Evaluation of Efficiency in Pooling Strategies for Some Mixed Linear and Nonlinear Models in Longitudinal Studies Pubblico

Abstract

For repeated measures or longitudinal studies, the cost and the effort of applying expensive laboratory assays to assess biomarker levels can be prohibitive. To mitigate assay costs, it is relatively common to pool samples prior to performing a laboratory test. In certain settings, an optimal pooling design has been shown to minimize information loss. Pooling laboratory samples can also preserve biospecimens and avoid issues with limits of detection. In this thesis, we explore the efficiency of pooling strategies for estimating fixed effects and variance components under mixed linear models for normally distributed outcomes as well as a mixed nonlinear model assuming a gamma-distributed outcome. We conducted a series of simulations to assess and compare the pooling strategies and models discussed in section 2. We evaluate the efficiencies of different pooling design strategies with initial simulation studies under standard Gaussian assumption-based linear mixed models. We also examine the efficiency of within-individual pooling a right-skewed outcome under the gamma model. The design strategies are presented with simulations inspired by The HIV Epidemiology Research Study, with assumed parameters mimicking a prior longitudinal study that estimated average trajectories of HIV ribonucleic acid (RNA) over time.

Table of Contents

1.  Introduction .  .  .  .  .  .  .  .  .  .  .  .  .  .  .   .  .  .  .  .  .  .  .1

2.  Linear Mixed Effect Model .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .3

3.  Gamma Model .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .5

4.  Pooling Strategies .  .  .  .  .  .  .  .  .  .  .  .  .   .  .  .  .  .  .  .6

Pooling within Individuals, Type I   .  .  .  .  .  .  .  .  .  .  .7

Pooling across Individuals, Type II   .  .  .  .   .  .  .  .  .  .8

Mixed Pooling, Type III  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .9

5.  Simulation Study .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .10

 

6.  Discussion .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  . .  .  19

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
Degree	M.P.H.
Submission	Master's Thesis
Language	English
Research Field	Statistics
Parola chiave	Pooling Gamma Model Mixed Linear Model
Committee Chair / Thesis Advisor	Lyles, Robert, Emory University
Committee Members	Manatunga, Amita, Emory University
Partnering Agencies	Does not apply (no collaborating organization)

Ultima modifica

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Evaluation of Efficiency in Pooling Strategies for Some Mixed Linear and Nonlinear Models in Longitudinal Studies ()	2018-04-09 12:11:14 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions