Antimicrobial resistance trends in methicillin-resistant and methicillin- susceptible Staphylococcus aureus and Staphylococcus epidermidis isolates obtained from patients admitted to intensive care units. 2010-2015

Antimicrobial resistance trends in methicillin-resistant and methicillinsusceptible Staphylococcus aureus and Staphylococcus epidermidis isolates obtained from patients admitted to intensive care units. 2010-2015 Tendencias de resistencia antimicrobiana en Staphylococcus aureus y Staphylococcus epidermidis resistentes y susceptibles a meticilina aislados, obtenidos en unidades de cuidados intensivos, 2010-2015

Conclusión. Los resultados encontrados sugieren el replanteamiento de las estrategias de control de la resistencia antimicrobiana en el hospital objeto de estudio.

Introducción
The last 20 years have been characterized by the emergence and spread of antibiotic-resistant bacteria, which in turn have become a determining factor for the failure of antibiotic therapies, mainly in patients admitted to intensive care units (ICU). (1,2) This public health problem has serious implications on morbidity and mortality rates, and on public health systems stability, provided that both direct and indirect economic costs increase heavily due long-term hospitalizations and greater drug consumption. (3) To date, different strategies have been developed to minimize the impact this public health problem can have on health systems, including appropriate antibiotics prescription, cleaning and disinfection programs, antimicrobial stewardship programs, and epidemiological surveillance of antimicrobial resistance. (2,4) It has been estimated that in patients admitted to a hospital the risk of contracting a nosocomial infection is ~10%. In addition, said risk almost quadruples when the individual is admitted to the ICU, since antimicrobial therapy is a common practice in ICU patients. (5,6) Moreover, multidrug-resistant bacteria infection susceptibility may be higher in these critically ill patients (7)(8)(9), which limits treatment options and makes mandatory to determine antimicrobial resistance profiles among clinical isolates retrieved from said patients.
Given how antimicrobial resistance works, any modification to be made in antimicrobial management schemes should be based on both antimicrobial resistance surveillance data and bacterial etiology. (10) Changes made to the antimicrobial management schemes are not generally applicable to other hospitals, for they are based on the intrinsic characteristics of each hospital and depend on several aspects such as the type of hospital (facilities), antibiotics availability, and specific local data obtained from monitoring antimicrobial resistance. (11) Staphylococcus spp. are frequently cultured from hospitalized patients and are characterized by being resistant to multiple antimicrobials. Methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-resistant Staphylococcus epidermidis (MRSE) can be used as markers to evaluate antimicrobial resistance time trends in critically ill patients. (12- 19) Espinosa et al. (20), in a systematic review on antimicrobial resistance in gram-positive cocci in Colombian hospitals, reported a higher isolation frequency of methicillin-resistant staphylococci in ICU isolates in comparison to non-ICU isolates. (20) This finding is consistent with those of Villalobos et al. (16) and Villalobos et al. (21), who reported MRSA antibiotic resistance phenotypes and profiles in tertiary hospitals in Colombia. In this context, the importance of an active in vitro antimicrobial resistance surveillance program should be noted.
Effective surveillance of antimicrobial resistant organisms allows generating reports that provide necessary information for analyzing and monitoring resistance patterns; likewise, these surveillance data can be used for making decisions regarding the treatment. (22) The objective of the present study was to evaluate the temporal trends of MRSA and MRSE antibiotic resistance phenotypes from isolates obtained between 2010 and 2015 from ICU patients in a tertiary hospital located in Cartagena, Colombia.

Materials and methods
Study population All S. aureus and S. epidermidis isolates were obtained from patients admitted between January 2010 and December 2015 to three different ICUs (11 adult ICU beds, 11 pediatric ICU beds, and 6 neonatal ICU beds) in a tertiary hospital in Cartagena, Colombia. Samples collection was performed before chemotherapy was carried out.

Study design
A cross-sectional study was carried out to evaluate antimicrobial resistance trends of methicillin-resistant and methicillin-susceptible S. aureus (MRSA and MSSA, respectively) and methicillin-resistant and methicillin-susceptible S. epidermidis (MRSE and MSSE, respectively) by using isolates obtained from ICU patients as resistant markers.

Bacterial identification
A MicroScan® Microbiology System (Siemens Healthcare Diagnostics) was used to perform an automated identification process of the staphylococcal species.
Antimicrobial susceptibility tests A MicroScan® automated system (Dade Behring, Sacramento, USA) was used to determine antibiotic resistance/susceptibility by using the minimal inhibitory concentration (MIC) technique.
MIC interpretive standards for all antimicrobials tested were defined according to the standard breakpoints established by the Clinical and Laboratory Standards Institute (CLSI). (23) Isolates were classified as sensitive (S) or resistant (R); besides, strains with intermediate sensitivity (I) were included in the latter category.
Statistical Analysis IBM SPSS Statistics for Windows, version 23.0, was used for statistical analysis. Annual absolute and relative resistance frequencies were analyzed. For dichotomous outcomes effect size in terms of odds ratio (24) and associated 95% confidence intervals, 95%CI were used. Fisher´s exact test was performed in order to compare antibiotic resistance percentages between S. aureus and S. epidermidis isolates, and Chi-square test for linear trends was used to analyze the 6-years resistance trend.
All tests of significance were calculated with a two-tailed alpha of 5%, and multiple antibiotic resistance (MAR) index was determined using the procedure described by Krumperman (25).

