Participants randomized to Arm 5 will i) initiate ART on the day of enrollment [experimental], ii) receive ongoing care in a PWID-focused community-based site [experimental], and iii) receive routine adherence support should they experience treatment failure at 6 months [usual care].

Participants randomized to Arm 6 will i) initiate ART on the day of enrollment [experimental], ii) receive ongoing care in a government HIV clinic [usual care], and iii) receive enhanced adherence support should they experience treatment failure at 6 months [experimental].

Participants randomized to Arm 7 will i) initiate ART on a standard timeline [usual care], ii) receive ongoing care in a PWID-focused community-based site [experimental], and iii) receive enhanced adherence support should they experience treatment failure at 6 months [experimental].

Participants randomized to Arm 8 will i) initiate ART on the day of enrollment [experimental], ii) receive ongoing care in a PWID-focused community-based site [experimental], and iii) receive enhanced adherence support should they experience treatment failure at 6 months [experimental].

Participants randomized to standard ART initiation will not initiate ART on the day of trial enrollment, but only after linking to their assigned source of HIV care (either community-based HIV care or government-based HIV care, depending on randomization). In standard ART initiation, patients typically, complete an intake visit at the HIV clinic, with baseline laboratory testing, and return to the clinic approximately 2 weeks later to begin ART.

Participants randomized to community-based HIV care will be referred to PWID-focused integrated care centers (ICCs) for ongoing HIV clinical management - a prototype of decentralized HIV care in India. ICCs will provide free HIV care that will adhere closely to Indian HIV treatment guidelines and, when relevant, to local HIV treatment standards.

Participants randomized to government-based HIV care will be referred to government-based HIV clinics for ongoing HIV clinical management. Government-based clinics provide free HIV care that adheres closely to Indian HIV treatment guidelines

Participants who i) experience virologic failure after 6 months in the study and ii) are randomized to enhanced adherence support will receive an intensive, tailored adherence intervention lasting a maximum of 6 months, with two components: 1) tracking and outreach, and 2) psychosocial support and navigation. These will aim to equip PWID with skills to independently manage their ART using motivational interviewing and strengths-based case management.

Participants who i) experience virologic failure after 6 months in the study and ii) are randomized to routine adherence support will receive a guideline-based, HIV clinic-based adherence counselling intervention lasting a maximum of 6 months

HIV RNA levels in blood measured with GeneXpert 2 module, Xpert HIV-1 Viral Load XC Cartridge (Cepheid AB, Sweden). Lower limit of quantification 40 copies/mL

Research staff and outreach workers collected reports on participant deaths. Verified deaths required one of the following: i) hospital records or death certificate, ii) confirmation of death from a family member, or iii) confirmation of death from an eyewitness (usually another PWID). Rate will be presented as events per unit observation time.

Linkage to ART will be defined as collecting one or more ART prescriptions from a clinic by 3 months and 6 months, captured by medical record abstraction or, in the absence of medical record data, participant self-report of ART collection from a clinic.

Participants will be classified as adherent if they report taking ART in the prior 30 days and report adherence of 80% or higher using a visual analog scale (range: 0% to 100%), with higher numbers indicating higher adherence.

ART fill data will be abstracted from medical records. In phase 1, participants will be classified as adherent if they have at least one ART refill by 3 months (91 days) and have an MPR of 80% or higher in the period between the first ART fill in the clinic and 6 months (182 days).

In phase 2, participants will be classified as adherent if they have at least one ART refill by 3 months (91 days) following the second randomization (or prior to the second randomization) and have an MPR of 80% or higher in the period between the first ART refill (or the second randomization if the first ART refill was prior to the second randomization) and 6 months (182 days following the second randomization).

Investigators will measure both HIV-related and drug use-related stigma as separate constructs using a survey. Each construct is evaluated by three sub-scales: 1) anticipated healthcare stigma, 2) enacted healthcare stigma, and 3) internalized stigma. All stigma subscales will be queried using Likert scales that range from 0 to 3, with higher scores indicating more stigma. For anticipated and enacted stigma, participants will be classified as experiencing stigma if the Likert scale is >0 for any one item (i.e., analyzed as a dichotomous outcome). For the internalized stigma subscale, scores will be converted to a continuous mean outcome from 0 to 3.0, with higher scores indicating worse stigma.

Depression will be measured with the PHQ-9 questionnaire. The scoring range is 0 to 27, with higher values indicating more depression symptoms. Participants will be categorized as having at least moderate depression symptoms if the score is ≥ 10.

HIV treatment self-efficacy will be measured with a modified HIV Treatment Adherence Self-Efficacy Scale. Items will be averaged to calculate a self-efficacy score, ranging from 0 to 100, with higher scores indicating better self-efficacy.

New antiretroviral drug resistance will be defined as the detection of one or more reverse transcriptase or integrase gene DRMs at follow-up that was not present at baseline, among participants with HIV RNA ≥1000 c/mL at follow-up visit. DRMs will be interpreted using the online Stanford HIV Database