Investigating Missing Data Mechanisms in Substance Abuse Research: Analyzing the Missing-Not-At-Random (MNAR) Pattern of ‘Return to Use’ Data Open Access

Abstract

Missing data is a large challenge of substance use disorder (SUD), and addressing such challenges is critical to ensure the validity and reliability of data analysis and study outcome results. The study looks at Missing-Not-At-Random (MNAR) Patterns of ‘Return to Use’ Data from a from a longitudinal cohort study conducted by the Hazelden Betty Ford Foundation and Emory University, where ‘Return to Use’ has a high level of missingness and all other variables exceed 94 percent completeness. Analysis of survey respondents found responses of age, treatment completion status, appointment attendance among other variables that significantly differed between missing and non-missing Return to Use individuals, and regression analysis saw programming type and discharge approval as constant predictors for Return to Use missingness. Latent class analysis determined that there were not discrete subgroups within the dataset to define different domains of sustainable recovery. This paper contributes to understanding of ‘Return to Use’ missingness within SUD research and develops pathways to continue to evaluate reasoning and approaches to missing data.

Table of Contents

Table of Contents

Introduction.

Types of Missing Data

Strategies for Missing Data

Considerations for Substance Use Disorder Research

Data Description

Methods.

Study Population

Analyzing Missingness of Return to Use

Statistical Analysis

Results.

Summary of Missingness in Dataset

Characteristics of Missingness - Comparative Analysis

Baseline Modeling with Complete Case

Modeling with MICE Imputation

Latent Class Analysis

Conclusions.

Discussion

Bibliography.

About this Master's Thesis

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Rollins School of Public Health
Department	Biostatistics
Subfield / Discipline	Biostatistics - MPH & MSPH
Degree	M.S.P.H.
Submission	Master's Thesis
Language	English
Research Field	Biology, Biostatistics Psychology, Behavioral Sciences
Keyword	Substance Use Disorder Missing Data
Committee Chair / Thesis Advisor	Lance Waller, Emory University
Committee Members	Siara Sitar, Emory University

Last modified

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Investigating Missing Data Mechanisms in Substance Abuse Research: Analyzing the Missing-Not-At-Random (MNAR) Pattern of ‘Return to Use’ Data ()	2024-04-08 10:09:22 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions