Evaluation of respondent-driven sampling in seven studies of people who use drugs from rural populations: findings from the Rural Opioid Initiative

Table 5 Unweighted and RDS-weighted Relative Risk and Odds Ratio measures of association for the relationship between (a) fentanyl use, (b) heroin use, and (c) age and positive Hepatitis C Virus antibody status

	(a) Fentanyl use^a and HCV antibody status		(b) Heroin use^a and HCV antibody status		(c) Age (per 10 years)^b and HCV antibody status
*Relative Risk Regression*¹	RR	95% CI	RR	95% CI	RR	95% CI
Unweighted	1.35	1.18, 1.54	1.44	1.28, 1.62	1.02	0.96, 1.09
RDS-I: Homophily by age	1.36	1.19, 1.55	1.44	1.30, 1.60	1.03	0.98, 1.08
RDS-I: Homophily by drug of choice	1.35	1.17, 1.56	1.44	1.28, 1.62	1.02	0.96, 1.09
RDS-II: Individual network size/degree	1.34	1.06, 1.70	1.31	1.06, 1.61	1.06	0.97, 1.15
*Logistic Regression*²	OR	95% CI	OR	95% CI	OR	95% CI
Unweighted	2.07	1.56, 2.74	2.13	1.77, 2.57	1.05	0.91, 1.20
RDS-I: Homophily by age	2.07	1.57, 2.72	2.12	1.75, 2.57	1.08	0.98, 1.20
RDS-I: Homophily by drug of choice	2.02	1.48, 2.76	2.10	1.74, 2.54	1.04	0.89, 1.21
RDS-II: Individual network size/degree	1.93	1.23, 3.02	1.87	1.37, 2.55	1.12	0.94, 1.33

Abbreviations: CI Confidence interval, HCV Hepatitis C virus, OR Odds ratio, RR Relative risk
¹Relative risks estimated separately in each study using modified Poisson regression and combined across studies using random-effects meta-analyses
²Odds ratios estimated separately in each study using logistic regression and combined across studies using random-effects meta-analyses
^aReference period: past 30 days
^bAge modeled as a continuous variable

ISSN: 1471-2288