EA collected data and interpreted the results; AEM: collected data and analyzed results in the laboratory; ARM participated in study design and reviewed the manuscript; AT drafted, supervised and reviewed the manuscript. All Authors read and approved the final manuscript.