The epidemiology of acute febrile illness in rural South and Southeast Asian primary care, and evaluation of selected interventions to improve its management

Acute febrile illness (AFI) accounts for a major proportion of presentations in primary care, particularly in tropical low-income and middle-income countries (LMICs). The paradigm that tropical AFI is primarily attributable to malaria is no longer valid in South and Southeast Asia, yet there is precious little high-quality contemporary evidence on its epidemiology, especially in the rural, resource-constrained, and under-studied areas containing most of the regional population. Clinical and health policy decision-making is, thus, hampered leading to larger-scale problems such as antimicrobial over-prescription and development of resistance, and hospital overcrowding from poorly-targeted referrals. Viewing this pressing but neglected problem through a pragmatic, multi-disciplinary lens, this thesis aims to lay the foundations for the development of interventions to improve primary care management of AFI in rural South and Southeast Asia at both individual patient and population levels. Beginning with two multi-national prospective studies embedded in the South and Southeast Asian Community-based Trials Network Rural Febrile Illness project, it details the baseline contextual and epidemiological realities confronting health workers and policymakers in Bangladesh, Cambodia, Laos, Myanmar, and Thailand daily, which have thus far been poorly researched. The first showed that while there were considerable differences between these sites, all relied on primary health centres and village health/malaria workers as the main providers of primary healthcare. The key results from the second, an observational study, were that in the 82,760 patients enrolled by these two provider groups the most frequent diagnoses were upper respiratory tract infection (URTI, 61.0%) and fever of unclear source (30.7%). This study also showed that targeting of antibiotic prescriptions and identification of patients with potentially severe illness could be improved, although mortality was very low (0.01%). Building on the results, a variety of research methodologies based on real-world data were used to generate evidence on potential interventions, starting with one commonly used in high-income settings (pulse oximetry), followed by one for which the need is well-established but data for its development is lacking (a multiplex rapid diagnostic test for AFI). With regard to pulse oximetry, a systematic review concluded that it may assist clinicians in diagnosing and managing paediatric pneumonia, but for the greatest impact on patient outcomes should be implemented as part of a health systems approach. For the latter, the most important finding from a priority-setting Delphi survey was that it should be able to diagnose dengue and enteric fever at a minimum in adults and children. Health economic modelling studies showed that both these interventions were likely to be cost-effective in the settings which they were evaluated. The culminating work describes the development of a novel algorithmic electronic clinical decision support tool which integrates epidemiological and clinical data to guide diagnosis and management, and the methodology for its evaluation via a cluster-randomised trial in Cambodia. The generalisability of the knowledge gained from this thesis to other tropical LMICs, as well as the barriers to its application and implementation, are also discussed, along with potential solutions.