Epidemiology, risk factors and outcomes of multidrug-resistant bacteria colonization in a Moroccan Medical Intensive Care Unit

¹Medical Intensive Care Unit, Mohammed VI University Hospital, Cadi Ayyad University, 40080 Marrakech, Morocco, ²Microbiology Department, Mohammed VI University Hospital, Cadi Ayyad University, 40080 Marrakech, Morocco

^&Corresponding author
Abdelhamid Hachimi, Medical Intensive Care Unit, Mohammed VI University Hospital, 40080 Marrakech, Morocco

Introduction    

Nosocomial infections with multidrug-resistant bacteria (MDRB) is a major global health problem, particularly in low- and middle-income countries. According to the Extended Prevalence of Infection in Intensive Care (EPIC II) study which was a one-day point prevalence, led on May 8, 2007, in 1265 ICUs from 75 countries, about 35% of the bacteria isolated were MDRB [1]. They are responsible for higher morbidity and mortality, longer duration of mechanical ventilation, and intensive care unit (ICU) stay [2-4]. The risk of acquiring a nosocomial infection by MDRB has increased with the evolution of care practices and the recruitment of patients [3]. Colonization is the first step of infection. Organisms colonize hosts from the external environment, develop adherence mechanisms, and overcome the host defenses [5]. At ICU admission, the colonization prevalence extends to 10-15% in Europe and up to 40% in some Asian countries [6]. Indeed, cefotaximase (CTX-M)-producing Escherichia coli has spread globally even in the community, 5-47% in Africa, 7-44% in Southeast Asia, 2-12% in Europe, and 29-63% in the West Pacific area [7]. One of the main pillars against the spread of MDRB is the fight against cross-transmission by the hygiene of hospital environment and nursing staff awareness as well as the screening of carriers that constitute a reservoir from which these bacteria can spread [2-4]. This strategy is recommended by international guidelines for ICUs to face MDRB endemicity and outbreaks [8,9]. Screening can identify these patients and take precautions to isolate them to prevent cross-transmission. The purpose of this work was to specify the frequency of MDRB carriage, to identify risk factors for acquiring MDRB and outcomes in a medical ICU.

Methods     

A prospective cohort study was conducted in a 10-bed medical ICU in Mohammed VI University Hospital of Marrakech (Morocco). We included all patients (≥ 18 years old) admitted between January 1^st and December 31^st, 2015, except those hospitalized less than 24 hours. All patients had nasal and rectal swabs at admission and repeated weekly until discharge or death. Bacterial identification was performed using the PHOENIX 100 (Becton Dickinson) according to the manufacturer´s instructions. The interpretation of antimicrobial susceptibility was made according to the recommendations of EUCAST CASFM. The identification focused on extended-spectrum beta-lactamase Enterobacteriaceae (ESBLE), multidrug-resistant Acinetobacter baumannii (MRAB), multidrug-resistant Pseudomonas aeruginosa (MRPA), methicillin-resistant Staphylococcus aureus (MRSA) and carbapenemase-producing Enterobacteriaceae. (CPE). We defined the colonization as a positive swab at ICU-admission or during hospitalization without signs of infection.

The collected data included the following: age, gender, medical history, cause of admission, Acute Physiology and Chronic Health Disease Classification System II (APACHE II), invasive procedures, results of culture swabs, colonization pressure, treatments, and outcomes. Defined as the ratio of MDRB-colonized patients relative to all patients, the colonization pressure is a tool for assessing the importance of colonized patients as an organism vector in an environment and for a period; calculated according to the formula: (Positive samples x patients - days x 100/total patient-days) [10-13]. Categorical variables were expressed as a percentage, while numerical variables were expressed as average ± standard deviation or median (quartiles 25%, 75%). The analysis was done in univariate using the Mann-Whitney U-test or the student's t-test, as appropriate, for comparison of numerical variables and the Chi-square test or Fisher test for categorical variables. Multivariate logistic regression analysis was considered to estimate the strength between exposure and outcome variables (risk factors and prognostic factors). A difference is considered significant when p <0.05. Statistical analysis was performed using IBM SPSS Statistics version 10 for Windows.

Ethical aspects: informed consent was waived, and researchers analyzed only anonymized data. All research was conducted following the national guidelines and regulations. Which note that only interventional studies require permission.

Results     

Baseline characteristics: among 334 patients admitted, only 300 were included in the study; the remaining 34 patients died on admission or hospitalized for a few hours before transferring to another department. The mean age was 44.4±17 years with a male predominance in 57% of cases. Diabetes dominated medical history (16%). The main reason for ICU-admission was neuromeningeal diseases in 24% of cases. More than one-third of patients were hospitalized in the surgical/medical ward before ICU-admission and had three invasives procedures in the ICU. About half of the patients received antibiotherapy before ICU-admission. The mean APACHE II was 19 ± 7.4. More than two-third required vasopressor agents. The mean length of ICU-stay was 10.6 ± 6.8 (Table 1).

Characteristics of colonization: at the admission, of these 300, 76 cases were colonized. Thus, the MDRB prevalence was 25.3% (76/300). The following species were identified: ESBLE in 45%, followed by MRAB in 28%, CPE in 12%, MRSA in 8%, and MRPA in 7% of cases. For hospitalized patients more than 48 hours, 65 new cases were positive among the 144 non-carriers. The incidence (ICU-acquired colonization) was 45.14% (65/144). The mean time to colonization was 9.63±3.6 days. There were 55% of ESBLE, 21% of MRAB, 11% of CPE, 9% of MRSA, and 4% of MRPA.

