Antibiotic Use and Adverse Events

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): Estimates of antibiotic exposure in the U.S. population range from once every 3 years to almost twice yearly; almost half of U.S. outpatient antibiotic use is unnecessary. Antibiotic overuse has individual and societal consequences, including adverse drug events (ADEs) and escalating antimicrobial resistance. Efforts to limit overuse at the patient-provider encounter level have had limited success. Even relatively small individual ADE risks from antibiotic use could result in large population attributable risks. Few studies on antibiotic ADEs have included a control group of patients without antibiotic exposure. While randomized clinical trials are ideal for ensuring comparability between exposed and unexposed groups, they are not always feasible. Observational data could offer an efficient way to study individual consequences of antibiotic use, but studies need to address confounding issues, especially confounding by indication. Enhanced adjustment methods to ensure comparability between treatment and control groups would thus help us make optimal use of observational data. This retrospective cohort study uses observational data from the UK's General Practice Research Database to study ADEs consequent to antibiotic use. This study takes advantage of the fact that antibiotic treatment for acute respiratory infections (ARIs) is variable, reflecting clinician beliefs, patient preferences and underlying clinical and nonclinical factors. The primary aim is to compare the risk of a severe adverse event between patients prescribed antibiotics, conditional on an office visit for an ARI vs. the risk for patients with ARI office visits not exposed to antibiotics. The hypothesis is that antibiotic use is associated with an increased risk of adverse events; any increased risk can be attributable to ADEs. Secondary aims are to compare the risk of a severe adverse event between patients with ARIs exposed to different classes of antibiotic drugs, and to compare the risk of a less-severe adverse event between patients with ARIs exposed to antibiotics vs. the risk for unexposed patients. We will explore methods to control for confounding by patient clinical factors such as comorbid conditions and visit frequency. This study will support the applicant's development into a productive independent investigator. This study is relevant to public health as it proposes to assess the individual risk of antibiotic-related ADEs and refine methodologies to address confounding issues inherent with using observational data. Results would help plan a prospective study using electronic medical record data to more effectively incorporate individual risks and benefits into antibiotic prescribing decisions, minimize adverse outcomes of antibiotic prescribing, and decrease unnecessary antibiotic use.