Community Initiated Preventive Therapy for TB (CHIP-TB)

Background: The World Health Organization (WHO) currently recommends household contact investigation for new tuberculosis (TB) patients in low- and middle-income countries, with an emphasis on pediatric contacts. Although the aim of this policy is to find previously undetected TB patients and reduce transmission, such investigations represent a missed opportunity to start contacts without TB on preventive therapy (TPT). The WHO guidelines do not address the optimal implementation of contact investigation. The standard of care (SOC) in most settings, passive referral of pediatric contacts to the clinic by the index TB patient, has largely remained unsuccessful in practice. In 2017, the WHO estimated only 23% of eligible child contacts were started on TB preventive therapy. Household contact investigation has been shown to have a higher yield in active TB case finding, but is labor intensive, and may be challenging to implement in densely populated urban settings or informal settlements/slums. The WHO recently endorsed the use of a new TPT regimen (rifapentine and isoniazid weekly (3HP)) for both children and adults in high burden settings, and the programmatic roll out of this regimen offers the opportunity to simultaneously examine new strategies to improve the identification and treatment of pediatric TB household contacts. Objective: To compare the effectiveness of community-based versus facility-based child contact investigation and delivery of TB preventive care to inform the optimal implementation strategy for investigating pediatric household TB contacts. Study Design: Cluster-randomized trial in 32 clinics (16 clinics per arm) divided equally among South Africa and Ethiopia (8 clinics per arm per country). Methods: Newly diagnosed qualifying TB index patients (determined by South African or Ethiopian National TB guidelines) and participants' household child contacts will be recruited to participate. Local staff, including the relevant nurses and community health workers in the intervention and control clinics, will be trained to conduct contact investigation with a symptom-based approach for all child contacts under 15 years old in home and at the clinic. Data will be collected using routine medical files and then retrospectively abstracted by the research team. Thirty-two primary health clinics will be stratified (by TB case notification and by country) and randomized in 1:1 fashion to either community-based or facility-based delivery of care. Household child contacts under 15 years of age who screen negative for TB disease will be initiated on TPT by a healthcare worker (nurse, community health worker, etc.) either in the home or clinic setting. Children in the intervention arm who screen positive will first be sent to the nurse at the clinic for repeat screening. Children who have a persistently positive screen in the intervention arm and those with a positive screen in the control arm using South Africa's or Ethiopia's pediatric symptom screening tool will be referred to a physician at the district hospital for further investigation of TB disease, as is the standard of care in both settings. Investigators will compare clinic-level outcomes including proportion of household TB contacts under 15 years of age that were screened, initiated on TPT, and who completed TPT, and reasons for not completing TPT including loss to follow up and incident TB disease while on TPT.

All TB index patients who agree to participate will have a home visit by clinic staff who will perform: (1) contact enumeration (2) TB symptom screening of all children <15 years (3) Initiation of TPT for all asymptomatic children and (4) Referral of all symptomatic children less than 15 years, including those living with HIV. HIV testing will be offered to all child contacts 12 months of age and older. Those children less than 12 months will be referred to the clinic for HIV testing, if indicated by local guidelines. In South Africa, these home visits will occur by a combination of community health workers and professional nurses. In Ethiopia, home visits will occur by health extension workers supported by nurses.

Children less than 15 years living in the home of TB index patients who agree to participate in the study will be referred to clinic for TB symptom screening and initiation of TPT for all asymptomatic child contacts. Symptomatic child contacts will be referred to a physician for evaluation, as is currently the standard of care. Additionally, child contacts identified in any maternal and child health program will be referred to the TB clinic for TB symptom screening. HIV testing will be offered at the clinic for all child contacts and will be performed according to local guideline.

Community-based care delivery model for pediatric tuberculosis evaluation and initiation of preventive therapy

The cluster-level ratio of the number of household child contacts less than 15 years of age initiated on TB preventive therapy (TPT) per index patient, comparing the intervention to the control arm

TPT includes 3 months of weekly rifapentine and isoniazid, 3 months of daily rifampin and isoniazid or 6 months of isoniazid

The cluster-level ratio of the number of household child contacts less than 15 years of age identified per index patient, comparing the intervention to the control arm

We will calculate the estimated child contacts per household using Demographic and Health Survey data and prior community-based contact tracing studies in both South Africa and Ethiopia. We will measure identified child TB contacts using the index patient's clinic-based contact tracing report, the child contact management file and the community-level file maintained by community health teams.

