Impact of antibiotic intake on the incidence of extended-spectrum ß-lactamase-producing Enterobacterales in sub-Saharan Africa: results from a community-based longitudinal study.

OBJECTIVES: The global prevalence of intestinal extended-spectrum ß-lactamase-producing Enterobacterales (ESBL-PE) is approximately 17% in communities, with significant variations among regions. This longitudinal study aimed to assess the impact of antibiotic intake on the incidence of intestinal ESBL-PE in Ghanaian pharmacy customers outside of hospitals. METHODS: Screening for ESBL-PE was performed in four independent pharmacies in Kumasi, Ghana, using rectal swabs and an ESBL-PE-selective medium. Pharmacy customers purchasing antibiotics were recruited, and those buying non-antibiotic drugs served as controls. Participants who were negative for ESBL-PE provided follow-up swabs for up to 28 days. RESULTS: At baseline, 302 (75%) of 404 participants were colonized with ESBL-PE. Sixty-three participants who were negative for ESBL-PE at baseline received per-protocol follow-up, including 28 individuals who took antibiotics and 35 controls. The cumulative proportions of ESBL-PE in the antibiotics and control groups were 71% (20/28) and 54% (19/35) at the first follow-up (p 0.258), 86% (24/28) and 80% (28/35) at the second follow-up (p 0.741) and 86% (24/28) and 94% (33/35) at the third follow-up (p 0.393), respectively. DISCUSSION: The rate of intestinal ESBL-PE carriage among pharmacy customers outside of hospitals was higher than expected at baseline and further increased during the 28 days of follow-up, irrespective of antibiotic intake. This alarming finding needs to be considered in the antibiotic treatment of outpatients and emphasizes the urgent need for improved prevention strategies, development of new antibiotic drugs and potential future elimination strategies. Further longitudinal studies on ESBL-PE in African communities, also outside of pharmacy settings, are required.

Diagnostic performance of capillary and venous blood samples in the detection of Loa loa and Mansonella perstans microfilaraemia using light microscopy.

BACKGROUND: Loa loa and Mansonella perstans-the causative agents of loiasis and mansonellosis-are vector-borne filarial parasites co-endemic in sub-Saharan Africa. Diagnosis of both infections is usually established by microscopic analysis of blood samples. It was recently established that the odds for detecting Plasmodium spp. is higher in capillary (CAP) blood than in venous (VEN) blood. In analogy to this finding this analysis evaluates potential differences in microfilaraemia of L. loa and M. perstans in samples of CAP and VEN blood. METHODS: Recruitment took place between 2015 and 2019 at the CERMEL in Lambaréné, Gabon and its surrounding villages. Persons of all ages presenting to diagnostic services of the research center around noon were invited to participate in the study. A thick smear of each 10 microliters of CAP and VEN blood was prepared and analysed by a minimum of two independent microscopists. Differences of log2-transformed CAP and VEN microfilaraemia were computed and expressed as percentages. Furthermore, odds ratios for paired data were computed to quantify the odds to detect microfilariae in CAP blood versus in VEN blood. RESULTS: A total of 713 participants were recruited among whom 52% were below 30 years of age, 27% between 30-59 years of age and 21% above 60 years of age. Male-female ratio was 0.84. Among 152 participants with microscopically-confirmed L. loa infection median (IQR) microfilaraemia was 3,650 (275-11,100) per milliliter blood in CAP blood and 2,775 (200-8,875) in VEN blood (p<0.0001), while among 102 participants with M. perstans this was 100 (0-200) and 100 (0-200), respectively (p = 0.44). Differences in linear models amount up to an average of +34.5% (95% CI: +11.0 to +63.0) higher L. loa microfilaria quantity in CAP blood versus VEN blood and for M. perstans it was on average higher by +24.8% (95% CI: +0.0 to +60.5). Concordantly, the odds for detection of microfilaraemia in CAP samples versus VEN samples was 1.24 (95% CI: 0.65-2.34) and 1.65 (95% CI: 1.0-2.68) for infections with L. loa and M. perstans, respectively. CONCLUSION: This analysis indicates that average levels of microfilaraemia of L. loa are higher in CAP blood samples than in VEN blood samples. This might have implications for treatment algorithms of onchocerciasis and loiasis, in which exact quantification of L. loa microfilaraemia is of importance. Furthermore, the odds for detection of M. perstans microfilariae was higher in CAP than in VEN blood which may pre-dispose CAP blood for detection of M. perstans infection in large epidemiological studies when sampling of large blood quantities is not feasible. No solid evidence for a higher odds of L. loa microfilariae detection in CAP blood was revealed, which might be explained by generally high levels of L. loa microfilaraemia in CAP and VEN blood above the limit of detection of 100 microfilariae/ml. Yet, it cannot be excluded that the study was underpowered to detect a moderate difference.