Parental Vaccine Hesitancy and Declination of Influenza Vaccination Among Hospitalized Children

Abstract

Objectives: Parents frequently decline the influenza vaccine for their child during hospitalization. In this study, we aimed to assess the role of vaccine hesitancy in these declinations.

Methods: This cross-sectional survey study was conducted among English-speaking parents of influenza vaccine-eligible children who were hospitalized between October 2014 and April 2015. Between July 2015 and September 2015, parents were recruited via mail to complete the validated Parent Attitudes about Childhood Vaccines (PACV) survey (modified for influenza vaccination). PACV scores (0-100 scale) were dichotomized into scores of ≥50 (hesitant) and <50 (nonhesitant). The primary outcome was parental declination of the influenza vaccine for their child during hospitalization. A secondary outcome was the declination reason documented during hospitalization. The main independent variable was parental vaccine hesitancy status, determined by the PACV score. Multivariable logistic regression was used to examine the association between vaccine hesitancy and influenza vaccine declination, adjusting for sociodemographic, visit, and clinical characteristics. The relationship between vaccine hesitancy and declination reason was also explored.

Results: Of 199 parents (18% response rate), 24% were vaccine hesitant and 53% declined the influenza vaccine for their child during hospitalization. Vaccine hesitancy (versus nonhesitancy) was associated with declining influenza vaccination (adjusted odds ratio: 6.4; 95% confidence interval: 2.5-16.5). The declination reason differed by vaccine hesitancy status, with a higher proportion of parents who were hesitant versus nonhesitant reporting "vaccine concern" or "vaccine unnecessary."

Conclusions: Vaccine hesitancy was prevalent in this limited sample of parents of hospitalized children and associated with influenza vaccine declination. Additional investigation in a large, diverse, prospectively recruited cohort is warranted given the potential sampling bias present in this study.