Antimicrobial resistance and antibiotic consumption in a third level pediatric hospital in Mexico City

Introduction: The increasing resistance to antibiotics is a public health problem and an imminent therapeutic challenge in hospitals. In this report we aimed to analyze the relationship between antimicrobial resistance and antibiotic consumption in a third-level pediatric hospital. Methodology: A cross-sectional analysis was conducted using the information from the microbiology and pharmacy databases of the Pediatric Hospital “Doctor Silvestre Frenk Freund”, during the period 2015-2018. Prevalence of antimicrobial resistance by microorganisms and dispensed grams of selected antibiotics were calculated annually. Antibiotic resistance trend over the time was evaluated using the Chi-square trends test and to assess the correlation between the dispensed grams of antibiotics with their antimicrobial resistance prevalence, we calculated the Pearson's coefficient (r). Results: A total of 4,327 isolated bacterial samples were analyzed (56.5% Gram-positive and 44.5% Gram-negative). Most frequently isolated microorganisms were coagulase-negative staphylococci (CoNS), E. coli, K. pneumoniae, P. aeruginosa and S. aureus We found a significant increase in resistance to clindamycin and oxacillin for CoNS and significant decrease in nitrofurantoin and amikacin resistance for E. coli and K. pneumoniae. We observed a strong positive and statistically significant correlation between amikacin resistance prevalence and amikacin dispensed grams for P. aeruginosa (r = 0.95, p = 0.05). Conclusions: The antibiotic resistance profile showed by our study highlights the need of an appropriate antibiotic control use in the Hospital setting.


Introduction
The transmission of multidrug-resistant pathogens in hospitals, such as Acinetobacter baumannii carbapenem-resistant, Pseudomonas aeruginosa carbapenem-resistant and Enterobacteriacea carbapenem-resistant, ESBL-producing, is considered as an important public health threat [1]. The knowledge on local microbiology, the resistance patterns and their relationship with usage metrics of antibiotics are the principal measures considered by the "antimicrobial stewardship" strategy which tries to reduce multidrugresistant microorganism transmission among hospitals [2]. In this study we aimed to identify the trends in antibiotic resistance obtained by isolated microorganisms in clinical samples taken during hospitalization and the relationship between the use of antibiotics and their resistance in the Third-level Pediatric Hospital "Doctor Silvestre Frenk Freund", located in Mexico City, which has an average of 6,700 hospitalizations per year.

Methodology
A cross-sectional analysis was conducted using the information from the Pediatric Hospital microbiology database. First isolated bacterial species from blood, urine, normally sterile body fluids and aspiration puncture cultures were included. The samples corresponded to in-hospital patients from the period 2015-2018.
Microorganism identification and antimicrobial susceptibility were carried out using the broth microdilution technique through Vitek 2 automated system (BioMérieux, Lyon, France). Minimum inhibitory concentrations (MIC) were interpreted using the 2018 version CLSI criteria [3].
Annual prevalence of antimicrobial resistance by microorganisms was calculated with their corresponding 95% Confidence Interval (CI) and Chisquare trends were used to assess their behavior over time. Dispensed grams of selected antibiotics were calculated annually, using the information from the pharmacy database and to assess the correlation between those grams and the antimicrobial resistance prevalence; the Pearson's coefficient (r) was calculated and P values < 0.05 were considered statistically significant. WHONET Desktop 2019, IBM SPSS Statistics version 25 (IBM Inc.) and Tableau Desktop ® 2019 (Tableau® Software Inc.) were used to carried out the statistical analysis. The research was approved by the Ethical Committee Number 3603 of the Mexican Institute of Social Security with registration number R-2020-3603-013. All methods were performed in accordance with ethical standards and regulations from the institutional research committees and national laws and with the 1964 Helsinki declaration and its later amendments. Given that this study was based on the use of available databases with no personal identifiers, not formal informed consent was required.

Antimicrobial resistance prevalence
Staphylococci group, including CoNS and S. aureus, showed high resistance prevalence to Oxacilin and Clindamycin, while third generation cephalosporins, ciprofloxaxin and Trimethoprim/Sulfamethoxazole had the highest resistance prevalence for Enterobacteria group and A. baumannii (Table 1).
In the trend analysis of antimicrobial resistance over the time, we found a significant increase in resistance to clindamycin and oxacillin for CoNS and significant decrease in nitrofurantoin and amikacin resistance for E. coli and K. pneumoniae (Figure 1).

Correlation between dispensed grams of selected antibiotics and antimicrobial resistance
We observed a strong positive and statistically significant correlation between amikacin resistance prevalence and amikacin dispensed grams for P. aeruginosa in the Pediatric Hospital (r = 0.95, p = 0.05).

Discussion
As previous studies related to microbiology did, we also found that the most frequently isolated microorganisms were Gram-negative, mainly E. coli, K. pneumoniae and P. aeruginosa [4][5][6][7].
Similar to previous reports in Mexico, our study found an increasing tendency for MRSA in S. aureus [8,9], a high resistance for third generation cephalosporins in E. coli and K. pneumoniae [10,11], a multiple simultaneous antibiotic resistance pattern for A. baumannii [5,6,12] and an increasing resistance for carbapenems in P. aeruginosa [8]. However, carbapenems trends were stable for the main Gramnegative bacilli (E. coli and K. pneumoniae), which could be explained by the strict antibiotic use policy in this Hospital.
Among the limitations of this study, we can mention that all the microorganisms reported in the cultures (including colonization and contamination) were analyzed, so the found patterns could be different for the microorganisms and the main HAIs. Another important limitation was that the dispensed grams of antibiotics provided by the pharmacy may not faithfully reflect the consumption of antibiotics at the pediatric population, where the dose depends mainly on the patient's weight, the site of infection and the antibiotic therapy duration. Also, due to technical problems, no complete information was recorded for several antibiotics prescribed for S. aureus in 2015 and 2016, so the results are incomplete for this microorganism.
Despite the above mentioned limitations our study found an increase in oxacillin resistance in Staphylococci and a decrease in aminoglycosides resistance in Gram-negative microorganisms and also a correlation between amikacin resistance prevalence and amikacin dispensed grams for P. aeruginosa.

Conclusion
Our results highlight the need of an appropriate antibiotic use in the Hospital setting, in order to limit the increase of antimicrobial resistance.