Figure 3.

(A and B) The effect of ≥3 months (A) or <3 months (B) of isoniazid preventive therapy on tuberculosis incidence in household contacts 19 years old or younger by the isoniazid resistance status of the index patient adjusted for index case age, recreational drug use, household contact age, sex, bacillus Calmette-Guérin vaccination scar, nutritional status, being a student or not, tuberculosis history, household socioeconomic status, and household residential district. For definition of abbreviations, see Figure 2.