CADTH Health Technology Review

Incentives and Support Programs to Improve Adherence to Tuberculosis Treatment

Rapid Review

Authors: Charlotte Wells, Melissa Severn

Key Messages

• Three overviews of reviews and 11 systematic reviews were identified regarding the clinical effectiveness of adherence incentives in those who require assistance to complete their tuberculosis treatment.

• Four evidence-based guidelines were identified that provided recommendations regarding the use of adherence incentives in those who require assistance completing their tuberculosis treatment.

• The reported clinical effectiveness of adherence incentives for patients with tuberculosis was mixed. There were no detrimental effects of providing incentives, but there was also no conclusive evidence pointing to a clinical benefit. The overall quality of the included reviews was moderate to high.

• The included guidelines recommended that incentives and enablers be included as a part of a patient-centred strategy for treatment and for patients with active tuberculosis or patients at high risk; however, the evidence formulating these recommendations was of low certainty or quality. Two of the included guidelines were of high methodological quality, and 2 were of lower methodological quality.

Context and Policy Issues

Tuberculosis (TB) is an infectious disease caused by the bacterium Mycobacterium tuberculosis.¹ Initial infection leads to latent TB, which has no active symptomology as the disease is contained by the host’s immune defences. However, in a small subset of infected patients (5% to 10%), the infection can proceed to active TB, which has visible symptoms and high mortality.² Worldwide, TB kills more people than any other infectious disease. WHO’s End TB Strategy aims to reduce TB death by 90% by 2035.²

TB is curable, with TB treatment given in 2 phases: initial intensive treatment and continuation treatment. In the initial phase, it is recommended that the medication be given daily; in the continuation phase, the medication can be given daily or intermittently.³ As treatment is intensive and frequent, medication nonadherence is a problem and can lead to poorer patient outcomes and development of drug-resistant TB.⁴ The intensive nature of treatment can lead to significant barriers for people attending treatment, especially those in hard-to-reach populations or those who are poorer and cannot afford to take time off work or to travel to the clinic for treatment. Therefore, a multi-faceted, patient-centred treatment strategy is often cited as an option to help overcome these barriers. Components of a patient-centred strategy could include enablers, such as transportation vouchers and social service assistance, and incentives, such as food stamps, snacks and meals, and provision of housing, stipends, coupons, or cash.⁵ Incentives are generally defined as items or services that reward healthy behaviour and enablers are defined as items or services that remove barriers to accessing health care.⁶

This report is an upgrade of a previous CADTH report, Support Programs for Tuberculosis Treatment: Clinical Utility and Guidelines, with an updated search using broader search terms.⁷ The purpose of this review is to identify clinical studies of support programs, material incentives, or material enablers in the treatment of TB, and to summarize the clinical effectiveness of these interventions. Evidence-based guidelines were also identified, and the recommendations regarding support programs and incentives were summarized.

This report is a component of a larger CADTH Condition Level Review on TB. A condition level review is an assessment that incorporates all aspects of a condition, from prevention and detection to treatment and management. For more information on CADTH’s Condition Level Review of TB, please visit the TB project page on CADTH’s website.

Research Questions

1. What is the clinical effectiveness of adherence incentives and support programs in those who require assistance to complete their tuberculosis treatment?

2. What are the evidence-based guidelines regarding the use of adherence incentives and support programs in those who require assistance completing their tuberculosis treatment?

Methods

Literature Search Methods

A limited literature search was conducted by an information specialist on key resources including MEDLINE, the Cochrane Library, the University of York Centre for Reviews and Dissemination (CRD) databases, the websites of Canadian and major international health technology agencies, as well as a focused internet search. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were adherence incentives and tuberculosis patients. Search filters were applied to limit retrieval to health technology assessments, systematic reviews (SRs), meta-analyses, or network meta-analyses, any types of clinical trials or observational studies, and guidelines. The search was also limited to English-language documents published between January 1, 2014, and November 11, 2020.

Selection Criteria and Methods

One reviewer screened citations and selected studies. In the first level of screening, titles and abstracts were reviewed and potentially relevant articles were retrieved and assessed for inclusion. The final selection of full-text articles was based on the inclusion criteria presented in Table 1.

Exclusion Criteria

Articles were excluded if they did not meet the selection criteria outlined in Table 1, were duplicate publications, or were published before 2014. The year 2014 was chosen as the cut-off to match the search dates of the previous CADTH report.⁷ Systematic reviews with all relevant studies captured in other more recent or more comprehensive SRs were excluded. Primary studies were excluded from the report due to the abundance of SRs and overviews. SRs that were retrieved by the search were excluded if they were captured in 1 or more included overviews of reviews. Guidelines with unclear methodology were also excluded.