Ethical considerations
The study was approved by the Bioethics Committee of Universidad de San Buenaventura (Affidavit 005/August-19-2014), which classified it as a risk-free research according to the provisions of article 11 of Resolution 8430/1993, issued by the Colombian Ministry of Health (26), provided that the data used here were obtained from bacterial strains collected from patients admitted to ICUs. Likewise, the ethical principles for medical research involving human subjects outlined by the Declaration of Helsinki were followed. (27)
In addition, resistance percentages to ERI, CLI, gentamicin (GEN), CIP, and SXT were far higher in in MRSA than in MSSA isolates. A similar behavior was only observed between MRSE and MSSE isolates (Table 4).  Resistance phenotypes trends in MRSA and MRSE isolates were higher in the last year of the study (January-December 2015), with the exception of resistance to SXT, since SXT-resistant isolates were not identified during this period. On the other hand, for the whole six years of the study, the lowest resistance percentages in MRSA and MRSE isolates were found for SXT (27.9% and 42.5%, respectively) ( Table 5).

Discussion
Staphylococcus species have emerged as significant pathogens in both hospital-acquired and community-onset infections. (28)(29)(30)(31) Hospital isolates of bacterial strains tend to have higher rates of resistance to antimicrobial agents, especially those collected from ICU patients. (20) In this study, a higher frequency of methicillin-resistant staphylococci clinical isolates was reported in comparison with methicillin-susceptible isolates, which is consistent with the findings of national (16,(32)(33)(34)(35)(36)(37)(38) and international studies. (12, [39][40][41][42][43][44] In part, this could be explained by the relationship between antibiotic selective pressure and the emergence of antimicrobial-resistant organisms in hospitals. (45,46) Most methicillin-resistant strains, mainly MRSE, were isolated from patients who were admitted to the NICU. This finding is consistent with several works reporting that S. epidermidis is one of the most common bacterial pathogens in this population. (32,39,41,43,(47)(48)(49)(50)(51) Major risk factors for presenting these staphylococcal infections include prematurity, invasive procedures, prolonged hospitalization, and immune immaturity. (52,53) On the other hand, the percentage values of MRSA and MRSE strains isolated from blood cultures are similar to those reported by Cortes et al. (34) in bacteremic patients admitted to the ICUs of 33 Colombian hospitals. All these similarities suggest three possible scenarios at a national level: (i) failure of empirical antibacterial therapy (34,54), (ii) increased hospital stay lengths for clinically ill patients (55), and increased risk of mortality in ICU patients. (34,56) Also, it was found that MRSE isolates showed a higher resistance frequency to antistaphylococcal antibiotics such as ERI, CLI and SXT that MRSA strains did. Somehow, in the case of ciprofloxacin, this pattern was not observed. Similar results were described by Álvarez et al. (57) In an attempt to explain these findings, the authors of the present study suggest two key factors: selective pressure caused by antibiotic use and cross-transmission of resistant bacteria.
Significant differences between antimicrobial resistance rates of methicillin-resistant and methicillin-susceptible Staphylococcus isolates have also been reported by other authors, although they have done so when comparing coagulase-negative Staphylococcus strains isolated from septicemic patients. (31) About a third of the methicillin-resistant isolates were also resistant to other antibiotics including macrolides, lincosamides, quinolones, aminoglycosides, rifamycins, oxazolidinones, nitrofurans and antifolates. In addition, if a MAR index value >0.2 can be considered a multiple antimicrobial resistance characteristic (58), the presence of these multidrug-resistant strains could support the emergence of the three hypothetical scenarios described above.
Out of 43 MRSA isolates, 2 vancomycin-intermediate and vancomycin-resistant strains were identified in adult bacteremia cases. To date, there are no reports of VISA or VRSA isolates in Colombian hospitals. (21,33,(35)(36)(37)49,57) However, one study conducted in Caldas, Colombia, in 1996 (38), reported the existence of VRSA strains isolated from patients admitted to the ICU, but these isolates were identified using the disc diffusion method. Despite the fact that VRSA isolates presence has not been confirmed by a reference laboratory, this finding could be interpreted as an early warning of the occurrence of clinically relevant pathogens for which only limited therapeutic options are available.
A final aspect that should be noted in this context is that in a recent study Rossi et al. (59) reported vancomycin resistance transmission between MRSA strains isolated from blood cultures.

Conclusion
The most effective antibiotics in MRSA and MRSE isolates were daptomycin, linezolid and vancomycin. The significant and rising trend in the number of MRSA isolates may be an indication of the need to redefine existing strategies to minimize antimicrobial resistance. It is necessary to periodically conduct surveillance studies in all hospitals in the country to achieve effective control strategies of multi-drugresistant infections and to reduce antimicrobial resistance rates.

Conflicts of interest
None stated by the authors.

Funding
None stated by the authors.