The association between the use of oral fluoroquinolones and rare cephalic adverse events

Oral fluoroquinolones are one of the commonly prescribed antibiotic classes. Safety concerns on their use have been raised in recent decades. However, population-based evidence is lacking with regard to their safety profile in the cephalic region, including adverse events such as retinal detachment, seizure, and neuropsychiatric events. These events are rare, therefore population-based observational studies are more suitable to examine their association with oral fluoroquinolones as compared to randomised controlled trials. The increasing use of population-based studies has led to development of new observational study design which addresses the limitations of conventional study designs. The self-controlled case series design is one of the newer methods which performs within person comparison in the analysis to minimise bias from fixed confounding. It is suitable for exploring the association between short-term exposure and acute events, which is appropriate for studying the safety of antibiotics such as fluoroquinolones. The aim of this thesis is to investigate the association between oral fluoroquinolones and retinal detachment, seizure, and neuropsychiatric events using the self-controlled case series design to provide evidence to broaden the understanding of its safety profile. The studies were conducted using data from the Hong Kong Clinical Data Analysis and Reporting System to complement fluoroquinolone safety profile with an Asian perspective. Databases from Taiwan and the United Kingdom were also used for external validation. The first part presents the study on the association between oral fluoroquinolones and retinal detachment in Hong Kong and Taiwan. No increased risk was shown in both databases. The crude absolute risks among current oral fluoroquinolones users were small. A systematic review and meta-analysis was then conducted to review the findings of the published literature and to draw an overall conclusion. No significant increased risk was found, refuting the reported association. The second part details the study on the association between oral fluoroquinolones and seizure in Hong Kong and the United Kingdom. Significant increased risk was observed before the commencement of the exposure in both databases. The findings suggested that the indication for which oral fluoroquinolones were prescribed was associated with seizure instead of the exposure itself. The crude absolute risk of incident seizure was also very small in both databases. The last part presents the study on the association between oral fluoroquinolones and acute neuropsychiatric events. An additional nested case-control study was also conducted to demonstrate the robustness of the self-controlled case series analysis. Both analyses showed increased risk during and after the completion of oral fluoroquinolones exposure, which indicated that neuropsychiatric events were associated with the exposure. The crude absolute risk was small but was much higher than that reported in the published literature. The research undertaken in this thesis suggests that oral fluoroquinolones have a generally safe neurological and ocular profile except for the possibility of association with acute neuropsychiatric events. Healthcare databases provide a useful platform for new observational study designs to broaden the understanding of medication safety. Future studies can be conducted using different databases to ensure robustness and to expand generalisability of the findings.published_or_final_versionPharmacology and PharmacyDoctoralDoctor of Philosoph