Bacteremia in critical care units at Bugando Medical Centre, Mwanza, Tanzania: the role of colonization and contaminated cots and mothers’ hands in cross-transmission of multidrug resistant Gram-negative bacteria

Abstract Background Multidrug resistance (MDR) is a major clinical problem in tertiary hospitals in Tanzania and jeopardizes the life of neonates in critical care units (CCUs). To better understand methods for prevention of MDR infections, this study aimed to determine, among other factors, the role of MDR-Gram-negative bacteria (GNB) contaminating neonatal cots and hands of mothers as possible role in transmission of bacteremia at Bugando Medical Centre (BMC), Mwanza, Tanzania. Methods This cross-sectional, hospital-based study was conducted among neonates and their mothers in a neonatal intensive care unit and a neonatology unit at BMC from December 2018 to April 2019. Blood specimens (n = 200) were sub-cultured on 5% sheep blood agar (SBA) and MacConkey agar (MCA) plates. Other specimens (200 neonatal rectal swabs, 200 maternal hand swabs and 200 neonatal cot swabs) were directly inoculated on MCA plates supplemented with 2 μg/ml cefotaxime (MCA-C) for screening of GNB resistant to third generation cephalosporins, r-3GCs. Conventional biochemical tests, Kirby-Bauer technique and resistance to cefoxitin 30 μg were used for identification of bacteria, antibiotic susceptibility testing and detection of MDR-GNB and screening of potential Amp-C beta lactamase producing GNB, respectively. Results The prevalence of culture confirmed bacteremia was 34.5% of which 85.5% were GNB. Fifty-five (93.2%) of GNB isolated from neonatal blood specimens were r-3GCs. On the other hand; 43% of neonates were colonized with GNB r-3GCs, 32% of cots were contaminated with GNB r-3GCs and 18.5% of hands of neonates’ mothers were contaminated with GNB r-3GCs. The prevalences of MDR-GNB isolated from blood culture and GNB r-3GCs isolated from neonatal colonization, cots and mothers’ hands were 96.6, 100, 100 and 94.6%, respectively. Significantly, cyanosis (OR[95%CI]: 3.13[1.51–6.51], p = 0.002), jaundice (OR[95%CI]: 2.10[1.07–4.14], p = 0.031), number of invasive devices (OR[95%CI]: 2.52[1.08–5.85], p = 0.031) and contaminated cot (OR[95%CI]: 2.39[1.26–4.55], p = 0.008) were associated with bacteremia due to GNB. Use of tap water only (OR[95%CI]: 2.12[0.88–5.09], p = 0.040) was protective for bacteremia due to GNB. Conclusion High prevalence of MDR-GNB bacteremia and intestinal colonization, and MDR-GNB contaminating cots and mothers’ hands was observed. Improved cots decontamination strategies is crucial to limit the spread of MDR-GNB. Further, clinical presentations and water use should be considered in administration of empirical therapy whilst awaiting culture results.