Association between seasonal influenza vaccination and antimicrobial use in Japan from the 2015-16 to 2020-21 seasons: from the VENUS study

Abstract

Background: Seasonal influenza vaccination might be considered an antimicrobial resistance (AMR) countermeasure because it can reduce unnecessary antimicrobial use for acute respiratory infection by mitigating the burden of such diseases.

Objectives: To examine the association between seasonal influenza vaccination and antimicrobial use (AMU) in Japan at the community level and to examine the impact of influenza vaccination on the frequency of unnecessary antimicrobial prescription for upper respiratory infection.

Methods: For patients who visited any healthcare facility in one of the 23 wards of Tokyo, Japan, due to upper respiratory infection and who were aged 65 years or older, we extracted data from the Vaccine Effectiveness, Networking, and Universal Safety (VENUS) study database, which includes all claims data and vaccination records from the 2015-16 to 2020-21 seasons. We used the average treatment effect (ATE) with 1:1 propensity score matching to examine the association of vaccination status with frequency of antibiotic prescription, frequency of healthcare facility consultation, risk of admission and risk of death in the follow-up period of the same season (from 1 January to 31 March).

Results: In total, 244 642 people were enrolled. Matched data included 101 734 people in each of the unvaccinated and vaccinated groups. The ATE of vaccination was -0.004 (95% CI -0.006 to -0.002) for the frequency of antibiotic prescription, -0.005 (-0.007 to -0.004) for the frequency of healthcare facility consultation, -0.001 (-0.002 to -0.001) for the risk of admission and 0.00 (0.00 to 0.00) for the risk of death.

Conclusions: Our results suggest that seasonal influenza vaccination is associated with lower frequencies of unnecessary antibiotic prescription and of healthcare facility consultation.

Figure 1.

Results of modified Poisson regression analyses. (a) RR of each variable to frequency of antibiotic prescription. (b) RR of each variable to frequency of healthcare facility consultation. (c) RR of each variable to risk of admission. (d) RR of each variable to risk of death.

Figure 2.

Balance of covariates before and after propensity score matching. Red squares represent data before adjustment and blue circles represent data after adjustment. Vertical dashed lines represent 0.10. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.