Selection effect of needle exchange in Anchorage, Alaska

Area of Science:

• Public Health
• Epidemiology
• Substance Abuse Research

Background:

• Participation bias, or selection bias, can compromise studies evaluating needle-exchange programs (NEPs).
• Understanding this bias is crucial for accurately assessing the impact of NEPs on public health outcomes.

Purpose of the Study:

• To investigate the presence and extent of selection bias among injection drug users (IDUs) participating in a randomized clinical trial of a needle-exchange program.
• To identify factors associated with the actual utilization of NEPs among randomly assigned participants.

Main Methods:

• A two-arm randomized clinical trial comparing an experimental NEP condition with a pharmacy sales condition.
• Analysis of follow-up time and logistic regression to predict NEP usage within the NEP arm.
• Inclusion of IDUs randomly assigned to the NEP condition, comparing users versus non-users.

Main Results:

• No statistically significant difference in time to follow-up between the NEP and pharmacy sales conditions.
• A highly significant difference in time to follow-up between IDUs who used the NEP and those who did not within the NEP condition.
• Higher levels of drug use (heroin and amphetamine injection, sharing equipment) were significant predictors of NEP utilization.

Conclusions:

• Selection bias is demonstrably present among injection drug users who utilize needle-exchange programs.
• The findings suggest that individuals with higher-risk behaviors are more likely to engage with NEPs, potentially confounding effectiveness studies.
• Further research is needed to develop methodologies that account for and mitigate selection bias in NEP research.