Carbapenem-Resistant Enterobacteriaceae among In-Patients of Tertiary Hospitals in Southwest, Nigeria

The present study assessed the incidence and risk factor associated with the occurrence of carbapenem-resistant Enterobacteriaceae (CRE) among hospitalized patients at three tertiary hospitals in Southwest, Nigeria. A cross-sectional study was performed over a six-month surveillance period in the locations with a total of 300 blood and urine samples. A structured close-ended questionnaire was also administered to all subjects for review of demographics and potential risk


Introduction
The increase in the rate of antimicrobial resistance exhibited by bacteria, especially the Enterobacteriaceae family, is a threat to public health (Albiger et al., 2015). This threat decreases the ability to successfully treat numerous infectious diseases, while simultaneously increasing health risks for vulnerable patients. Carbapenemresistant Enterobacteriaceae (CRE) are able to inactivate carbapenems which are revered to possess higher antibacterial spectrum against Gram negative bacteria than penicillin, cephalosporins and other β-lactams (ECDC, 2013;Jeong et al., 2015).
CRE produce seriously difficult to treat infections (urinary tract infections, septicemia, pneumonia, or intraabdominal infections) in debilitated and immunecompromised patients, in association with prolonged hospitalization and increased fatality, with mortality rates up to 50% (Guh et al., 2015). Few studies have been carried out on CPE strains in Nigeria, with reports of incidence ranging between 10-33.5% (Yusuf et al., 2012a(Yusuf et al., , 2012bEjikeugwu et al., 2013;Motayo et al., 2013;Mohammed et al., 2015;Oduyebo et al., 2015). Although carbapenems are unusually prescribed in Nigerian hospitals, there still remains a necessity for a focused approach on the dissemination of CRE. Also, CRE has the disposition to spread quite promptly, especially in an hospital environment where the risk of transmission is much higher; therefore a low incidence is not to be ignored, but rather a charge to curb further spread.
The objective of the present study was to determine the occurrence, frequency of distribution and risk factors associated with the colonization of CRE in three tertiary hospitals situated in Southwest Nigeria.

Description of study area
The study was conducted in three states in the Southwestern part of Nigeria. The region's population is predominantly heterogenous with people of varied religious and ethnic beliefs. The study was carried out in three tertiary hospitals from Ekiti, Osun and Oyo states (Fig. 1).

Study population and description
The study was carried out among all consenting age groups (children, adults and aged) who were hospitalized and also met the inclusion criteria in the selected hospitals that were surveyed. The survey/ sample collection was carried out in all three locations from October 2016 to March 2017.

Identification of isolates
The isolates were identified by using the Hi25 TM Enterobacteriaceae identification kit (HiMedia Laboratories Ltd., India) and PCR.

Determination of susceptibility to antibiotics
Solidified Mueller Hinton agar (Rapid Lab, UK) plates were seeded with 100 µl of the standardized organisms and were spread evenly over the total surface area of the agar using a glass spreader. Multiple antibiotics discs containing eight of the aforementioned antibiotics were carefully and firmly placed on the surface of the agar using sterile forceps. For imipenem and meropenem, single discs of each of the antibiotics were placed. Afterwards, the plates were incubated at 37 °C for 18-20 hours. After incubation, the zones of inhibition generated by the antibiotics were measured. The values for each organism against the antibiotics were interpreted using the breakpoints interpretative criteria of Clinical and Laboratory Standard Institutes (CLSI, 2016).

Statistical analysis
The descriptive statistics of the data obtained in the study was analyzed and/ or plotted using SPSS 22.0, SigmaPlot. The odd ratio was calculated using MEDCALC and Vassarstats online statistical software to measure the association of variables as risk factor in CRE colonization. The p-value for significance was at p ≤ 0.05.

