Estimating Lymphatic Filariasis Morbidity in Haiti Using Respondent-Driven Sampling: A Simulation Study Public

Abstract

 

Accurately estimating the size and composition of hidden populations is important and needed in public health for a number of reasons. In this thesis, we focus on one such hidden population: those suffering morbidity from Lymphatic Filariasis (LF). Over time, LF can lead to lymphedema (fluid collection and severe swelling of extremities) and/or hydrocele (severe swelling of the scrotum in males). These can lead to reduced mobility, financial hardships, and social isolation for affected individuals, which causes them to become hidden. Respondent Driven Sampling (RDS) is a method developed for assessing the population size of hidden populations where the sample accrues by referrals based on an assumption of social network ties between affected individuals and/or their families or support groups. Using RDS to estimate the numbers of individuals experiencing lymphedema and/or hydrocele in conjunction with the Successive Sampling – Population Size Estimation (SS-PSE) method has potential to produce an accurate estimate for the morbidity of LF with associated levels of uncertainty. Having an accurate estimate will not only aid in ongoing efforts regarding the surveillance of LF in Haiti and, potentially, aid in elimination efforts for neglected tropical diseases (NTDs) globally, but could also be implemented more widely across many public health areas with hidden populations to provide accurate estimates requiring less cost, less time, and fewer resources in general.  To explore whether RDS will work to accurately estimate our population of interest, lymphedema and hydrocele, we simulate the population of interest and the diffusion of RDS on that population with varying levels of connectivity and true population sizes. SS-PSE, a Bayesian approach, allows different prior information and prior precision to be incorporated into our estimates. We show that the method captures the true simulated population size in the posterior probability intervals through varying levels of connectivity, true population sizes with sensitivity to the choice of prior.  The results highlight the importance of initial seed choice in RDS and outline several areas for continued research. However, caution should be used when interpreting results since the simulation carried many assumptions.

Table of Contents

Table of Contents

INTRODUCTION............................................................................................................. 1

METHODS.......................................................................................................................... 6

Respondent-Driven Sampling........................................................................................... 6

Sample Size and Seed Selection........................................................................................ 8

Simulation........................................................................................................................... 10

Networks and graphs........................................................................................................ 10

Background parameters for lymphedema and hydrocele in Haiti.............................. 11

SS-PSE................................................................................................................................. 13

ANALYSIS.......................................................................................................................... 16

DISCUSSION..................................................................................................................... 23

Conclusions......................................................................................................................... 23

Strengths and limitations................................................................................................... 26

Future work......................................................................................................................... 27

REFERENCES................................................................................................................... 29

APPENDIX......................................................................................................................... 31

 

 

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 Health Sciences, Public Health
Mot-clé	Hidden Populations Lymphatic Filariasis Successive Sampling Respondent-Driven Sampling Population Size Estimation Haiti
Committee Chair / Thesis Advisor	Waller, Lance A., Emory University
Committee Members	Lyles, Robert, Emory University
Partnering Agencies	Does not apply (no collaborating organization)

Dernière modification

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	Estimating Lymphatic Filariasis Morbidity in Haiti Using Respondent-Driven Sampling: A Simulation Study ()	2018-04-09 12:16:37 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions