Invited Commentary: The Use of Population Attributable Fractions in Studies of Vaccine Hesitancy

Abstract

Vaccine hesitancy-the delay or refusal of vaccines despite their availability-has been linked to lower vaccination rates and outbreaks of vaccine-preventable diseases. Using cross-sectional surveys of 78,725 parents and other family members in the United States, Nguyen et al. (Am J Epidemiol. 2022;191(9):1626-1635) calculated the population attributable fraction (PAF) of vaccine hesitancy on nonreceipt of recommended childhood vaccines, including influenza vaccine. The PAF is readily calculated: p(rr - 1)/rr, where p is the proportion of those hesitant among nonvaccinated individuals, and rr is the risk ratio of nonvaccination between those hesitant over those nonhesitant. By vaccine, the PAF ranged from 6.5% for nonreceipt of the hepatitis B vaccine birth dose to 31.3% for nonreceipt of the diphtheria-tetanus-pertussis vaccine dose 3. For nonreceipt of influenza vaccine, the PAF varied geographically, with relatively high values in some Northeast (e.g., New York at 22.6%) and Northwest (e.g., Oregon at 23.0%) states and lower values in certain Southern (e.g., Louisiana at 7.5%) and Mountain West (e.g., Utah at 8.8%) states. The PAF can elucidate the contribution of vaccine hesitancy on nonvaccination in different circumstances. Future studies can apply this technique in different populations and incorporate different measures of vaccine hesitancy.

Keywords: childhood vaccines; influenza vaccines; population attributable fraction; vaccine confidence; vaccine hesitancy.