Results
Demographic characteristics and risk factor assessment All age groups specified for the study participated in all three hospitals except at Oyo state where none of the subjects was older than 60 years. Equal ratio of the genders was used in all the three locations (Table 1). Over 80% of the subjects have been on hospitalization within 3 weeks in the three hospitals, while 14-18% were hospitalized for over 3 weeks. Also, more of the participants in the three hospitals (66%, 88% and 89% respectively) had no recent surgical procedures. On the evaluation of antibiotics use, it was gathered that most of the participants in the three hospitals were on antibiotics within 1-3 weeks (64%, 78% and 75%). A similar trend of 21% or 22% was recorded in all locations for use of antibiotics from 3 weeks to 3 months. Exposure to catheter or other invasive devices during hospitalization in the three hospitals ranged between 17-42%. Subjects with previous diagnosis of UTI ranged from 28-34% in the hospitals, while records of previous hospitalization were < 40% (Table 2).

Inclusion criteria for CRE screening
• Patients who had been hospitalized within 3 weeks and above • Patients who undergone surgery • Patients with urinary tract infections • Patients with invasive devices such as catheter • All age groups who met the aforementioned criteria

Ethical consideration
Ethical approval for the research was obtained foremost from the Ethics Review Board of the Faculty of Life Sciences, University of Ilorin. Then, approval was obtained from the Ethics Review Committee of the selected hospitals after which informed consent was obtained from the subjects and/or their caretakers.

Data collection
Questionnaires were administered to all subjects who participated in the study. The response to the questionnaires was used to collate information about the socio-demographic characteristics such as age, gender and socioeconomic status (level of education and occupation) of the participants. The questionnaire also provided information on the patients' medical history, use of antibiotics and patients' exposures to the risk factors associated with the colonization of CRE.

Sample collection
Blood or urine samples were obtained from subjects who met the inclusion criteria, placed in an ice chest and transported to the laboratory for analysis.

Procedure for isolation of Enterobacteriaceae
Blood sample: Following the collection of the blood samples directly into brain heart infusion broth (Rapid Labs, UK), the samples were incubated for 5 days at 37 °C. After 5 days, the blood culture was subcultured by streaking a loopful of the culture suspension on solidified MacConkey and Blood agars (Oxoid, UK) and incubated at 37 °C for 24-48 hours.
Urine sample: the samples obtained were analyzed within 6 hours of collection. A loopful of urine was aseptically picked from the urine sample and streaked on a solidified CLED (Lab M, UK) agar; then incubated at 37 °C for 24 hours.

Antibiotic susceptibility test
Isolates that showed multiple resistances to one to five antibiotics were categorized as multi-drug resistant (MDR); resistance to six to nine antibiotics was categorized as extensive drug resistance (XDR), while complete resistance to all antibiotics was categorized as pan-drug resistance (PDR). In TE, out of the 21 isolates tested, 28.6% exhibited MDR, while 71.4% were XDR. All the isolates showed complete resistance to augmentin, ampicillin, ceftazidime and cefuroxime. Resistance to other antibiotics was as follow: ciprofloxacin (61.9%), gentamicin (57.1%), ofloxacin (57.1%), nitrofurantoin (19%), imipenem (19%) and meropenem (19%) (Fig. 2a). Notably, there was a fairly low level of resistance against the carbapenems (imipenem and meropenem), with four of the twenty-one isolates being resistant (Fig. 3).

Incidence proportion and risk factor association
The incidence proportion of CRE in TE was 6%, while the incidence rate was 3 cases per 100 people; in TOY, 7% with an incident rate of 3.5 cases per 100 people; in TOS, 10%, while incident rate was 5 cases per 100 people. Age (p = 0.01) and exposure to invasive devices (p = 0.05) were significantly associated with CRE colonization. The duration of hospitalization (p = 0.06), prolonged use of antibiotics (p = 0.08) and previous hospitalization (p = 0.08) were also associated with CRE colonization.
The odds of CRE colonization was 68% higher in patients of ages 51 years and above. Patients who had been on hospitalization for over 3 weeks had 59% higher odds of CRE colonization compared to patients who were on hospitalization within 3 weeks (Table 4). Gender was not a confounding factor in CRE colonization (p = 0.28) although odds of CRE colonization was 16.1% lower in males. Presence of underlying medical condition (p = 36) and recent surgical procedures (p = 45) were also not associated with CRE colonization (Table 4).