Critical Appraisal of Individual Studies

The included publications were critically appraised by 1 reviewer using the following tools: A MeaSurement Tool to Assess Systematic Reviews 2 (AMSTAR 2)⁸ for SRs, with additional considerations for overviews of reviews, and the Appraisal of Guidelines for Research and Evaluation (AGREE) II instrument⁹ for guidelines. Summary scores were not calculated for the included studies; rather, the strengths and limitations of each included publication were described narratively.

Summary of Evidence

Quantity of Research Available

A total of 374 citations were identified in the literature search. Following screening of titles and abstracts, 331 citations were excluded and 43 potentially relevant reports from the electronic search were retrieved for full-text review. Four potentially relevant publications were retrieved from the grey literature search for full-text review. Of these potentially relevant articles, 29 publications were excluded for various reasons, and 18 publications met the inclusion criteria and were included in this report. These comprised 3 overviews of reviews, 11 SRs, and 4 evidence-based guidelines. Appendix 1 presents the PRISMA¹⁰ flow chart of the study selection.

Additional references of potential interest are provided in Appendix 6.

Summary of Study Characteristics

Additional details regarding the characteristics of included publications are provided in Appendix 2.

Study Design

Three overviews of reviews and 11 SRs were identified.^4,11-23 The overviews of reviews were published in 2018¹¹ and 2017.^12,13 The SRs were published in 2020,¹⁴ 2019,^4,15 2018,^16-19 2017,^20,21 and 2016.^22,23 There were 7 SRs with meta-analyses, but not all SRs with a meta-analysis included relevant information within the meta-analyzed data (i.e., the relevant data were narratively described).^{14-16,18,19,22,23} The date ranges for the overviews of reviews were up to 2017¹¹ and 2016.^12,13 The date ranges for the SR searches were up to 2018,^4,14 2017,^11,15,19 2016,^12,13,17 2015,^18,20,21,23 and 2014.²²

There was significant overlap between the SRs. Details on the extent of overlap in the studies is provided in Appendix 5.

The scope of the included SRs was broader than the scope of the present report, with the exception of Richterman et al. (2018).¹⁹ This was because the majority of SRs included other interventions to promote adherence to treatment,^{4,16-18,20,23} interventions to promote adherence or completion of screening,^14,21,22 or additional populations such as patients with HIV,^11,20 vulnerable populations without TB,¹¹ or patients in low-income countries.^12,13 Appendix 2 details how many primary studies were included in each SR and how many of those included studies that were relevant to this report.

Four guidelines were identified. Two of these guidelines were separate chapters of an overarching guideline (the Canadian Tuberculosis Standards).^3,24 The developing institutions for the identified guidelines were the WHO,²⁵ the National Institute for Health and Care Excellence (NICE),²⁶ and the Canadian Thoracic Society in collaboration with the Public Health Agency of Canada.^3,24

A comprehensive literature search (or multiple literature searches) was used to inform the included guidelines. The authors of the WHO guideline searched for randomized controlled trials (RCTs) that had direct comparisons and used teams of experts to develop the recommendations through consensus and discussion.²⁵ The authors of the NICE guidelines searched for reviews, RCTs, or observational studies, and used a guideline development group who developed the recommendations through consensus.²⁶ The Canadian Tuberculosis Standards were not clear on the guideline development process, but noted that they included “all published evidence” on the topics of interest.^3,24

The evidence was assessed by the publication authors using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) or a modified GRADE in all included guidelines.^3,24-26 The rating system for each guideline is reported in Appendix 2.

Country of Origin

The overviews were conducted by first authors from the UK,¹¹ Chile,¹² and South Africa,¹³ but the remainder of the authors in each of these reviews were from a variety of different countries.

The SRs were conducted by first authors from Canada,^14,15 Spain,⁴ the US,^16,17,19,20 Brazil,¹⁸ the Netherlands,^21,23 and Sweden.²²

The 4 guidelines were intended for use globally,²⁵ in the UK,²⁶ and in Canada.^3,24

Patient Population

The overviews of reviews examined vulnerable populations who are food insecure or malnourished,¹¹ and people in low-income countries.^12,13 Only data regarding people in low-income countries or vulnerable populations who also have TB were extracted.

The SRs examined patients who have latent TB,^14,17,22 drug-resistant TB,¹⁵ active TB,^16,18,19,21 either latent or active TB,^4,23 or patients who have TB and a co-occurring substance use disorder.²⁰

For the included guidelines, the target populations were patients with drug-susceptible TB,²⁵ children, young people and adults with latent TB,²⁶ and patients with active TB.^3,24

Interventions and Comparators

The overviews of reviews examined interventions such as community-based supplementary feeding programs,¹¹ alternative delivery and implementation strategies,¹² and financial arrangements, including incentives.^12,13 These interventions were compared to other strategies or no strategies or incentives.^11-13

The SRs examined interventions such as interventions designed to promote adherence to treatment,^4,16,21,22 interventions that addressed barriers to treatment compliance,¹⁷ material or financial incentives or support,^{14,15,18,19,23} educational programs,^14,18 psychosocial or psycho-emotional support,^15,23 food incentives or support,¹⁸ and contingency management interventions.²⁰ The general interventions to promote adherence or address barriers to treatment or screening included support programs, material incentives, or material enablers.^{4,16,17,21,22}

