Estimating the number of injection drug users in greater Victoria, Canada using capture-recapture methods

Population size estimation is critical for planning public health programmes for injection drug users. Estimation is difficult, as these populations are considered 'hidden' or 'hard to reach'. The currently accepted population size estimate for greater Victoria, Canada is between...

Full description

Saved in:
Bibliographic Details
Published in:Harm reduction journal 2014-03, Vol.11 (1), p.9-9, Article 9
Main Authors: Xu, Yuan, Fyfe, Murray, Walker, Liz, Cowen, Laura L E
Format: Article
Language:English
Subjects:
Adult
Behavior
British Columbia - epidemiology
Drug use
Epidemiologic Methods
Epidemiology
Female
Harm reduction
Health services
Humans
Male
Methods
Needle exchange programs
Probability
Public health
Socioeconomic Factors
Source studies
Substance Abuse, Intravenous - epidemiology
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Population size estimation is critical for planning public health programmes for injection drug users. Estimation is difficult, as these populations are considered 'hidden' or 'hard to reach'. The currently accepted population size estimate for greater Victoria, Canada is between 1,500 and 2,000 individuals, which is dated prior to the year 2000, and is likely an underestimate. We used three mark-recapture methods (the Lincoln-Petersen estimator, Huggins' model, and Pledger's model) to estimate population size using cross-sectional survey data collected in 2003 and 2005. Data come from a closed population with two time-ordered samples from the same source. We compare our estimates with the currently accepted estimate that is based on the registry of a Victoria needle exchange. All methods provided population size estimates that were higher than the currently accepted estimate. Huggins' method produced wider confidence intervals. Point estimates of population size from the three methods ranged from 3,329 to 3,342. Our estimates will aid health authorities in planning for harm reduction programmes. Repeating the methods as further phases of I-Track data become available will ensure that the population estimates remain up to date.
ISSN:1477-7517
1477-7517
DOI:10.1186/1477-7517-11-9