Evidence Synthesis for Infectious Diseases

Abstract

Despite reductions in mortality due to improved sanitation and hygiene, the advent of antibiotics and childhood vaccination, infectious diseases continue to pose a threat to human health worldwide. This underscores the need for policy and action based on the best scientific evidence. The goal of this thesis is to advance evidence syntheses for infectious diseases. This goal is achieved through four chapters, where each chapter presents a unique evidence synthesis methodology solution to a unique infectious diseases research challenge, with examples drawn from streptococcal toxic shock syndrome (STSS), skin and soft tissue infections (SSTIs) and coronavirus disease 2019 (COVID-19). Inferences of prognostic factors from individual studies are limited. To address this challenge, the first chapter applies meta-analysis to increase the statistical power to detect associations between prognostic factors and outcomes in STSS patients. Network meta-analysis (NMA) provides the statistical framework to compare both treatment options directly, that is, based on within trial comparisons, and indirectly, where treatment options not previously studied in head-to-head comparisons are assessed through the use of a common comparator. For SSTIs, many antibiotics of potential interest have not been assessed in head-to-head randomized trials. The second chapter applies NMA to address this challenge and evaluate the comparative effectiveness of antibiotics for the treatment of SSTIs. Clinicians, patients, guideline bodies and government agencies faced challenges in interpreting the results from COVID-19 trials that were being published at a rate never encountered previously. The third chapter applies living NMA to present a complete, broad, and up-to-date view at the time of publication of the evidence base addressing prophylaxis for COVID-19. Multivariate network meta-analysis (MVNMA) builds on NMA by incorporating evidence from correlated outcomes. How MVNMAS are being conducted and reported in the healthcare literature is unclear. The fourth chapter is a scoping review and is the first study to systematically identify, map and describe this information for MVNMAs. It may help inform future conduct of MVNMAs and research on standardizing MVNMA methods. Further, these advancements in evidence syntheses for infectious diseases may better inform clinical and public health guidance, policy, and future research, which can have both patient- and population-level impacts.