Risk factors of MDRB colonization: according to the multivariable model using logistic regression model, the most significant independent risk factors for MDRB colonization were transfer from another ICU (OR 9.8, 95% CI 1.8-54.5, p=0.009) or a private structure (OR 36, 95% CI 5-262, p=0.001), prior antibiotics before ICU admission (OR 5.2, 95% CI 1.92-13.96, p=0.001), central venous catheter use (OR 4.6, 95% CI 1.06-20, p=0.04), and colonization pressure (OR 1.3, 95% CI 1.16-1.45, p=0.001) (Table 2).

Outcomes: the overall mortality rate was 40.3%. The mortality, mechanical ventilation and nosocomial infection rates were higher in colonized than non-colonized patients (52% vs 30%; p=0.001), (62% vs 42%; p=0.001) and (56% vs 18%; p=0.001), respectively. In the multivariable model by logistic regression model, the most significant independent prognostic factors were history of chronic renal failure (OR 10.3, 95% CI 1.4-74.1, p=0.02) or diabetes (OR 7.3, 95% CI 1.2-45, p=0.03), infection at admission (OR 11.5, 95% CI 2.3-58.2, p=0.003), need of mechanical ventilation (OR 22.8, 95% CI 2.4-214, p=0.006) and its duration (OR 1.8, 95% CI 1.31-2.56, p<0.001), and MDRB colonization (OR 22.4, 95% CI 3.2-153, p=0.002) (Table 3).

Discussion     

Our results suggested that the prevalence of MDRB colonization was 25.3% and the incidence was 45.14%. The independent risk factors of MDRB colonization were the transfer from another ICU or a private structure, prior antibiotics before ICU admission, central venous catheter use, colonization pressure, length of stay before ICU admission and ICU stay. Moreover, MDRB acquisition was associated with an increased need for mechanical ventilation, hospital and ICU length of stay, and mortality rate. The independent predictors for ICU mortality were age, history of chronic renal failure or diabetes, infection at admission, need for mechanical ventilation and its duration, MDRB colonization, and ICU length of stay. In our setting, one Moroccan study in a surgical ICU of a university hospital showed 29.4% as an overall incidence of digestive ESBLE colonization, with one independent risk factor: current antibiotic therapy [14]. As reported in other regions, the frequency varied and depended on region and population. In the Americas (the United States of America and Brazil), the colonization at admission varied between 2.2 and 6.3% and during ICU stay between 2.5 and 6% [15-17]. In France, at admission, it was between 4 and 15.4%, and during ICU stay between 2 and 13.2% [18-20]. Our findings are similar to south-east Asia and Tunisia. In south-east Asia, at admission, the carriage was between 28 and 33%, and during ICU hospitalization between 15 and 42% [21-23]. In Tunisia, 21% at admission and 43% during ICU stay [24].

Several studies, in different regions, showed various predicting factors of the acquisition of MDRB in the ICUs. In a French study, the risk factors for MDRB acquisition were prior use of piperacillin/tazobactam, surgery and hospitalization in a room after a patient colonized by MDRB, mechanical ventilation, tracheostomy, and sedation [19]. In Singapore, MDRB risk factors were antibiotic therapy, renal failure, hepatopathy, central venous catheter, and recent surgery [25]. In China, the independent risk factors of MDRB colonization at ICU admission were prior antibiotic therapy more than two types, prior use of broad-spectrum antibiotics within 3 months, duration of prior antibiotic administration, and hospitalization days before ICU admission > 9 days [26]. In Korea, men gender, history of admission within one year, co-colonization with multidrug-resistant Acinetobacter baumannii and extended-spectrum β-lactamases-producing bacteria, and exposure to glycopeptide antibiotics were independent predictors of carbapenem-resistant Enterobacteriaceae carriage [27]. In another recent French work, colonization pressure, mechanical ventilation and the presence of an arterial catheter were independent predictors for ICU-acquired MDRB [13] as well as high SAPS III, severe chronic obstructive pulmonary disease, the need of mechanical ventilation, central venous catheter or hemodialysis catheter in Brazil [28]. The risk factors in our study are similar to these findings in some points.

Several recommendations are proposed to deal with MDRB carriage. Derde et al. in a multicentric European study, highlighted enhancing hand hygiene and chlorhexidine body-washing to reduce the acquisition [29]. Besides, fecal microbiota transplantation has emerged as a promising therapeutic option in this field [30,31], so many clinical trial are currently underway [32]. Our work has some limitations. It was monocentric in a university hospital. Consequently, the conclusions cannot be generalized. Also, we did not evaluate the standard precautions of hygiene and workload. Besides, this study was centered on all MDRB with different frequency, risk factors, and prognosis. Finally, there is any information about antibiotic use in the community. Previous studies showed an association between MDRB carriage and mortality rate [13,27,28,33]. Furthermore, in a systematic review, 16.5% of cases developed an infection during stay among carbapenem-resistant Enterobacteriaceae carriers [34].

Conclusion     

Admission prevalence and incidence of MDRB carriage were high compared to developed countries. Therefore, the hospital needs to reinforce precautions to minimize cross-transmission and infections in critically ill patients, such as compliance with hand hygiene and control cleaning, as well as implementing antibiotic stewardship programs.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

IO, FM, AOT, and AT: contribution to conception and design and acquisition of data; drafting the article; final approval of the version to be published. NS, ME and AH: contribution to conception and design, analysis and interpretation of data; drafting the article and revising it critically for important intellectual content; and final approval of the version to be published. All the authors have read and agreed to the final manuscript.

Tables     

Table 1: baseline characteristics of the patients

Table 2: risk factors associated with MDRB colonization among participating patients in univariable analysis and multivariable analysis by a logistic regression model

Table 3: predictors for ICU mortality by multivariable analysis by a logistic regression model

References