Background: Clinical guidelines and policies often fail to achieve high levels of delivery of intended clinical interventions. The difference in what investigators know works and what is actually delivered at the clinic-level to patients, is known as the "science-to-service gap." In the realm of tuberculosis (TB) prevention, this gap is reflected in <20% of TB preventive therapy (TPT)-eligible persons living with HIV (PLWH) being offered or initiated on isoniazid preventive therapy (IPT) in many settings. Recent innovation in TPT have brought new pharmacological options allowing for shorter courses, intermittent dosing, or both. A 12-dose once-weekly rifapentine and isoniazid (3HP) regimen has been demonstrated to be effective and well tolerated. This regimen has several potential advantages over IPT; however, if patients are never assessed for 3HP eligibility and 3HP is not prescribed, TPT packets will remain on pharmacy shelves and the potential health benefits will not reach those who need it.

The overarching goal of this study is to identify a generalizable approach to overcome current barriers to delivery of TPT in order to achieve high levels of TPT delivery during routine care in public clinics. Investigators are proposing a choice architecture that makes prescribing TPT the "default" or standard option and that for TPT not to be prescribed will require a choice by a clinician to "opt-out" of TPT for a specific patient.

Methods: Investigators will use a cluster randomized design with the larger IMPAACT4TB (I4TB) program to deliver 3HP to countries in Africa, Asia, and Latin America. A subset of countries and clinics within these I4TB countries will be included with each clinic the unit of randomization. Clinics within study countries will be randomized to one of two strategies: (1) standard implementation within the UNITAID project (clinic training on TPT along with posters and other standard medication material) and (2) choice architecture default TPT.

Clinical process data will be used to assess the effectiveness of each strategy to determine the proportion of PLWH (1) screened for TB preventive therapy, (2) eligible for TPT, and (3) prescribed TPT.

Significance: Identifying a pragmatic approach will lead the way for improving TPT prescribing across the study sites. It will furthermore contribute to implementation science at large in describing implementation strategies that may be applied to clinic-level implementation of other innovations.

BACKGROUND

Clinical guidelines and policies often fail to achieve high levels of delivery of the intended clinical intervention. The difference in what investigators know works and what is actually delivered at the clinic-level to patients, is known as the "science to service gap." In the realm of TB prevention, this "science to service" gap is reflected in <20% of isoniazid eligible PLWH actually receiving isoniazid in most settings. 3HP has several potential advantages over isoniazid for TB prevention; however, if patients are never assessed for 3HP eligibility and 3HP is not prescribed, packets of 3HP will remain on pharmacy shelves and the potential health benefits will not reach those who need it.

Failure of delivery of an innovation can occur for multiple reasons and at multiple levels: this includes an outer setting of national policy and funding and inner setting of providers embracing the innovation, feeling confident in prescribing it, and remembering to prescribe it. The failure with TB preventive therapy (TPT) delivery appears to be mostly due a failure of clinician having a realistic understanding in the benefits compared to risks of IPT, confidence in appropriate prescribing, and remembering to prescribe at the appropriate time.

Multiple approaches have been suggested and/or implemented to improve clinic-level delivery of recommended services. These include performance feedback, performance-based incentives, change agents, opt-out services, routinizing the service in normal care delivery, quality improvement officers or projects, clinic champions, electronic medical record "popups", etc. Not all of these approaches have been evaluated and of those that have been evaluated, success is uneven. Part of the reason for variable success is that some of these context specific approaches have failed in settings with variation in the context of the implementation and variation in the barriers to implementation.

Investigators are proposing to test an implementation strategy to achieve high levels of clinic-level delivery of TB preventive therapy. Investigators will conduct a cluster randomized trial of a clinician "optout" strategy, shifting clinician decision making from prescribing TPT to identifying individuals to whom not to prescribe TPT, and the standard implementation strategy for 3HP roll-out in the IMPAACT4TB project: training and visual aids (posters, etc.) in the clinic.

Training for clinic staff will be general on 3HP implementation and routine data recording as part of the IMPAACT4TB implementation. A separate training will occur for clinicians working in intervention clinics to minimize adoption of the routinization approach by control clinics.

The choice architecture "default" or "opt-out" strategy is based on behavioral economic theory. The key behavioral economic concepts are choice architecture for which the opt-out approach makes delivery of TPT the default option - rather than the choice option. This approach has been effective in multiple clinic implementation areas for other services. In addition to shifting decision making by making TPT the default option it also normalizes TPT to be appropriate for all PLWH, rather than those with perceived exposure, risk, etc. Finally, by making TPT the default it prevents providers from overlooking TPT at initial encounter and after ruling out TB disease if TPT was delayed due to concern for TB disease. Investigators further hypothesize, and propose to assess, that through the process of normalizing or routinizing TPT to all (versus selecting for eligibility) the self-efficacy of providers regarding TPT prescribing will increase.

STUDY OBJECTIVES

Primary Objective: To use a feasible, low-resource strategy to achieve levels of TPT delivery substantially higher than routine implementation in primary care public clinics.

