CONFERENCE PROCEEDING
Structural fragility in digital public health campaigns: A longitudinal network analysis of U.S. e-cigarette prevention (2014–2023)
More details
Hide details
1
Social Data Collaboratory, NORC at the University of Chicago, Chicago, Illinois, USA
2
Departments of Communication and Political Science, University of Illinois-Chicago, Chicago, Illinois, USA
Tob. Prev. Cessation 2026;12(Supplement 1):A74
ABSTRACT
BACKGROUND-AIM:
Digital public health campaigns often report robust reach and engagement metrics, yet these surface-level indicators obscure deeper structural vulnerabilities that limit long-term influence. When recent e-cigarette prevention campaign messages outperformed oppositional messaging (vaping advocacy and opposition to regulations), prevention campaigns faced structural challenges to maintain visibility and influence within contested digital spaces. We introduce the concept of structural fragility, defined as a measurable decline in the collective capacity of a class of actors to maintain network visibility amid algorithmic and commercial pressures. This study assesses structural fragility among health organizations and professionals across nearly a decade of Twitter discourse surrounding U.S. e-cigarette and tobacco prevention campaigns (2014–2023).
METHODS:
We classified users into five categories (Health, Commercial, Vape Community, Organic, Other) based on account characteristics. Using longitudinal network analysis, we tracked three dimensions of structural fragility: positional (amplification and centrality), community (fragmentation and homophily), and core (presence in the network’s k-core). Annual directed retweet networks were constructed, and metrics such as modularity, in-degree, eigenvector centrality, and k-core composition were computed using NetworkX and Louvain algorithms.
RESULTS:
Over time, the network (active users) contracted by 99.7%, fragmented into homophilous clusters (modularity rose from 0.16 to 0.84), and displaced health actors from the network’s core—most notably in 2015, coinciding with FDA’s Deeming Rule proposal. These changes reflected a “bunker effect” where public health persisted in broadcasting content even as they became increasingly isolated from the network’s core. Per-capita amplification efficiency for health actors declined by 95.7%, with remaining influence concentrated among a small elite.
CONCLUSIONS:
These findings challenge exposure-based evaluation models and reveal how platform architectures consistently demote and isolate institutional health communication. We propose a network-informed framework for campaign design and policy resilience that accommodates diverse harm reduction perspectives while safeguarding evidence-based prevention messaging. Implications extend to health professionals, authorities, and policymakers seeking effective digital governance. Tobacco endgame, strategies and policies to permanently end the tobacco