Discussion
Demographic characteristics of the subjects Over 50% of the participants in each location belonged to the age group 21-50 years old; less than 10% were aged 6-10 years old and ≤ 11% were 70 years and above. For the present study, a higher number of the participants were adults (21-50 years) and thus may be considered to be a mixed population of individuals at the peak of their immunity, as well as those with a probable decline in immune system efficacy. The gender of the respondents was evenly distributed in all the locations. This was done in order to critically assess the possible influence of gender on colonization or infection by CRE.

Isolation of Enterobacteriaceae
Enterobacteriaceae are known to be the most representative bacteria family in the cause of nosocomial infections and outbreaks (Cai et al., 2017). Nosocomial agents have the propensity to spread at an alarming rate among patients with incidence rate at about 4.5 cases per 100 people (Cunha et al., 2016;Demiraslan et al., 2017). Notably in the hereby study, Escherichia coli and Enterobacter aerogenes were the most dominant. The high frequency of E. coli has also been reported by Yusuf et al. (2012), Motayo et al. (2013) and Mohammed et al. (2015) in hospitals at Kano and Abeokuta, Nigeria. It can therefore be said that in this study, bacteria belonging to the Enterobacteriaceae family that are principal etiology of nosocomial infections have been isolated, which may be an indicator of an existing or impending nosocomial infection. Overall, 57.6% of the isolates were recovered from urines samples, while 42.4% were from blood samples. The higher frequency in urine samples observed was similar to the reports of 57.1% by Mohammed et al. (2015) and 52% by Lixandru et al. (2015).

Antibiotic resistance of isolates
The antibiotics used in the study belonged to different classes of antibiotics including aminoglycosides, fluoroquinolones, penicillins, nitrofurans, beta-lactam, cephalosporins and carbapenems. All the 59 Enterobacteriaceae isolates recovered from blood and urine samples in the three hospitals were multi-drug resistant. The isolates considered as most susceptible were resistant to at least 4 out of 10 antibiotics. Complete resistance to second and third-generation cephalosporins (cefuroxime and ceftazidime) was also observed in all the isolates. This observation is higher than the report of about 52.1% resistance to cephalosporins by Mohammed et al. (2015) in a hospital in Kano state, Nigeria. Resistance to cephalosporins has been reported to be an indicator of beta lactamase production and also, carbapenemase production in Enterobacteriaceae (Huang et al., 2014).  Fig. 3. Susceptibility of the isolates to meropenem and imipenem; A represents meropenem; B represents imipenem According to the CLSI (2012) guidelines on the phenotypic determination of ESBL production in Enterobacteriaceae, non-susceptibility of isolates to cephalosporins (ceftazidime, ceftriaxone, cefotaxime or cefuroxime) or beta-lactam (aztreonam or amoxicillinclavulanate) could be taken as potential ESBL producers. The results of the present study are therefore suggestive that these isolates carry any of the aforementioned resistance genes.

Incidence of CRE and risk factor association
The overall incidence of CRE in EKSUTH, LAUTECH Ogbomoso and LAUTECH Osogbo at the time of study from October 2016 to March 2017 was 7.7% (12.8 cases per 1,000 people). Age and exposure to invasive devices were significantly associated with CRE colonization. More so, CRE distribution was higher in patients aged 51 years and above. According to the concept of immunosenescence which explains the change in immune responses as a result of aging, individuals belonging to older age groups are more susceptible to infectious diseases, cancer development and auto immune (Castelo-Branco and Soveral, 2013;Simon et al., 2015). Montencino-Rodriguez et al. (2013) also reported that individuals that are 70 years and above have an increased vulnerability to infectious diseases. Duration of hospitalization and prolonged use of antibiotics were also factors implicated in CRE colonization. Although presence of a medical condition and surgical procedures are existing risk factors of CRE acquisition, they were not confounding factors in the hereby study. Gender of the patients was also not associated with CRE colonization.

Conclusions
The study revealed the incidence of carbapenemresistant Enterobacteriaceae (CRE) among in-patients receiving healthcare in three selected tertiary hospitals in Southwest, Nigeria. Also, the strong association of age with CRE colonization may pose high complications in older patients. Succinctly, the low incidence of CRE isolates in the study locations at the time of the survey does not nullify possible incidence in subsequent surveys. It is therefore crucial to establish a surveillance program that will ensure routine check to allow early detection of CPE in order to prevent outbreaks.