Prevalence and Incidence of Carriage of Carbapenem Resistant Enterobacteriaceae among Patients Admitted to Mulago Hospital General Intensive Care Unit

Abstract

Background: Carbapenem Resistant Enterobacteriaceae CRE) are among the major causes of nosocomial infections contributing to increase morbidity and mortality in the ICU. Carbapenem being the last choice of beta lactams for the management of Gram-negative bacteria associated infection; clinicians have limited options for the management of such infections. However, hardly any published data establishes the incidence and factors associated with carriage of CRE among clients admitted to the ICU in Uganda. Methods: This was a prospective cohort study of 75 patients recruited in the Mulago hospital general Intensive Care Unit between May 2019 and August 2019. Samples collected at 0 hours, 72 Hours, 144 hours, 216 hours 288 hours and at discharge. Kirby Bauer disk diffusion test was performed to determine antimicrobial sensitivity patterns and the presence of CRE was determined using the carbapenem inactivation test, EDTA test, and Boronic acid test. Prevalence of CRE carriage at baseline was determined and factors associated with CRE carriage were determined by logistic regression. Incidence of carbapenem resistant Enterobacteriaceae on ICU ward was also determined and the predictors determined using the Cox-proportional hazard model. Results: Patients had a median age in years (IQR) of 37.0 (27.0-52.0) and 41/75 (54%) were female. All patients had visited a health facility prior to ICU admission. The prevalence of CRE carriage at baseline (at 0 hours) was 17.6% (95% CI 9.7-28.2). Klebsiella at 4/13(38.46%), Citobacter 5/13(38.46%) and Escherichia coli 4/13(30.7%) at baseline Incidence of CRE acquisition on the ICU ward was 1.5 persons/1000 person-hours. Sedative analgesic therapy (OR 0.01,95 Cl 0.0003-0.070, p=0.032), and buying ready to eat vegetables (OR O.O4, 95 Cl 0.001-0.98 p=0.049) increased the risk of CRE carriage at baseline. Parenteral nutrition was associated with CRE acquisition (OR 9.25,95 Cl. 1.03-82.76, p=0.047). Conclusion. Based on literature from Mexico, Greece and Spain we had a higher prevalence and incidence of CRE. We recommend a study with a bigger sample size, perform molecular relatedness and identify the genes associated with CRE in our setting. Further more active surveillance, isolation of CRE positive patients and continuous medical education on CRE.