We will calculate the estimated child contacts per household using Demographic and Health Survey data and prior community-based contact tracing studies in both South Africa and Ethiopia. We will define screening as the child's documented visit to the TB clinic or a documented visit to the home by the community health team (where the child was present) within four months of the index patient's TB clinic admission.

The cluster-level proportions of estimated child contacts under 15 years who are initiated on TB preventive therapy

We will calculate the estimated child contacts per household using Demographic and Health Survey data and prior community-based contact tracing studies in both South Africa and Ethiopia. Clinic staff or community health workers will document TPT initiation on the child contact management file.

The cluster-level proportions of estimated child contacts under 15 years who complete TB preventive therapy

We will calculate the estimated child contacts per household using Demographic and Health Survey data and prior community-based contact tracing studies in both South Africa and Ethiopia. Clinic staff or community health workers will document TPT completion on the child contact management file.

The cluster-level ratio of the number of household child contacts less than 15 years of age initiated on TB preventive therapy per index patient, comparing the intervention to the control arm in South Africa

The cluster-level ratio of the number of household child contacts less than 15 years of age initiated on TB preventive therapy per index patient, comparing the intervention to the control arm in Ethiopia

The cluster-level ratio of the number of household child contacts less than 15 years of age initiated on TB preventive therapy per index patient, comparing the intervention to the control arm among those less than 5 years of age

The proportion of newly diagnosed HIV-positive children, by study arm, who are referred for initiation of antiretroviral therapy by the community health team

The proportion of newly diagnosed HIV-positive children, by study arm, who are initiated on antiretroviral therapy

The proportion of identified child contacts in the intervention arm whose caregiver agrees for them to be screened by community-based healthcare workers in their home

Proportion of TB index patient households in the intervention arm who refused a household visit due to stigma

Proportion of TB index patient households in the intervention arm who refused a household visit due to lack of time

Proportion of TB index patient households in the intervention arm who refused a household visit due to well status of the child

Proportion of TB index patient households in the intervention arm who do not receive a home visit because the family and/or child was not home on three separate attempts

Proportion of TB index patient households in the intervention arm who refused a household visit due to fear of disclosure

Inclusion Criteria: Inclusion Criteria for Index Patient in South Africa: Adult pulmonary TB patient being treated for pulmonary TB disease (with bacteriologic confirmation including smear, GeneXpert® MTB/RIF and/or liquid mycobacterial culture) in one of the participating clinics Willing to have a home visit and disclose their diagnosis to household members Lives in the catchment areas of a study clinic Age 18 years or older Provides informed consent Inclusion Criteria for Index Patient in Ethiopia: Adult or child being treated for pulmonary TB disease (clinical with or without microbiologic diagnosis) in one of the participating clinics Willing to have a home visit and disclose their diagnosis to household members Lives in the catchment areas of a study clinic Age 18 years or older Provides informed consent Inclusion Criteria for Contact in South Africa: Child less than 15 years old living in the household of the TB index patient Caregiver willing to provide informed consent Children 7 years and older must also provide assent Inclusion Criteria for Contact in Ethiopia: Child less than 15 years old living in the household of the TB index patient Caregiver willing to provide informed consent Children 12 years and older must also provide assent Exclusion Criteria: Exclusion Criteria for Index Patient in South Africa and Ethiopia: Extrapulmonary TB with no evidence of concurrent pulmonary TB Evidence of rifampin and/or isoniazid resistance on GeneXpert® MTB/RIF or drug sensitivity testing performed on M. tuberculosis isolates identified on culture Household has already participated in study (more than one index patient in the household) and child contacts under 15 years have already been evaluated and remain on TPT and/or TB treatment Exclusion Criteria for Contact in South Africa: Household child contact of a drug-resistant TB index patient. (Drug resistance defined either by the presence of rifampin resistance on GeneXpert® MTB/RIF OR drug sensitivity testing on isolates identified by culture). Exclusion Criteria for Contact in Ethiopia: Household child contact of a drug-resistant TB index patient. (Drug resistance defined either by the presence of rifampin resistance on GeneXpert® MTB/RIF OR drug sensitivity testing on isolates identified by culture). Household contact 15 years and older