Vivarium CSU Latent Tuberculosis Infection Simulation

List of abbreviations

Label	Definition
AcTB	Active tuberculosis
HBCs	High TB burden countries
LMICs	Low- and middle-income countries
LTBI	Latent tuberculosis infection
PLHIV	People living with HIV
U5 HHC	Children under 5 living in a household with a person with AcTB
TB	Tuberculosis
TPT	Tuberculosis preventive treatment
6H	Daily isoniazid monotherapy administered daily for six months
3HP	A short-course combination treatment, rifapentine and isoniazid, administered for three months once a week

1.0 Background

1.1 Project overview

This project used Vivarium framework to model the impact of treating LTBI on TB disease burden in five HBCs: India, South Africa, Philippines, Ethiopia, and Peru. The simulation is designed to estimate the extent to which the scale-up of TPT (6H and 3HP) reaching 95% of two high-risk populations (U5 HHC and PLHIV) in five HBCs improves health outcomes and reduces the disease burden of TB over a 5-year timeframe from 2020 to 2024, compared to a “business-as-usual” projection.

1.2 Literature review

TB is the leading infectious cause of death globally, with particularly high disease burden in LMICs. The GBD 2017 study estimates that 1.9 billion people globally may harbor LTBI. [GBD-2017] TPT to reduce the risk of progression from LTBI to AcTB is important not only at the individual level but also at the population level in order to achieve TB elimination per the WHO End TB Strategy. WHO recommends that high-risk populations, such as children under 5 who are household contacts of persons with AcTB and people living with HIV, should receive TPT. [Global-Tuberculosis-Programme-2018] Our client has a long history in TB control and offers programmatic support with partners from the global Stop TB initiative. As part of its commitment to reduce the burden of TB, the client developed 3HP which may increase adherence and treatment completion rates compared to 6H.

2.0 Modeling aims and objectives

To assess the impact of scaling up different treatments for LTBI on TB disease burden over five years. The project provides information on the relationship between different treatments and health outcomes such as the incidence of AcTB, deaths due to AcTB, and DALYs due to AcTB. The simulation forecasts a baseline and two intervention scenarios and the resulting health outcomes from years 2020 through 2024 for five HBCs.

3.0 Causal framework

3.1 Causal diagram

Note

link to DAGs page use round circles with DAGs

Outcome (O):

Most proximal determinant/exposure (E):

Confounders (C):

Effect modifiers:

Mediators (M):

4.0 Intervention

Scale-up of TPT coverage among two high-risk groups: U5 HHC and PLHIV.

4.1 Simulation scenarios

The project compared the health outcomes in three scenarios:

• Baseline (6H as projected): The baseline scenario projects GBD country-specific demographic and disease trends into the future, as well as the historic coverage of 6H treatment for U5 HHC and PLHIV, thus simulating what would happen if these trends continue “business as usual.”

• Intervention scenario 1 (6H scale up): This scenario is identical to the baseline scenario but scales up 6H to reach 95% coverage for U5 HHC and PLHIV by 2025. Thus, this scenario compares what would happen if countries were to increase coverage, utilizing the same treatment regimen (6H) as seen in the baseline scenario.

• Intervention scenario 2 (3HP scale up): This scenario is identical to the baseline scenario but scales up 3HP to reach 95% coverage for U5 HHC and PLHIV by 2025. Thus, this scenario compares what would happen if countries were to increase coverage and utilize a new treatment regimen (3HP).

Treatment coverage in baseline and intervention scenarios:

• Baseline coverage data

• Intervention coverage data shift

• Python code for adjusting coverage data

The country- and risk-group-specific TPT coverage data is informed by country-specific National Strategic Plan and WHO global tuberculosis report.

5.0 Vivarium modeling components

5.1 Vivarium concept model

5.2 Demographics

5.2.1 Population description

• Cohort type: Prospective closed cohort

• Time span: Jan 1, 2019 to Dec 31, 2024

• Time step: 30 days

• Size of largest starting population: 100,000

• Youngest start age and oldest end age: 0-95+ years

• Location: India, South Africa, Philippines, Ethiopia, and Peru

5.3 Models

5.3.1 Cause model

see LTBI cause model

5.3.2 Risk factor model

see LTBI risk factor model

5.3.3 Treatment model

see LTBI treatment model model

5.4 Desired outputs

Output shell table

Location	Year	Age group	Sex	Risk group	Scenario	Treatment group	Outcome
Ethiopia	2019	0 to 4	Female	General population	Baseline (6H as projected)	6H adherent	Active TB Incidence count (cases)
India	2020	5 to 14	Male	PLHIV	Intervention 1 (6H scale up)	6H non-adherent	Active TB Incidence rate (cases per 100,000 person-years)
Peru	2021	15 to 59	Both	U5 HHC	Intervention 2 (3HP scale up)	3HP adherent	DALYs due to Active TB (per 100,000 person-years)
Philippines	2022	60 plus				3HP non-adherent	DALYs due to HIV resulting in other diseases (per 100,000 person-years)
South Africa	2023	All ages				Untreated	Deaths due to Active TB (per 100,000 person-years)
	2024					All	Deaths due to HIV resulting in other diseases (per 100,000 person-years)
							Deaths due to other causes (per 100,000 person-years)
							Person-Years
							Treatment Coverage (proportion)
							Ylls due to other causes (per 100,000 person-years)

6.0 Validation and verification

Todo

Add V&V strategy

7.0 Limitations

• We assume same efficacy for 6H and 3HP as literature evidence shows that 6HP is non-inferior to 3HP on protecting patients progress from LTBI to active TB.

• We assume perfect screening for active TB, which may not be the case in reality (e.g., some individuals eligible for treatment may not receive it, while others that have already progressed to active TB may receive treatment for LTBI, depending on quality of screening).

• We assume same TB disease duration for both HIV-positive population and HIV-negative population.

• We do not account for the reduced risk of onward TB transmission by people treated for LTBI (“transmission dynamics”), which likely leads to a more conservative estimate of treatment impact.

• The relative risk of hosuehold contact exposure does not quantify the relationship between TB risk, household size, and income.

• Treatment for active TB is not assessed in this simulation.

• Adverse events are not captured in this simulation.

• Cost effectiveness analysis is not included in this simulation.

8.0 References

[GBD-2017]

James SL, Abate D, Abate KH, et al. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. The Lancet 2018; 392: 1789–858.

[Global-Tuberculosis-Programme-2018]

Global Tuberculosis Programme. Latent tuberculosis infection: updated and consolidated guidelines for programmatic management. 2018 http://www.ncbi.nlm.nih.gov/books/NBK531235/ (accessed Jan 8, 2020).

Country-specific TPT coverage sources:

• [Ethiopia] Federal Democratic Republic of Ethiopia National Strategic Plan Tuberculosis and Leprosy Control 2006-2013 EC (2013/14-2020). Ministry of Health.

• [India] National Strategic Plan for Tuberculosis Elimination 2017-2025. Revised National Tuberculosis Control Programme. Central TB Division, Directorate General of Health Services, Ministry of Health with Family Welfare, Nirman Bhavan, New Delhi, India. 2017.

• [South_Africa] South Africa’s National Strategic Plan for HIV, TB and STIs 2017-2022. South African National AIDS Council.

• [Peru] Tuberculosis in the Americas 2018. Pan American Health Association and World Health Organization; 2018.

• [Philippines] 2017-2022 Philippine Strategic TB Elimination Plan: Phase 1 (PhilSTEP1). Department of Health, Philippines.