Pseudomonas aeruginosa in the ICU: prevalence, resistance profile, and antimicrobial consumption

Abstract INTRODUCTION: Pseudomonas aeruginosa is one of the main pathogens causing infection in intensive care units (ICUs) and usually presents antimicrobial resistance. METHODS: Data were obtained from ICUs between 2010 and 2013. RESULTS: P. aeruginosa had a prevalence of 14.5% of which 48.7% were multidrug resistant. We observed increasing resistance to carbapenems and polymyxin B and growing consumption of aminoglycosides, meropenem, ceftazidime, and polymyxin B. The regression impact between resistance and consumption was significant with respect to amikacin, imipenem, meropenem, and polymyxin B. CONCLUSIONS: Monitoring antimicrobial consumption and resistant microorganisms should be reinforced to combat antimicrobial- and multi-drug resistance.

Antimicrobial resistance is a public health concern 1 . Among resistant pathogens, the bacterium Pseudomonas aeruginosa is of note because it is often associated with high rates of morbidity and mortality in patients in intensive care units (ICUs) 2 . P. aeruginosa is resistant to several antibiotics and has the ability to rapidly develop resistance to new antimicrobials. Moreover, P. aeruginosa was the first bacterium to present multidrug-resistant (MDR) phenotypes 3 . Therefore, it is important to monitor the consumption of antibiotics, especially in a hospital environment, to avoid the development of resistance and improve the therapeutic efficacy of these drugs against P. aeruginosa.
To address the above issues, this study aimed to describe the prevalence of resistant and MDR isolates of P. aeruginosa in nine ICUs of a university hospital, as well as to describe the consumption of antimicrobial agents using a defined daily dose (DDD) and investigate the relationship between the consumption of and resistance to antibiotics in the nine ICUs.
This was a retrospective descriptive study, conducted from January 2010 to December 2013 in ICUs of the Hospital São Paulo -university hospital (HSP-HU/Unifesp) in the city of São Paulo (Brazil).
For resistance and susceptibility analysis, data from the Clinical Laboratory of the hospital were used. Samples were obtained from patients hospitalized in at least one of the ICUs selected during the study period. P. aeruginosa isolates were considered to be MDR if they showed resistance to three or more classes of antimicrobial agents, with resistance to at least one antibiotic in each class 4 .
Data on the consumption of antimicrobials (in grams) were provided by the Supplies System of the hospital, and the hospitalization rate in the study period was provided by the Hospital Statistics Sector. Antimicrobial agents were classified according to the Anatomical Therapeutic Chemical system, and consumption was measured in DDD per 100 bed-days 5 .
Susceptibility tests performed on samples from anal and rectal swabs for Klebsiella pneumoniae carbapenemase surveillance and vancomycin-resistant enterococci surveillance were excluded from the study.
For the organization and analysis of the data, Microsoft® Office Excel for Mac, version 16.27 (Microsoft Corporation) and the Statistical Package for Social Sciences version 20 (IBM Corp, Chicago, IL, USA) were used.
The resistance trend of P. aeruginosa, monthly consumption of antimicrobial agents, and the association between antimicrobial agent consumption and resistance of P. aeruginosa were analyzed using a linear regression model. The β coefficient indicates the association between a dependent variable and independent variables analyzed by the linear regression model. The significance level was set at 0.05.
The remaining analyses were for normality, linearity, and homoscedasticity. For normality analyses, the Kolmogorov-Smirnov (KS) test (significance>0.05) was performed and all the variables obtained were considered normal, except for the data on consumption of ceftazidime and polymyxin resistance. The homoscedasticity and linearity analyses were conducted by graphical analyses. Homoscedasticity was analyzed using the standardized variables against the standardized predicted values.
This study is part of the project "Prevalent microorganisms in ICU: antimicrobial agents consumption and resistance profiles," approved by the Research Ethics Committee (opinion 921.500), in accordance with Resolution No. 466 of December 12, 2012, of the National Health Council.
Of 6,473 microorganisms isolated, 939 were identified as P. aeruginosa (14.5%). P. aeruginosa was the second most prevalent bacterium during the study period, after Acinetobacter baumannii (17.3%). The prevalence of P. aeruginosa isolates resistant to antimicrobial agents by year is presented in Table 1.
In this study, P. aeruginosa was the second most frequently isolated bacterium, representing 14.5% of the total isolates, which demonstrates the clinical importance of this microorganism. The results obtained are consistent with those of other studies, such as the one performed in ICUs of a public hospital in Ceará (Brazil), wherein P. aeruginosa represented 33.8% of the isolates 6 , and that performed in an ICU of Anesthesiology in Turkey, wherein the prevalence of P. aeruginosa was 14.29% 2 , demonstrating that the high prevalence of this microorganism is common in various countries.
The resistance profiles of P. aeruginosa against the tested antibiotics are presented in Table 1. It is noteworthy that Biswall et al. (2014), in India, reported similar resistance prevalence results for aztreonam (41.38%) and piperacillin/tazobactam (34.5%), as well as zero resistance to colistin (or polymyxin E). However, the results were different for amikacin (81.03%), gentamicin (81.03%), ceftazidime (70.68%), imipenem (18.9%), and meropenem (13.79%) 7 . Based on these data, there may be variations in microbial resistance profiles depending on the study site, which shows the importance of determining resistance profiles to be able to select the most appropriate antimicrobial therapy.
It is also worth noting that the prevalence of resistance of P. aeruginosa to imipenem and meropenem showed an increasing trend in the study period. These data concur with the results of a study carried out at a hospital in the state of Santa Catarina (Brazil), which indicated that resistance to carbapenems was increasing. Thus, in 2004, the resistance of P. aeruginosa to imipenem was 6.06%, increasing to 48.09% in 2008; for meropenem, the resistance increased from 6.89% in 2004 to approximately 20% in 2008 8 . Similarly, Xu et al. (2013) in a study performed at a hospital in China found increasing resistance to imipenem (β=4.620; p<0.001) and meropenem (β=4.624; p<0.001) 9 . Although carbapenems are one of the most potent classes of antimicrobials against P. aeruginosa, increasing resistance to these drugs is an endemic problem in several countries; therefore, the World Health Organization (WHO) considers the development of new drugs for this microorganism a priority 10 .
In addition, the present study showed a decreasing prevalence of resistance of P. aeruginosa to amikacin and piperacillin/tazobactam, which differs from the results of Xu et al. (2013), who showed that resistance to amikacin (β=0.488; p=0.332) and piperacillin/tazobactam (β=4.346; p<0.001) was increasing 9 .
Although the resistance of P. aeruginosa to polymyxin B was found to be low (0.9%) in the present study, resistance to

