Effects of Subject-level Characteristics on Influenza Illness and Vaccination Open Access

Elkind, Polina (2016)

Permanent URL: https://etd.library.emory.edu/concern/etds/wp988k19x?locale=en


A cohort study of 1,426 subjects utilized 2012-2013 influenza season data to carry out two objectives: firstly, to determine risk factors associated with influenza contraction, as well as the effect of vaccination against influenza infection, adjusting for various subject-level characteristics (race, sex, age, household size, and health risks); and, secondly, to evaluate the associations between these characteristics and vaccination status. A Cox proportional hazards regression model indicated that those who were both effectively vaccinated and from 4-member households (HR=0.45, p=0.006) were the least likely to contract influenza when compared to their respective reference group. Being 6 months-8 years of age (HR=1.55, p=0.047) was associated with a higher risk of contracting influenza. Adjusted vaccine effectiveness in the overall population was found to be 55% (CI95% [20, 74]). Adults experienced significant protection, with a VE of 49% (CI95% [2, 74]), but neither age category for children indicated significant protection from the flu due to effective vaccination. VE was also significant and protective for individuals from 4-member households (56%, CI95% [23, 75]). Further, having been 6 months-8 years of age (OR=1.47, CI95% [1.14, 1.90]) or 9-17 years of age (OR=1.60, CI95% [1.22, 2.10]) were protective characteristics and yielded statistically significant associations with vaccination status. Having these characteristics increased the odds - in comparison to each characteristic's reference group - that an individual received a vaccination. Additionally, the interaction between health risks and sex indicated that females with health risks (OR=2.46, CI95% [1.46, 4.12]), females without health risks (OR=1.43, CI95% [1.14, 1.79]), and males with health risks (OR=4.42, CI95% [2.24, 8.72]) all experienced significantly greater odds of vaccination when compared to males without health risks. Subjects who were of black (OR=0.65, CI95% [0.43, 0.96]) or other/unknown race (OR=0.52, CI95% [0.35, 0.79]), or lived in a household consisting of 5+ members (OR=0.78, CI95% [0.63, 0.97]) were less likely to be vaccinated when compared to their respective reference groups.

Table of Contents

1. Introduction and Review of the Literature p.1

2. Methodology p.3

2.1. Study Overview p.3

2.2. Objectives p.4

2.2.1. Objective 1: Evaluating Factors Associated with Contracting Influenza. p.4

2.2.2. Objective 2: Evaluating Factors Associated with Receiving an Influenza Vaccination p.5

2.3. Statistical Methods p.6

2.3.1. Methods for Univariate Analysis p.6

2.3.2. Methods for Objective 1: Cox Proportional Hazards Regression Model p.6

2.3.3. Methods for Objective 2: Logistic Regression Model p.7

3. Results p.8

3.1. Descriptive Statistics and Univariate Results p.8

3.2. Results for Objective 1 p.10

3.3. Results for Objective 2 p.14

4. Discussion p.17

5. Limitations and Future Analyses p.18

6. References p.20

7. Appendices p.22

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
Subfield / Discipline
  • English
Research Field
Committee Chair / Thesis Advisor
Committee Members
Last modified

Primary PDF

Supplemental Files