Premium
Reduced bias for respondent‐driven sampling: accounting for non‐uniform edge sampling probabilities in people who inject drugs in Mauritius
Author(s) -
Ott Miles Q.,
Gile Krista J.,
Harrison Matthew T.,
Johnston Lisa G.,
Hogan Joseph W.
Publication year - 2019
Publication title -
journal of the royal statistical society: series c (applied statistics)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.205
H-Index - 72
eISSN - 1467-9876
pISSN - 0035-9254
DOI - 10.1111/rssc.12353
Subject(s) - respondent , estimator , sampling (signal processing) , sample (material) , statistics , econometrics , human immunodeficiency virus (hiv) , population , sampling bias , nonprobability sampling , transmission (telecommunications) , sample size determination , medicine , environmental health , computer science , mathematics , virology , telecommunications , chemistry , chromatography , political science , detector , law
Summary People who inject drugs are an important population to study to reduce transmission of blood‐borne illnesses including human immunodeficiency virus and hepatitis. We estimate the human immunodeficiency virus and hepatitis C prevalence among people who inject drugs in Mauritius. Respondent‐driven sampling (RDS), which is a widely adopted link tracing sampling design used to collect samples from hard‐to‐reach human populations, was used to collect this sample. The random‐walk approximation underlying many common RDS estimators assumes that each social relationship (edge) in the underlying social network has an equal probability of being traced in the collection of the sample. This assumption does not hold in practice. We show that certain RDS estimators are sensitive to the violation of this assumption. To address this limitation in current methodology, and the effect that it may have on prevalence estimates, we present a new method for improving RDS prevalence estimators using estimated edge inclusion probabilities, and we apply this to data from Mauritius.