(238) Association between repeated antibiotic prescribing and hospitalization among people living with dementia: a population-based cohort study

Background: Whether repeated prescriptions of some antibiotic classes can cause seizure or psychosis-related and all-cause hospitalization among people with dementia remains unknown.

Objectives: To understand the association of repeated antibiotic prescriptions with seizure or psychosis-related and all-cause hospitalization among people living with dementia.

Methods: This population-based cohort study retrieved data from electronic healthcare records from Hong Kong. Individuals with dementia and aged 65 or above with at least one seizure-related (penicillins and cephalosporins) or psychosis-related (penicillins, macrolides, quinolones and sulphonamides) high-risk antibiotic prescriptions in the past six months from 2004 to 2019 were included. The number of high-risk antibiotic prescriptions in the past six months was counted. Poisson regression models were fitted to estimate incidence rate ratios (IRRs) adjusted by propensity scores and 95% confidence intervals (CIs). The IRRs compare the seizure-, psychosis-related and all-cause hospitalization risks after a high-risk antibiotic prescription in the three higher quartiles of prior antibiotic use to the lowest, with a follow-up time of 30 days.

Results: We identified a total of 79,367 and 71,920 dementia patients who are aged 65 or over with seizure- and psychosis-related high-risk antibiotic prescriptions, respectively. The adjusted IRRs of seizure-related and all-cause hospitalization are progressively incremental and mostly statistically significant as the frequency of antibiotic prescriptions increases (The highest IRR is 1.57 (95%CI 1.16-2.15) for seizure in 30 days after prescription and 2.66 (95%CI 2.62-2.71) for all-cause). The adjusted IRRs of psychosis-related hospitalization are statistically non-significant, but the adjusted IRRs of all-cause hospitalizations are statistically significant (The highest IRR is 2.65* (95%CI 2.61-2.69)).

Conclusions: Frequent prescribing of high-risk antibiotics in the past six months in people with dementia is associated with increased risk of seizure-related and all- cause hospitalizations but is not associated with an increased risk of psychosis-related hospitalization. The findings suggest that clinicians should carefully balance the benefits of repeated antibiotic courses against their potential risks.