Primary Outcome: proportion of TPT eligible/potentially eligible PLWH initiating ART at clinics who are prescribed TPT

Secondary Outcomes:

• Proportion of TPT eligible/potentially eligible PLWH established on ART who are prescribed TPT
• Proportion of all unique HIV patients during study period prescribed TPT (3HP and IPT)
• Adverse events among TPT recipients and among non-TPT recipients
• TPT completion among those who initiate TPT
• TB diagnoses among TPT recipients
• Acceptability and fidelity of choice architecture approach to clinicians

METHODS

Study Design:

Investigators have designed a cluster randomized trial of a clinic-level implementation strategy for a standard of care intervention (TPT to PLWH). Clinics within study countries will be randomized to one of two strategies: (1) standard implementation within the I4TB project (clinic training on TPT along with posters and other standard medication material) and (2) "choice architecture" / "opt-out" TPT prescribing. Because of the clinic-level nature of the implementation strategies, all PLWH receiving care at a clinic will be exposed to the standard implementation or TPT choice architecture implementation. Clinical process data will be used to assess the effectiveness of each strategy to determine the proportion of PLWH prescribed TPT.

NOTE: Implementation of 3HP will only occur in any of the study countries after rifapentine is registered for in-country use for TPT. No off-label or un-registered drug use will occur as part of this study. This study is being conducted within the IMPAACT4TB project focused on implementing 3HP and the study will be implemented alongside 3HP rollout in these countries.

Study Setting:

Malawi, Mozambique, and Zimbabwe

Site selection:

Outpatient clinics will be selected for inclusion based on the following criteria:

1. The clinic is supported by the in-country collaborating partner (through PEPFAR or other international funding)
2. The clinic is prescribing both ART and TPT
3. Clinic has been selected for 3HP introduction
4. Clinics will be included starting sequentially with the lowest TPT performing clinics. Ideally, no clinics with >50% TPT prescribing will be included.
5. Preferred characteristics: Another potential study clinic within 15 km; use of similar of prescribing approach to multiple other clinics.

Each partner supports 20-30 clinics as part of the I4TB project. Of these approximately 22 will be selected for inclusion in each country. No additional considerations, beyond what is listed above, will be used for clinic selection.

Randomization:

Clinic randomization to study arm will be conducted publically (either in a single location, or if blocking by country, in each country). Randomization will be achieved either through computer randomization or through drawing marked balls from a bag. Randomization will be stratified by country and within each country by clinic ART population and urban / rural location.

Implementation Strategy initiation:

Following randomization, optimization and adaptation of the implementation strategy for that clinic will be discussed between clinic staff, 3HP country implementers, and research staff. Core elements of the novel strategy (choice architecture default to 3HP prescribing) will be maintained while allowing adaptation of specific algorithm tools, posters, clinical mentoring approaches, and pre-printed, or ink-stamp prescriptions in patient clinic files (prescriptions in most countries are written in the clinic-maintained patient paper file rather than a slip of paper or electronic prescription system) or EMR modifications for electronic prescribing.

Choice Architecture Strategy delivery:

Clinic staff will be responsible for strategy delivery for all patient interactions. Research staff will provide training and guidance for the "choice architecture" arm. Research staff will also work with the clinics to develop appropriate clinical stationary, ink stamps, stickers, or EMR modifications for prescribing, and reminder systems (e.g. written by pharmacy in clinic file, post-it on clinic file, post-it on lab results). The goal of this approach is for TPT prescribing to occur routinely and as part of ART prescribing. This is in contrast to considering prescribing only at the end of a long algorithm that includes TB and other assessments. With this approach, a patient will "automatically" be prescribed TPT unless the clinician specifically decides participants are not a candidate due to active TB treatment or other clinical reasons.

In-country research coordinators will travel to clinics to monitor strategy implementation and provide support as appropriate within the strategy procedures.

Steps for "opt-out" strategy delivery:

1. Providers will receive general training on TPT benefits, indications, and contra-indications.
2. Providers will be provided with updated ART and TPT prescribing approach. This will be adapted to the country context and may include a pre-printed prescription pad (Mozambique), adaptations in clinical stationary (Zimbabwe), or adaptations to an electronic prescribing system (Malawi).
3. The pharmacy or clinician (if the clinician dispenses) will dispense ART, cotrimoxazole, and TPT as prescribed

Standard care delivery:

Providers will receive training on benefits, indications, and contra-indications for TPT. Providers will use the standard approach of the "default" being to not prescribe. Only if providers specifically write for TPT will it be dispensed by a pharmacy or the provider. In-country research coordinators will travel to clinics to monitor strategy implementation and provide support as appropriate within the strategy procedures.

Inclusion criteria:

• As this is a health system strategy with the intervention at the clinic level, all adult PWH receiving care at that clinic will be exposed to the strategy and included in the overall proportion of PWH seen at the clinic during the study period who receive 3HP

Exclusion criteria:

• None

De-identified aggregate data will be made available one-year after completion of all study activities.

One year after completion of study activities

Contact PI

The data identified by clinic allocation will be blinded to the PI and co-investigators and study statistician until completion of comparative analysis. There will be no other blinding.