Table 3.

Associations between the outcome of treatment non-completion and the variables treatment support, origin, age and sex (n = 726)

Covariates	Univariable			Multivariable
	n	c OR	p	a OR	SE	p	95% CI
Origin
Foreign-born	595	1 (ref)		1(ref)
Norwegian-born	131	2.1	0.016	0.8	0.342	0.602	0.34–1.84
Age group
< 35	561	1 (ref)		1 (ref)
> 35 years	165	1.9	0.026	1.7	0.521	0.093	0.92–3.09
Sex
Female	299	1 (ref)		1 (ref)
Male	427	0.5	0.017	0.6	0.093	0.106	0.36–1.10
Treatment support. Model with interaction term for origin: LRT p-value of interaction term = 0.049
Effect of treatment support on treatment completion in foreign-born (multivariate wald p-value 0.017)
Self-administereda	174	1 (ref)		1 (ref)
DOT dailyb	151	0.2	0.005	0.3	0.150	0.017	0.10–0.80
DOT weeklyc	261	0.4	0.009	0.4	0.162	0.027	0.22–0.91
Effect of treatment support on treatment completion in Norwegian-born (multivariate wald p-value 0.468)
Self-administereda	80	1 (ref)		1 (ref)
DOT dailyb	28	1.7	0.374	2.1	1.332	0.229	0.62–7.26
DOT weeklyc	19	0.9	0.928	1.0	0.868	0.961	0.20–5.34

OR Odds ratio, SE standard error, DOT Direct Observed Treatment for part of or the full treatment period

^aSelf-administered include those who managed their treatment themselves or were given weekly pill boxes

^bdaily DOT: those who were administered daily treatment under direct observation

^cweekly DOT: those who were administered weekly rifapentine and isoniazid under direct observation