EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Toward Proactive Overdose Prevention Through Spatiotemporal Cluster Prediction

Zhengyu Lin, Haoran Qiu, Elena García

Academic Journal of Applied Sciences · 20.05.2026 · 1 (3) · 121-129

DOI: 10.54097/swk9h978 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Academic Journal of Applied Sciences, online erschienen am 20.05.2026. Verfasst von Zhengyu Lin und 2 weiteren Autor:innen. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Drug overdose mortality in the United States has reached unprecedented levels, with opioid-involved deaths surpassing 80,000 annually and continuing to escalate despite sustained public health investment. The reactive architecture of most existing prevention strategies—deploying naloxone, expanding treatment capacity, or routing outreach workers only after overdose clusters have already been documented—has proven insufficient to prevent loss of life at scale. This paper proposes and evaluates a proactive spatiotemporal cluster prediction (SCP) framework designed to shift overdose surveillance from retrospective hotspot mapping toward prospective risk forecasting. By integrating historical overdose death records, socioeconomic covariates derived from the American Community Survey (ACS), and space-time permutation scan statistics with Bayesian hierarchical regression, the proposed framework generates one-year-ahead census-tract-level predictions of overdose cluster probability. Applied to publicly available county and tract-level mortality data from 2015 through 2022, the model identifies the top 10% of predicted high-risk tracts as the locus of approximately 74% of observed overdose deaths in held-out evaluation years. Findings indicate that socially vulnerable, economically deprived, and historically clustered tracts account for a disproportionate share of forward-predicted high-risk areas, and that proximity to naloxone access points independently attenuates predicted cluster formation risk. The paper discusses implications for pre-positioning harm reduction resources, mobile health service routing, and policy-level allocation of overdose prevention funding.

Zitieren

APA

Lin, Z., Qiu, H., & García, E. (2026). Toward Proactive Overdose Prevention Through Spatiotemporal Cluster Prediction. Academic Journal of Applied Sciences, 1(3), 121-129. https://doi.org/10.54097/swk9h978

Vancouver

Lin Z, Qiu H, García E. Toward Proactive Overdose Prevention Through Spatiotemporal Cluster Prediction. Academic Journal of Applied Sciences. 2026;1(3):121-129. doi:10.54097/swk9h978

BibTeX
@article{Lin2026,
  author = {Lin, Zhengyu and Qiu, Haoran and García, Elena},
  title = {Toward Proactive Overdose Prevention Through Spatiotemporal Cluster Prediction},
  journal = {Academic Journal of Applied Sciences},
  year = {2026},
  volume = {1},
  number = {3},
  pages = {121--129},
  doi = {10.54097/swk9h978}
}
Kostenlos starten →