Regimen factor is also part of certain interaction terms. Blank cells indicate not applicable or not tested. Use of a less stringent entry criterion (P ≤ .2) produced the same final model.

Abbreviations: 3HP-DOT, 3 months of directly observed once-weekly rifapentine (maximum dose, 900 mg) plus isoniazid (maximum dose, 900 mg); 9H-SAT, 9 months of daily self-administered isoniazid (maximum dose, 300 mg); CI, confidence interval; HIV, human immunodeficiency virus; LTBI, latent tuberculosis infection; NCT, noncompletion of treatment; OR, odds ratio; Ref., reference.

^a Includes North American Indian and other participants in the United States and Canada.

^b History of living in a correctional institution for ≥1 month before enrollment.

^c History of homelessness or living in a shelter or single-room occupancy for ≥6 months before enrollment.

^d Use: affirmative response to a question asking whether the participant ever drank alcoholic beverages. Abuse: score of ≥2 on the Cut down, Annoyed, Guilty, and Eye-opener questionnaire [9].

^e Smoking was ascertained on the basis of the patient's statement of whether smoking was current at time of enrollment.

^f Enrollment site was analyzed as the random effect variable of a generalized linear mixed model. Two-tailed P value was calculated on the basis of the z score (estimate/standard error).

^g Missing ≥1 of the first 3 directly observed therapy (DOT) sessions for the 3HP regimen or ≥1 of the 3 monthly clinic visits for the 9H regimen, followed by a DOT or a monthly visit, respectively. The variable includes those who did not receive any study dose (n = 148). In the univariate analysis, the need for an interpreter among non-US- or non-Canadian-born participants (n = 3133 [use = 1355; no use = 1778]) resulted in statistical significance (OR, 0.55; 95% CI, .40–.77; P = .001) among participants who did not complete treatment attributed to an adverse event.