Disentangling the wicked problem of antibiotic resistance
Omulo, Sylvia Adhiambo
MetadataShow full item record
Antimicrobial resistance (AMR) is currently an urgent public health concern worldwide. For low-income regions such as sub-Saharan Africa, absence of surveillance data and systematic investigations of risk factors prevents clear assessment of the impact that AMR has in these communities. There is also no empirically-based guidance for intervention strategies. In chapter one, we use PRISMA guidelines to review 116 journal articles from eastern Africa concerning AMR in people and animals, with the aim of understanding the scale of the problem and contributors to the problem. Over the last 40 years, AMR has reportedly increased, with unregulated antibiotic use invoked to explain the high prevalence of AMR in the region. Nevertheless, we note critical gaps: (i) consistent speculation about the risk factors for AMR without critical examination; (ii) no standardized system of testing and reporting of AMR data; and (iii) lack of community-based studies on AMR in people. Chapter two reports the outcome of two cross-sectional surveys on antibiotic use in an informal settlement (slum) in Kenya. Results from this survey show that residents generally have limited and inconsistent knowledge of antibiotics, including a perceived effectiveness of antibiotics for colds and flu. Chapter three explores how variation in lab resources and deviations from prescribed laboratory protocols might impact estimates of AMR prevalence. For the set of factors investigated, i.e. culture conditions, sample storage, isolate selection and repeated re-isolation, we observe that subtle variations are unlikely to significantly alter prevalence estimates. In chapter four, we use a longitudinal study design to examine how sanitation and antibiotic use impact household- and individual AMR loads in households living within an informal settlement in Kenya. While there is a relationship between reported antibiotic use and resistance to sulfamethoxazole and trimethoprim in children, most of the variance in the distribution of resistant bacteria is explained by sanitation variables. We conclude that although it is possible to design educational interventions based on our findings (chapter 2), such interventions are not likely to have a significant impact on community-level AMR when sanitation challenges continue to impact these communities (chapter 4).