245 - Concordance between claims and survey-based measures of children’s influenza vaccination

Poster Number: 245
Publication Number: 245.1756

Background: The study of seasonal vaccines is complicated by measurement error due to inadequate capture in claims data and recall error for surveys. Understanding the constraints of the available data can increase awareness of the potential misclassification of individual vaccine receipt, help interpret findings in light of these limitations, and ultimately lead to improved study design.

Objective: 1) To evaluate concordance between parent report of a child’s flu vaccine receipt and insurance claims documenting receipt over the course of two seasons (2019-2020, 2020-2021) and 2) to determine whether lack of concordance between parental report and claims varies by reported site of vaccination.

Design/Methods: In 2022, we surveyed 517 U.S. parents of children aged 2- to 18-years covered by a non-profit commercial health insurance plan. Children were eligible if insurance claims indicated inconsistent vaccination, defined as no influenza vaccination claims in the 2019-2020 influenza season and ≥1 vaccination claim in the 2020-2021 season. For each season, we assessed concordance between claims and parental report and parent report of location of vaccination using descriptive statistics. We used chi square tests to compare demographics of those who were concordant or discordant.

Results: For the 2019-2020 season, less than half of parental reports (44%) were concordant with claims data indicating no influenza vaccination. For the 2020-2021 season, almost two-thirds of parental reports (64%) were concordant with claims indicating vaccination. Very few parental reports (10%) were concordant with claims for both seasons. Of the discordant individuals who reported vaccinating their child in the 2019 season in the absence of a claim for flu vaccination, the majority reported that their child received the vaccine at a site that typically bills insurance, such as a doctor’s office (81%) or a pharmacy (5%) (Figure 1). There were no significant demographic differences between those whose responses were concordant or discordant with claims in both seasons.

Conclusion(s): Concordance between parental report of child influenza vaccination and insurance claims is low, particularly in more distant seasons. This highlights the possibility that claim capture of flu vaccination may be inconsistent and that parental recall may be poor beyond the immediately preceding influenza season. Further work is needed to identify a gold standard data source for reliable influenza vaccine capture to increase accuracy of future studies.