5/6
this antibiotic was observed to increase (β=0.401, p=0.005). This result is worth mentioning as polymyxins are considered the last line of defense against gram-negative bacteria and have been classified by the WHO as critically important for human medicine 11 .
Among the isolates of P. aeruginosa, 48.7% were classified as MDR. This result is similar to the prevalence of 30.5% observed in a multicenter study of patients admitted to hospitals in Europe and the United States 4 . Similarly, a study carried out at a Spanish hospital indicated that 33.3% of P. aeruginosa isolates were classified as MDR 12 . It is also worth noting that in Brazil, In the present study, a downward trend in imipenem consumption was identified, while meropenem consumption was increasing. In the study by Vojtová et al. (2011), increasing consumption of both imipenem (from 0.33 in 1999 to 1.60 in 2008) and meropenem (from 0.88 in 1999 to 3.56 in 2008) was observed 15 .
In addition, a decreasing trend in the consumption of cefepime was observed in the present study, whereas ceftazidime showed an increasing trend. In the study by Vega et al. (2015), the consumption (in DDD/100 beds-day) of cefepime and ceftazidime decreased between 2008 and 2011 from 10.19 to 1.48 and 2.2 to 0.25, respectively 14 .
Furthermore, a growing trend of consumption was found for polymyxin B, perhaps because of the increased prevalence of PaMDR in the study period and because polymyxins are currently the last option for monotherapy against resistant and PaMDR 3 , and are therefore extensively used for this purpose.
By analyzing the correlation between the prevalence of resistance to and the consumption of antimicrobial agents, we observed a negative association between the consumption of amikacin and imipenem and prevalence of P. aeruginosa resistance to these antibiotics. However, even though the consumption of amikacin increased during the study period, we cannot conclude that the decrease in the prevalence of resistance was only due to the possible effectiveness of the antimicrobial since it is known that the decrease in prevalence of bacterial resistance also depends on other factors 1 such as increased surveillance, campaigns to combat the spread of resistance, and changes in microbial resistance profiles.
It is still noteworthy that a positive impact and association between the consumption of meropenem and polymyxin B and the prevalence of resistance of P. aeruginosa to these antibiotics were detected. Xu et al. (2013) obtained similar results for meropenem (β=1.241; p<0.001) but different results for imipenem (β=1.238; p<0.001) 9 . As for polymyxin B, the result obtained is worrisome as polymyxins are the last option for monotherapy against resistant and PaMDR 3 . However, it is important to note that in the present study, P. aeruginosa showed a low resistance profile to polymyxin B.
The present study has the following limitations: (i) the presented consumption data were general data for the nine selected ICUs, not specific data for patients with P. aeruginosa infection; and (ii) other factors that influence resistance were not considered, although antimicrobial consumption is one of the most important factors 1 .
Thus, a prospective study using specific antimicrobial consumption data and data from each patient with P. aeruginosa infection, as well as the analysis of all associated factors and/or resistance influencers, may help better establish the correlations between bacterial resistance and antimicrobial consumption, as well as between bacterial resistance and other associated factors and/or resistance influencers.
In the present study, we determined the prevalence of resistant and MDR isolates of P. aeruginosa, as well as the levels of antibiotic consumption in ICUs of a university hospital in São Paulo (Brazil). The results of this work emphasize the importance of monitoring the consumption of antimicrobial agents and combating the increase of microbial resistance, including multidrug resistance.