UTIs are among the most common infections worldwide and caused predominantly by uropathogenic Escherichia coli (UPEC), which could lead to recurrence, renal damage, sepsis, or even death [6]. Thus early and appropriate empirical antibiotic therapy is of great importance. UTIs are a significant cause of morbidity in infant boys, older men, and females of all ages [1]. David et al. characterized the etiological profile of UTI in children < 2 years of age, showing that Escherichia coli proportion increased with age while Klebsiella spp. and Enterococcus spp. proportions decreased with age, with the average percentages of 56.9%, 14.1%, and 11%, respectively [14]. While the data of adults were different from children to some extent [6, 8, 9]. Thus the etiological profile and antibiotic susceptibility may change with age. However, few studies systematically compared the data of different ages, especially when subdivided by different ward types (outpatient, inpatient, ICU, or emergency) or gender (male or female). We presumed the underlying factors in these subgroups may be different, and age could be an important factor when considering empirical antibacterial therapy.
Although Escherichia coli in this study was still the dominant urinary pathogen in pediatric (34%, n = 362), adult (43%, n = 3416), and geriatric (44%, n = 1617), the percentage was lower than the report from China mainland (49.7%) [15], Taiwan (60.5%) [16] and Ethiopia (52.7%) [8]. One possible reason could be the proportion of complicated UTIs were higher than uncomplicated UTIs in our study, as both urology and nephrology departments in our hospital are the top 1 key discipline in China, with more intractable UTI patients visiting. The etiological profiles were also different in specific age populations, such as pediatrics [15], veterans [16], and pregnant women [17]. We also found the etiological profile was more diverse in adult than pediatric and geriatric patients, while the diversity was low in newborn. This is probably because the proportion of adult was the largest among all age categories, and the newborn were mostly from inpatients due to nosocomial or complicated UTIs. Furthermore, we analyzed the adult and geriatric patients subdivided by wards types. The proportion of Escherichia coli decreased, while some nosocomial infection associated pathogens (Klebsiella peumoniae, Pseudomonas aeruginosa, and Acinetobacter baumannii) increased in ICU and emergency department. More broad-spectrum antibiotics were used in ICU and emergency department compared with outpatient and non-ICU inpatient, which showed different etiological profile in these age categories in different ward types.
The dominant uropathogen in newborn was Eenterococcus faecium (45%) instead of Escherichia coli (20%). The proportion was lower than previous studies from West Asia and Europe (50%-57%) [18, 19], thus high proportion of non-E. coli uropathogens in newborns should be noticed when considering appropriate treatment. The percentage of Escherichia coli increased with age while the percentage of Enterococcus faecium decreased with age, which was in accordance with previous study [14]. The resistance rate of Enterococcus faecium was high to multiple available antibiotics in both newborn and pediatric, except vancomycin, linezolid, and teicoplanin. Markus et al. found that UTI caused by Enterococcus spp. was associated with male predominance, underlying urinary abnormality, and inappropriate empirical antibiotic therapy [20]. One possible explanation for the high percentage of Enterococcus faecium in newborn and pediatric of this study was that, a large proportion of them was from inpatient, and might be nosocomial infections. Thus more attention should be paid to urinary Enterococcus faecium in newborn and pediatric patients.
Antibiotic resistance is becoming a serious global health problem [21], and updated surveillance of antimicrobial susceptibility of specific type of infection is of great importance for initial empirical therapy [9]. The resistance rates of ampicillin, sulfamethoxazole-trimethoprim, and cephalosporins for Escherichia coli were higher in newborn and pediatric than in adult and geriatric overall, probably due to the inappropriate use of the these drugs, especially in the community setting. The resistance rates of carbapenems, pipercillin-tazobactam, and nitrofurantoin were < 5% in Escherichia coli overall, thus could be used for empirical treatment of UTI (Fig. 2). The resistance rate of Klebsiella pneumoniae was more serious compared with Escherichia coli (Fig. 3). The resistance rates of cephalosporins, fluoquinolones and carbapenems were higher in adult than in geriatric (P < 0.05). The overall resistance rates of Enterococcus faecium were higher than Enterococcus faecalis (Figs. 4 and 5, P < 0.05). The resistance was more serious in adult and geriatric, with the trend of increasing with age.
This study has certain limitations: Firstly, due to its single-center and retrospective nature, the patients’ selection bias existed and detailed clinical characteristics were not available. Secondly, although we enrolled a large sample size of 13308 non-repetitive patients from a tertiary general hospital, which has the top 1 discipline of nephrology and urology in China during a 12-year period, our data may not fully stand for the community or other specific population (e.g. mild or asymptomatic UTI patients), thus the conclusion should be extended with caution. Nevertheless, our results emphasized that the etiological profile and antibiotic susceptibility patterns of UTI patients were different among different age categories from large sample size, which are fundamental for designing proper antimicrobial therapy and provided the basis for high-quality research in the future.