Supervised smoking facility access, harm reduction practices, and substance use changes during the COVID-19 pandemic: a community-engaged cross-sectional study

Table 3 Bivariable multinomial logistic regression of SSF usage since COVID-19, Vancouver, Canada (2020)

Characteristic	Model 1a: SSF usage since COVID-19 relative to no change:
	A: Increased usage		B: Decreased usage
	RRR	95% CI	RRR	95% CI
Demographics
Age (/ age < 30)
30–39	1.55	0.64–3.76	0.75	0.24–2.38
≥ 40	1.78	0.71–4.50	1.13	0.35–3.57
Homeless (/housed)^†	1.18	0.57–2.41	1.31	0.50–3.45
Men (/women + non-binary)	0.62	0.27–1.41	0.49	0.17–1.41
SSF access issues, ever (/no)
Yes	1.27	0.59–2.75	5.36*	1.14–25.2
Behavioural changes since COVID-19
Substance use (/ no change)
Increased	6.94**	2.93–16.44	2.10	0.63–6.96
Decreased	0.73	0.13–4.07	3.50	0.87–14.13
Used alone more often (/ no change)	1.99	0.90–4.40	2.71	0.87–8.40
Mental health since COVID-19
Mental health (/ much worse)
Somewhat worse	0.70	0.25–2.01	0.40	0.10–1.64
About the same	0.51	0.19–1.33	0.50	0.29–2.09
Community connection since COVID-19 (/ more connected)
About the same	1.09	0.33–3.61	0.57	0.14–2.29
Less connected	1.08	0.33–3.49	0.59	0.15–2.28
Model statistic: p < 0.05

Model is estimated with each comparison relative to no change in SSF usage as the reference group
CI confidence interval, RRR relative risk ratio
/ omitted reference category, *p < 0.05, **p < 0.001, Analytic significance level is set to p = 0.05, ^†refers to the month prior to the baseline questionnaire

ISSN: 1477-7517