The predictors of full, partial and no COVID-19 vaccination among immigrants and non-immigrants

Abstract

Objective: To determine covariates linked to each COVID-19 vaccination level (any, partially and fully vaccinated) for working-age adults including immigrants.

Methods: Logistic regression models tested the three vaccination levels of a cohort (n = 119,373), randomly-selected from Israeli electronic medical records (years of 2019-2022), consisting of five groups (the total cohort, native-born, and Russia/USSR-born, USA-born, and France-born immigrants). Covariates identified in the existing literature were categorized according to Andersen's Behavioral Model's factors: predisposing (demographic characteristics and health behaviors), enabling (facilitating/impeding healthcare access) and need (risk factors).

Results: Obtaining any vaccinations, the most commonly used operationalization of COVID-19 vaccination status, exhibited the strongest model for all five groups, while the rarely tested dependent variables of partially and unvaccinated exhibited the weakest models, particularly for immigrant groups. Only the enabling factor's variables of socioeconomic status and/or family physician contact were consistently linked to all COVID-19 vaccination levels for the five groups.

Conclusions: Preventive health actions are hampered by the collection of disease-related data that is relevant to older adults but not to working-age adults or to at-risk populations, such as immigrants.

Keywords: COVID-19; Immigrants; Preventive care; Receiver operating characteristic (ROC); Vaccination status.