A latent process model for forecasting multiple time series in environmental public health surveillance

Kathryn T. Morrison, Gavin Shaddick, Sarah B. Henderson, David L. Buckeridge

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

This paper outlines a latent process model for forecasting multiple health outcomes arising from a common environmental exposure. Traditionally, surveillance models in environmental health do not link health outcome measures, such as morbidity or mortality counts, to measures of exposure, such as air pollution. Moreover, different measures of health outcomes are treated as independent, while it is known that they are correlated with one another over time as they arise in part from a common underlying exposure. We propose modelling an environmental exposure as a latent process, and we describe the implementation of such a model within a hierarchical Bayesian framework and its efficient computation using integrated nested Laplace approximations. Through a simulation study, we compare distinct univariate models for each health outcome with a bivariate approach. The bivariate model outperforms the univariate models in bias and coverage of parameter estimation, in forecast accuracy and in computational efficiency. The methods are illustrated with a case study using healthcare utilization and air pollution data from British Columbia, Canada, 2003-2011, where seasonal wildfires produce high levels of air pollution, significantly impacting population health.

Original languageEnglish
Pages (from-to)3085-3100
Number of pages16
JournalStatistics in medicine
Volume35
Issue number18
Early online date16 Feb 2016
DOIs
Publication statusPublished - 15 Aug 2016

Fingerprint

Public Health Surveillance
Latent Process
Multiple Time Series
Environmental Health
Public Health
Surveillance
Process Model
Forecasting
Health
Air Pollution
Environmental Exposure
Univariate
Outcome Assessment (Health Care)
British Columbia
Laplace Approximation
Morbidity
Model
Canada
Mortality
Computational Efficiency

Keywords

  • Air pollution
  • Bayesian hierarchical models
  • INLA
  • Latent processes
  • Time series

Cite this

A latent process model for forecasting multiple time series in environmental public health surveillance. / Morrison, Kathryn T.; Shaddick, Gavin; Henderson, Sarah B.; Buckeridge, David L.

In: Statistics in medicine, Vol. 35, No. 18, 15.08.2016, p. 3085-3100.

Research output: Contribution to journalArticle

Morrison, Kathryn T. ; Shaddick, Gavin ; Henderson, Sarah B. ; Buckeridge, David L. / A latent process model for forecasting multiple time series in environmental public health surveillance. In: Statistics in medicine. 2016 ; Vol. 35, No. 18. pp. 3085-3100.
@article{2f3a9e584e7a4909893bd7e07aec3447,
title = "A latent process model for forecasting multiple time series in environmental public health surveillance",
abstract = "This paper outlines a latent process model for forecasting multiple health outcomes arising from a common environmental exposure. Traditionally, surveillance models in environmental health do not link health outcome measures, such as morbidity or mortality counts, to measures of exposure, such as air pollution. Moreover, different measures of health outcomes are treated as independent, while it is known that they are correlated with one another over time as they arise in part from a common underlying exposure. We propose modelling an environmental exposure as a latent process, and we describe the implementation of such a model within a hierarchical Bayesian framework and its efficient computation using integrated nested Laplace approximations. Through a simulation study, we compare distinct univariate models for each health outcome with a bivariate approach. The bivariate model outperforms the univariate models in bias and coverage of parameter estimation, in forecast accuracy and in computational efficiency. The methods are illustrated with a case study using healthcare utilization and air pollution data from British Columbia, Canada, 2003-2011, where seasonal wildfires produce high levels of air pollution, significantly impacting population health.",
keywords = "Air pollution, Bayesian hierarchical models, INLA, Latent processes, Time series",
author = "Morrison, {Kathryn T.} and Gavin Shaddick and Henderson, {Sarah B.} and Buckeridge, {David L.}",
year = "2016",
month = "8",
day = "15",
doi = "10.1002/sim.6904",
language = "English",
volume = "35",
pages = "3085--3100",
journal = "Statistics in medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Inc.",
number = "18",

}

TY - JOUR

T1 - A latent process model for forecasting multiple time series in environmental public health surveillance

AU - Morrison, Kathryn T.

AU - Shaddick, Gavin

AU - Henderson, Sarah B.

AU - Buckeridge, David L.

PY - 2016/8/15

Y1 - 2016/8/15

N2 - This paper outlines a latent process model for forecasting multiple health outcomes arising from a common environmental exposure. Traditionally, surveillance models in environmental health do not link health outcome measures, such as morbidity or mortality counts, to measures of exposure, such as air pollution. Moreover, different measures of health outcomes are treated as independent, while it is known that they are correlated with one another over time as they arise in part from a common underlying exposure. We propose modelling an environmental exposure as a latent process, and we describe the implementation of such a model within a hierarchical Bayesian framework and its efficient computation using integrated nested Laplace approximations. Through a simulation study, we compare distinct univariate models for each health outcome with a bivariate approach. The bivariate model outperforms the univariate models in bias and coverage of parameter estimation, in forecast accuracy and in computational efficiency. The methods are illustrated with a case study using healthcare utilization and air pollution data from British Columbia, Canada, 2003-2011, where seasonal wildfires produce high levels of air pollution, significantly impacting population health.

AB - This paper outlines a latent process model for forecasting multiple health outcomes arising from a common environmental exposure. Traditionally, surveillance models in environmental health do not link health outcome measures, such as morbidity or mortality counts, to measures of exposure, such as air pollution. Moreover, different measures of health outcomes are treated as independent, while it is known that they are correlated with one another over time as they arise in part from a common underlying exposure. We propose modelling an environmental exposure as a latent process, and we describe the implementation of such a model within a hierarchical Bayesian framework and its efficient computation using integrated nested Laplace approximations. Through a simulation study, we compare distinct univariate models for each health outcome with a bivariate approach. The bivariate model outperforms the univariate models in bias and coverage of parameter estimation, in forecast accuracy and in computational efficiency. The methods are illustrated with a case study using healthcare utilization and air pollution data from British Columbia, Canada, 2003-2011, where seasonal wildfires produce high levels of air pollution, significantly impacting population health.

KW - Air pollution

KW - Bayesian hierarchical models

KW - INLA

KW - Latent processes

KW - Time series

U2 - 10.1002/sim.6904

DO - 10.1002/sim.6904

M3 - Article

VL - 35

SP - 3085

EP - 3100

JO - Statistics in medicine

JF - Statistics in medicine

SN - 0277-6715

IS - 18

ER -