(283) An Assessment of The Risk of Anaphylaxis Following Influenza Vaccination in The Korean Elderly: Implementing Near-Real Time Vaccine Safety Surveillance Methods

Sunday, August 28, 2022

8:00 AM – 1:45 PM CEST

Location: Hall C

Publication Number: 1328

Presenting Author(s)

Na-Young Jeong
– Ewha Womans University, Seodaemungu, Gyeonggido, South Korea

Co-Author(s)

Heejin Kim
– Ewha Womans University, Seodaemungu, South Korea
EUNSUN LIM
– Seodaemun-gu, South Korea
Byung-Joo Park
– Department of Preventive Medicine, Seoul National University College of Medicine, Seoul, Republic of Korea., Republic of Korea
Nam-Kyong Choi
– Department of Health Convergence, College of Science & Industry Convergence, Ewha Womans University, Seoul, Republic of Korea, Republic of Korea

Background: Influenza vaccines are administered for the elderly every year under the National Immunization Program (NIP) in Korea. However, changes of vaccine antigenic components and the administration policy in each flu season may lead to a new risk of adverse events such as anaphylaxis. Near-real time surveillance is needed to detect the risk of adverse events as early as possible, but it has not yet been implemented in Korea.

Objectives: To examine whether there was an increased risk of anaphylaxis following influenza vaccination in the elderly between 2015/16 to 2017/18 season and the feasibility of weekly near-real time surveillance

Methods: This study included the elderly who received influenza vaccines during three flu seasons using a large-linked database of NIP database and the national health insurance claims database. We excluded people who received more than one flu vaccination for each season. For cases of anaphylaxis, patients who diagnosed with ICD-10 code T78.2 or T88.6 and prescribed with epinephrine or corticosteroids and who have no anaphylaxis history during the 183 days before vaccination were identified. We used a self-controlled risk interval (SCRI) design in which risk window (RW) was days 0-2 and control window (CW) was days 7-13 [CW1] or days 14-20 [CW2] following vaccination to estimate incidence rate ratios (IRRs). As sequential analyses, we used SCRI design with the binomial-based maxSPRT to estimate log-likelihood ratio weekly in each season. The endpoint of surveillance was set when the signal was detected, maximum sample size was reached, or the season ended. Then we compared to the result of sequential and non-sequential methods in terms of the direction and the time of signal detection.

Results: We identified 3,715,453 elderly who received influenza vaccines. For CW1, the risk of anaphylaxis in RW was statistically significant in 2016/17 season (IRR=1.98, 95% CI=1.04-3.79) and 2017/18 season (IRR=2.22, 95% CI=1.20-4.10), excluding 2015/16 season. However, the risk of anaphylaxis was not significantly increased when using CW2. There was no significant risk of anaphylaxis in sequential analysis in 2015/16 season, but it was difficult to find the temporal advantage. Signal was detected in the 2016/17 and 2017/18 seasons, and we confirmed that the signal could be detected early in sequential analysis as the surveillance ended in the 6th and 4th weeks, respectively.

Conclusions: In this study, analyses compared to CW1 confirmed that the risk of anaphylaxis significantly increased after flu vaccination in the 2016/17 and 2017/18 seasons. In the case of sequential analysis, it was confirmed that signal could be detected generally earlier than non-sequential analysis.