Date of Conferral

11-5-2025

Date of Award

November 2025

Degree

Ph.D.

School

Public Health

Advisor

David Segal

Abstract

Timely digital surveillance is essential because clinical reporting often lags. This quantitative, longitudinal correlational study tested whether weekly Google searches in Austria signal changes in COVID-19 cases and vaccination uptake, and short-term case-vaccination dynamics, guided by Rogers’ Diffusion of Innovations Theory. Weekly Google Trends indices (0-100) and nationally reported surveillance totals (weekly counts of new cases and vaccine doses) were aggregated; cases: N = 175 weeks (Feb 2020–Jun 2023); vaccinations: N = 132 weeks (Dec 2020–Jun 2023). Time series correlations and lagged growth models showed that symptom-related searches consistently rose 1 week before reported cases (Δlog cases_t on Δlog searches_{t−1}: β = 0.72–1.33 across symptom/general/testing/fever, all p ≤ .015; R² = .16–.42); brand-specific vaccine searches (Pfizer, Moderna, AstraZeneca) preceded increases in vaccinations by ~1 week (β = 0.155–0.239, p = .003 ≤ .001; R² ≈ .06); and booster shot searches preceded vaccination activity by ~4 weeks (β = 0.130, p = .003; R² = .038). Cases and vaccinations showed no meaningful same-week association (Spearman’s ρ = −0.033, p = .709), and short-lag analyses provided little evidence that growth in cases drove growth in vaccinations (Δlog vaccinations_t on Δlog cases_{t−k}, k = 0–2: β = 0.10–0.16, all p ≥ .096; R² ≤ .021). Influenza-related search models were small or null, supporting COVID-19 specificity (cases at +1 wk: β = 0.115, p = .001, R² = .043; vaccinations at lags 0–2: ns). The positive social change implications from these findings may help support the use of Google Trends as a complementary, short-term, early-awareness surveillance approach for timelier targeted outreach by the Austrian public health authorities.

Download

Included in

Epidemiology Commons

Share

 
COinS