Date of Award

January 2015

Document Type

Open Access Thesis

Degree Name

Master of Public Health (MPH)

Department

School of Public Health

First Advisor

Daniel M. Weinberger

Abstract

BACKGROUND: Pneumococcal conjugate vaccine (PCV) is now recommended for use in adults in the United States. Because vaccination of children with PCVs protects adults from the targeted serotypes, local variations in PCV uptake among children could influence the effectiveness of direct immunization of adults. The effect of variations in PCV7 and PCV13 uptake in children will be compared to assess the possible effectiveness of PCV vaccination in adults.

METHODS: We obtained ZIP Code-level invasive pneumococcal disease (IPD) data from an active population-based surveillance system in Connecticut (1998-2013) and ZIP Code-level data on immunization with 3+ or 4+ doses of PCV from the state immunization registry. To estimate the decline in PCV serotypes in each age group during the PCV7 and PCV13 eras, we fit logistic regression models to estimate the proportion of IPD cases in a month caused by a vaccine serotype. To determine whether SES factors and/or vaccine uptake was associated with adult IPD cases we fit logistic regression models to estimate the proportion of IPD cases among adults >40 years that was caused by PCV-targeted serotypes. Covariates included ZIP Code-level socioeconomic indicators (e.g. percent black, income), percent of the population that received at least 3 (3+) or at least 4 (4+) doses of PCV (mean centered), and a linear spline to control for the average rate of decline across all ZIP Codes. The data were split into the 2000-2009 PCV7 era and the 2010-2013 PCV13 era to compare the effect of the two vaccines. In the PCV7 era we selected the best model for estimating the proportion of IPD cases among adults >40 years that was caused by PCV7-targeted serotypes by ranking the models by AIC score. In the PCV13 era we conducted AIC based model averaging to estimate the proportion of IPD cases among adults >40 years that was caused by PCV13-targeted serotypes.

RESULTS AND SIGNIFICANCE: The PCV7 serotypes declined rapidly as causes of IPD in children while the decline was slower and delayed in adults. A similar pattern was seen in the PCV13 serotype decline. ZIP Codes that had a higher proportion of children that did not complete the four-dose PCV7 series had a higher proportion of adult IPD cases caused by PCV7 serotypes. This trend is not yet evident in adults during the PCV13 era. Local variations in PCV uptake might influence the effectiveness of PCVs in preventing pneumococcal disease in adults.

Comments

This is an Open Access Thesis.

Share

COinS