Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Spatiotemporal modeling of opioid abuse and dependence outcomes using Bayesian hierarchical methods

Sumetsky, Natalie (2017) Spatiotemporal modeling of opioid abuse and dependence outcomes using Bayesian hierarchical methods. Master's Thesis, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF
Submitted Version

Download (623kB) | Preview

Abstract

Opioid addiction is a major public health concern that presents a significant disease burden. In the past decade, drug overdose rates have soared. More research is necessary to inform policy and to ensure provision of proper care to individuals and communities in need. This thesis explores spatiotemporal models to assess ecological and demographic factors associated with opioid addiction risk on a ZIP-code level in Pennsylvania.
Bayesian hierarchical models are commonly used to explore complex spatiotemporal disease trends. Markov chain Monte Carlo (MCMC) simulations are a valuable albeit computationally costly tool in fitting models of this class. A newer method, integrated nested Laplace approximation (INLA), offers improved computational efficiency with comparable results for models with latent Gaussian fields. For example, a 2014 cross-sectional model discussed in this thesis took 5581 seconds to run using MCMC simulations, while INLA offered comparable results in seven seconds. Cross-sectional and longitudinal misalignment models with opioid abuse and dependence outcomes are compared using both methods.
Higher outcome risk is associated with areas with greater proportions of 45- to 64-year-olds, higher density, more retail clutter and manual labor establishments per square mile, higher unemployment, lower median income, and greater proportion of residents below the 150% poverty line. As regional needs differ, identifying high-risk community-level factors and locations carries great public health significance. Interventions and preventive efforts could then be tailored specifically to areas where the disease burden is greatest.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Sumetsky, Natalienms77@pitt.edunms770000-0001-9800-8753
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Thesis AdvisorAnderson, Stewartsja@pitt.edusja
Committee MemberMair, Christinacmair@pitt.educmair
Committee MemberBuchanich, Jeaninejeanine@pitt.edujeanine
Committee MemberChang, Joycechangj@pitt.educhangj
Date: 25 September 2017
Date Type: Publication
Defense Date: 27 July 2017
Approval Date: 25 September 2017
Submission Date: 23 July 2017
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 47
Institution: University of Pittsburgh
Schools and Programs: Graduate School of Public Health > Biostatistics
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: Bayesian hierarchical models, spatiotemporal models, conditional autoregressive models, Markov chain Monte Carlo simulation, integrated nested Laplace approximation, opioid abuse, opioid dependence
Date Deposited: 25 Sep 2017 14:23
Last Modified: 25 Sep 2017 14:23
URI: http://d-scholarship.pitt.edu/id/eprint/33054

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item