Estimating influenza vaccine effectiveness among older adults using an integrated administrative database and the implications of potential bias: A population-based cohort study in Japan.
{"title":"Estimating influenza vaccine effectiveness among older adults using an integrated administrative database and the implications of potential bias: A population-based cohort study in Japan.","authors":"Ayu Kasamatsu, Yuichiro Yahata, Wakaba Fukushima, Hirofumi Sakamoto, Kaori Tanaka, Miwa Takigawa, Kaori Izu, Yuko Nishino, Motoi Suzuki, Hajime Kamiya","doi":"10.1016/j.vaccine.2024.126488","DOIUrl":null,"url":null,"abstract":"<p><strong>Background: </strong>Japan lacks an established framework for routine seasonal influenza vaccine effectiveness (SIVE) assessment at the national and municipal levels. This study aimed to estimate SIVE among older adults using an innovative population-based administrative database linking medical fee claims data with vaccination records, while also exploring its potential bias.</p><p><strong>Methods: </strong>In this retrospective population-based cohort study, we assessed SIVE against medically attended influenza during the 2017/18 season among older adults aged ≥65 years in a Japanese city. A Cox proportional hazards model was used to estimate hazard rate ratios, treating vaccination status as time-dependent. To explore potential biases, multivariate logistic regression analysis was used to investigate the association between vaccination status and acute respiratory infection (ARI) diagnosis and trauma/injury during the non-influenza season.</p><p><strong>Results: </strong>This study included 82 % (n = 110,892) of the city's older adult population, with 39.7 % vaccination coverage. The estimated SIVE was 2.9 % (95 % confidence interval: -6.2-11.2), showing no statistical significance. Similarly, subgroup analyses by age and comorbidities revealed no significant protective effect of SIVE. In the non-season analysis, adjusted odds ratios of vaccination were significantly higher for ARI [1.3 (1.3-1.4)] and trauma/injury [1.2 (1.1-1.2)]. However, no significance was observed for hospitalizations with these diagnoses, which include severe conditions less associated with healthcare-seeking behaviors [0.9 (0.8-1.1) and 0.8 (0.6-1.0), respectively].</p><p><strong>Conclusions: </strong>No significant SIVE was observed during the 2017/18 season. Our real-world observational study, based on medical fee claims data, indicates a potential underestimation of SIVE owing to bias related to healthcare-seeking behaviors.</p>","PeriodicalId":94264,"journal":{"name":"Vaccine","volume":"42 26","pages":"126488"},"PeriodicalIF":0.0000,"publicationDate":"2024-12-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Vaccine","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1016/j.vaccine.2024.126488","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"2024/10/31 0:00:00","PubModel":"Epub","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Background: Japan lacks an established framework for routine seasonal influenza vaccine effectiveness (SIVE) assessment at the national and municipal levels. This study aimed to estimate SIVE among older adults using an innovative population-based administrative database linking medical fee claims data with vaccination records, while also exploring its potential bias.
Methods: In this retrospective population-based cohort study, we assessed SIVE against medically attended influenza during the 2017/18 season among older adults aged ≥65 years in a Japanese city. A Cox proportional hazards model was used to estimate hazard rate ratios, treating vaccination status as time-dependent. To explore potential biases, multivariate logistic regression analysis was used to investigate the association between vaccination status and acute respiratory infection (ARI) diagnosis and trauma/injury during the non-influenza season.
Results: This study included 82 % (n = 110,892) of the city's older adult population, with 39.7 % vaccination coverage. The estimated SIVE was 2.9 % (95 % confidence interval: -6.2-11.2), showing no statistical significance. Similarly, subgroup analyses by age and comorbidities revealed no significant protective effect of SIVE. In the non-season analysis, adjusted odds ratios of vaccination were significantly higher for ARI [1.3 (1.3-1.4)] and trauma/injury [1.2 (1.1-1.2)]. However, no significance was observed for hospitalizations with these diagnoses, which include severe conditions less associated with healthcare-seeking behaviors [0.9 (0.8-1.1) and 0.8 (0.6-1.0), respectively].
Conclusions: No significant SIVE was observed during the 2017/18 season. Our real-world observational study, based on medical fee claims data, indicates a potential underestimation of SIVE owing to bias related to healthcare-seeking behaviors.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
