Addressing social determinants to end tuberculosis

Leave no one behind. This is the overarching pledge of the Sustainable Development Goals; a pledge that is far from being realised. In 2016, more than 4 million people with tuberculosis were estimated to be undiagnosed or their care and treatment were unknown.¹ In the same year, nearly a fifth of the people who were diagnosed and known to be treated for tuberculosis had adverse outcomes, including 1·3 million deaths.¹ One reason that millions of people affected by tuberculosis are left behind is an absence of coordinated, international action to combat poverty and inequality.

Despite renewed interest in addressing the social determinants of tuberculosis, there remains a stark global disparity in disease burden and access to care.² Low-income and middle-income countries (LMICs) bear the highest tuberculosis burden¹ and people with tuberculosis are often vulnerable and impoverished.³ Conversely, in high-income countries, improvements in living conditions, increased social expenditure per person, and strategies to address the social determinants of health have been associated with decreased tuberculosis prevalence.^4–6 WHO’s 2015 End TB Strategy acknowledges the need to reduce inequalities in tuberculosis prevention and care, including through provision of social protection and poverty alleviation for households affected by tuberculosis.⁷

For the past two decades, we, the Innovation for Health and Development (IFHAD) team, have been working with impoverished periurban shantytown communities of Callao, Peru, to generate evidence on the feasibility, acceptability, and impact of socioeconomic support for people affected by tuberculosis. Our cohort study of households affected by tuberculosis defined a threshold above which costs of accessing “free” tuberculosis care became catastrophic, predicting loss to follow-up, treatment failure, treatment non-completion, tuberculosis recurrence, and death.³ This threshold was endorsed in the End TB Strategy, which mandated that “zero TB-affected families should face catastrophic costs by 2025”.⁷ However, little rigorous evidence is available to guide cost mitigation strategies for tuberculosis- affected households.

The Innovative Socioeconomic Interventions Against Tuberculosis (ISIAT) project addressed this knowledge gap by offering a panel of diverse psychosocial and economic support to enhance access to tuberculosis care and increase prevention and cure.⁸ Education, community mobilisation, psychological and social support were integrated with povertyreduction interventions, including food and cash transfers, linkage to universal health coverage, microcredit, microenterprise, and vocational training. Implementation increased successful tuberculosis treatment completion by patients with tuberculosis and preventive therapy completion by their household contacts.⁸

The most acceptable, feasible components of ISIAT were then combined into a standardised intervention that consisted of integrated social support (household visits and “TB Clubs” involving peer-led mutual support groups) and economic support (conditional cash transfers), evaluated in the Household-Randomised Evaluation of a Socioeconomic Intervention to Prevent Tuberculosis (HRESIPT).⁹ HRESIPT showed that patients offered the intervention were more likely to achieve treatment success,⁹ their contacts more likely to initiate preventive therapy,¹⁰ and the household less likely to incur catastrophic costs.¹¹ The intervention was simplified further, including relaxation of cash transfer conditionality, a stronger focus on participant empowerment, and consolidation of a civil society of “tuberculosis survivors” who provide improved peer support to affected households. The impact and costeffectiveness on long-term cure, tuberculosis casefinding, and risk-targeted tuberculosis prevention¹² are being assessed in the Community Randomized Evaluation of a Socioeconomic Intervention to Prevent Tuberculosis (CRESIPT) trial. CRESIPT’s results will be instructive, but in the meantime important questions remain largely unaddressed.

First, while socioeconomic support for tuberculosis- affected households has had positive impacts on cure and prevention in a few settings,^13–15 including Peru,^8–11,16,17 its wider transferability is unknown. Socioeconomic interventions are likely to benefit from adaptation to, and evaluation in, settings with diverse patient demographics (eg, rates of poverty, substance use, HIV, and tuberculosis drug resistance), health and social care systems, and infrastructure.

Second, current global policies focus on providing tuberculosis-affected households with economic support to prevent catastrophic costs associated with treatment, travel to receive care, and time off work. Indeed, economic incentives and enablers for tuberculosis have a long history. However, it is striking that in ISIAT, HRESIPT, and CRESIPT, participants consistently valued social support more highly than economic support.^16,17 This is important because social support is, generally, much less expensive than economic support but can be more labour intensive. Social or psychosocial support might have a similar impact on tuberculosis prevention and outcome measures to economic or combined support.¹⁵ Unlike economic support, which has its own target inthe End TB Strategy in the form of the catastrophic costs, social support does not have a global indicator and is in danger of being overlooked. Furthermore, there has, to our knowledge, been no robust trial to evaluate social versus economic versus socioeconomic support versus existing standard of care for tuberculosis-affected households. The results from Peru highlight the potential value of engaging tuberculosis-affected households and civil society to establish what type of socioeconomic support communities would find valuable, acceptable, locally appropriate, and sustainable.

Third, cost-opportunity and cost-effectiveness of socioeconomic support for tuberculosis-affected households are likely to depend on targeted strategies that focus on specific tuberculosis-affected beneficiaries, but relevant analyses are lacking. There is an evidence gap about the optimal allocation of already scarce resources. For example, it is unknown whether a fairly expensive, labour-intensive intervention for high-risk groups (eg, people who are homeless, incarcerated, living in extreme poverty, or have multidrug-resistant tuberculosis) would be more cost-effective and impactful on rates of tuberculosis treatment success in national tuberculosis programmes than a cheaper, simpler intervention that supported all tuberculosis-affected households.

Fourth, it is not known how a disease-orientated approach that provides tuberculosis-specific support focused on tuberculosis-affected households, as in ISIAT, HRESIPT and CRESIPT, might enhance and be complemented by tuberculosis-sensitive approaches. Tuberculosis-specific approaches focus on supporting households after diagnosis and throughout treatment, and their success is measured by short-term, predominantly tuberculosis-related indicators. Tuberculosis-sensitive approaches involve strengthening national social protection strategies and adapting them to be more inclusive to people at risk of tuberculosis infection and disease, such as Brazil’s Bolsa Família social welfare programme, which has been associated with improved tuberculosis treatment outcomes and reduced incidence.^18,19 Although initially expensive,²⁰ tuberculosis-sensitive interventions could lead to improved long-term population outcomes in health (including tuberculosis), food security, and poverty alleviation. Thus, a combination of tuberculosis-sensitive and tuberculosisspecific support is likely to be the most impactful approach to comprehensively address the social determinants of tuberculosis.

Finally, consideration should be given to how socioeconomic support would be funded (eg, governmental, charitable, crowd-sourcing), by whom it will be delivered (eg, national tuberculosis programmes, the third sector, civil society), how it might be improved by streamlined service delivery (eg, ambulatory care, decentralisation, out-of-hours services), and the timeframes, indicators, and outcome measures used to judge its success (eg, tuberculosis-related, broader health outcomes, psychosocial and economic impact).

Socioeconomic interventions for tuberculosis can be achievable and impactful^8,9 but require local adaptation tailored to meet the needs of diverse populations and underserved groups. To end tuberculosis, medicines must be integrated with socioeconomic interventions that fight poverty.