Carbapenems are the most effective drugs for the treatment of severe infections with gram-negative bacteria due to their broad antimicrobial spectrum and high stability for hydrolysis by most β-lactamases, including extended-spectrum β-lactamases (ESBLs) and AmpC cephalosporinases [16]. But unreasonable use of antibiotics leads to the emergence of CRE, which is becoming more and more serious, and the treatment option is limited [7, 17, 18]. Therefore, understanding the hazard factors for antibiotic-induced CRE is very important in the early selection of empirical antibiotic program. In this study, we explored the relationship between the use of antibiotics and CRKP and CREC respectively. Compared with CSE, the risk factors of CRE in clinical samples were gender, principal diagnosis, urinary cannula, vascular cannula, mechanical ventilation, blood transfusion or use of blood products, respiratory failure, and ICU admission.
Previous studies comfirmed that antibiotic exposure was a risk factor for CRE infection [19, 20]. In our study, the first significant finding was that the difference in antibacterial therapy was significant between CSKP group and CRKP group, and the proportion of combined antibacterial therapy in CRKP group was higher. Increased exposure to one antibiotic increased the effect of exposure to other antibiotics on the risk of CRKP infection [21]. Therefore, combined use of antibiotics would increase the selection pressure of antibiotics, allowing carbapenem-resistant bacteria to produce various carbapenem resistance mechanisms [22]. This result was consistent with that in CSEC group and CREC group. During the study, βlactamase inhibitors were the most commonly used antibiotics.
The second significant finding was that carbapenem, lactamase inhibitors and quinolones were associated with higher 30-day CRKP hazards than other antibiotics after survival analysis. Among the antibiotics, carbapenems the most principal hazard factor, which was consistent with previous studies [2, 23–29]. The use of carbapenem may promote the production of carbapenemase, such as K pneumoniae carbapenemase and metallo-β-lactamases, which could increase the production of CRE [30]. Other mechanisms of carbapenem resistance include outer membrane porin expression loss combined with extended-spectrum β-lactamase (ESBL) and AmpC enzyme, change of antimicrobial target and high expression of efflux pump [31–33]. The result suggested that restriction of carbapenems was associated with a significant reduction in the incidence of CRKP. However, for CREC, the 30-day hazards were significantly higher in patients with β-lactamase inhibitor and 3rd-cephalosporins than other antibiotics.
Last but not least, we found the hazard of CRKP induction decreased with the increase dose of lactamase inhibitors, but there was no significant change in the hazard ratio of CRKP induction with the increase dose of quinolones or carbapenems. The result indicated that the risk of CRKP was higher in low-dose lactamase inhibitor exposure. Therefore, we suggested that adequate dose of lactamase inhibitors should be used in the treatment of these patients. However, there was no relevant research, and further studies were necessary to explore the mechanism. We also found that increasing the use of carbapenem or quinolones did not lead to a significant increase in the resistance of carbapenem among Enterobacteriaceae which was similar to a few previous studies [34–36]. But some studies showed that carbapenem or fluoroquinolone use had a positive relationship with the incidence of CRE [35, 37]. The result was still controversial, which might be explained by the influence of multiple interactive factors that had an impact on induction of CRE. As a preceding result, hazard of CRKP induction decreased might be responsible for the increasing dose of β-lactamase inhibitors. Moreover, there was an obvious characteristic of "parabolic curve" for the hazard of CREC induction due to lactamase inhibitors, and the hazard value gradually increased with the dose, reached the maximum at 24g, and then gradually decreased from 26g. In other words, with the increasing use of lactamase inhibitors, the risk of CREC increased and reached the maximum when the dose of lactamase inhibitors reached 24g, but gradually decreased when the dose exceeded 26g. But no existing study reported this correlation. This result may explain why the incidence of CREC was low in CRE. Although antibiotic exposure may lead to all clinically significant antibiotic resistance, the effects are not consistent and vary with the organism and antibiotic resistance mechanism. Large sample studies were needed to further clarify the mechanism of the effect of β-lactamase inhibitor doses on CRE. In this study, lactamase inhibitors had different drug resistance inducing effects on CRKP and CREC, which suggested that the maximum dose should not be blindly pursued in clinical medication. We should choose the appropriate treatment plan according to the specific characteristics of microbial distribution, or adjust the antimicrobial treatment strategy in real time according to the characteristics of microbial distribution, which raised higher requirements for the rational use of antibiotics.
There were several potential limitations in our study. Firstly, this study was a retrospective design conducted in a large tertiary A-level hospital, not a multicenter research. Secondly, Due to the low prevalence of CRE and the two types of CRE were analyzed according to the use of antibiotics, the sample size was not large enough. Thirdly, in our study, patients with CRKP were less likely to use carbapenems before detection of CRKP. So we found that there was no significant change in the hazard ratio of CRKP induction with the increase dose of carbapenems. But the effect of carbapenems on CRE was beyond all doubt. Fourthly, genotypic detection of drug resistance genes was not performed in the study. Therefore, multi-center studies with large sample size were necessary in the future to address these limitations to further confirm the relationship between CRE and antibiotic usage.