The SRs compared these support programs or incentives to control groups,^{4,14,15,17,20} no incentives or support,^15,17,22 other incentives or strategies,^17,18 or to usual care or standard support.^16,17,23

The guidelines examined treatment adherence interventions^25,26 and incentives and enablers to TB treatment.^3,24

Outcomes

The outcomes for the overviews of reviews and the SRs were treatment adherence^{4,11-13,15-17,20,22,23}; mortality^11,16,18; treatment completion, success, or failure^{4,11-14,16,19,20,22,23}; cure rate or microbiologic cure^16,18,19; weight gain¹¹; quality of life (QoL)¹¹; effectiveness^21,22; sputum conversion^4,11; loss to follow-up^15,16; default rate^16,18; and missed doses.^4,20

The guidelines examined outcomes related to treatment adherence,^3,24,26 treatment success,²⁵ treatment completion,²⁵ and sputum conversion²⁵ and used these outcomes as metrics to develop the recommendations.

Summary of Critical Appraisal

Additional details regarding the strengths and limitations of included publications are provided in Appendix 3.

Overviews of Reviews

The overviews of reviews were of high quality. All 3 reviews were published by Cochrane and therefore followed Cochrane methodology, which was provided and detailed.^11-13

The methods, research questions, inclusion criteria, and study details were clear. The studies used comprehensive literature searches, performed study selection and data extraction in duplicate, and provided excluded studies lists with reasons.

Pantoja et al.¹² and Wiysonge et al.¹³ did not include an analysis on overlap of the primary studies examined by the included SRs. Additionally, all 3 studies did not report on the sources of funding for the studies included in the reviews.^11-13

Systematic Reviews

The included SRs were of variable quality.^4,14-23

All the SRs^4,14-23 included a comprehensive literature search strategy (although Alipanah et al.¹⁶ only searched 1 database). Additionally, all the included SRs, with the exception of Herrmann et al.,²⁰ assessed study quality or risk of bias using validated tools, and clearly reported the quality of the included primary studies.^{4,14-19,21-23}

Eight SRs performed study selection in duplicate.^4,14,18-23 Three SRs had study selection performed by only 1 reviewer, which may have led to missed studies during selection.^15-17 Additionally, for data extraction, 4 SRs did not have either duplicate extraction or single extraction with an independent verifier.^4,15-17

Seven SRs conducted their final searches over 1.5 years before publication.^{14,15,17,18,20-22} This does not necessarily mean that the conclusions presented in the SRs were erroneous, but it could mean that studies published between the search date and publication date were potentially missed. Additionally, 5 SRs only included studies in English^14,16,17,20 (with 1 including English and Spanish only⁴), which limits the potential number of included studies and relevant information, especially for SRs that focused on individuals living in lower-income countries. Although information from low-income countries with a higher TB incidence may not be as relevant to the Canadian context, many low-incidence countries have primary languages that are not English, and this limitation on language may have eliminated potentially relevant studies from those countries.

Guidelines

The critical appraisal of the guidelines was performed in previous CADTH reports.

The guidelines by WHO,²⁵ NICE,²⁶ and the Canadian Tuberculosis Standards chapter on active tuberculosis³ were assessed in the CADTH report Treatment of Tuberculosis: A Review of Guidelines,²⁷ available at the CADTH website. The guideline from the Canadian Tuberculosis Standards chapter on identification of TB in high-risk populations²⁴ was assessed in the CADTH report Identification of Tuberculosis: A Review of the Guidelines.²⁸

More details on the quality of these guidelines can be found in those reports. Briefly, the WHO guideline²⁵ and NICE guideline²⁶ were assessed to be of high quality, with clear descriptions of scope, populations, target users, and recommendations. These guidelines had a systematic approach to evidence synthesis, with evaluation of the primary literature, and transparent literature search methods and recommendation development.^25,26

The Canadian Tuberculosis Standards chapter on active tuberculosis (chapter 5)³ and the Canadian Tuberculosis Standards chapter on identification of TB in high-risk populations (chapter 13)²⁴ were assessed to be lower quality compared with other identified guidelines. They had clear recommendations but limited detail on methods, such as recommendation development, research questions, the professions involved in the development, search methods, assessment of primary study quality, or external review.^3,24

Summary of Findings

Appendix 4 presents the main study findings and authors’ conclusions.

Clinical Effectiveness of Support Programs for Treatment of Tuberculosis

Treatment adherence and completion were the most common outcomes examined by the studies.

Treatment Acceptance (Initiation of Treatment)

In 1 SR by Barss et al., acceptance of treatment (i.e., a patient starting treatment) improved with both patient incentives and patient education supports. When patients were provided with an incentive, 49 additional patients per 100 patients recommended for treatment accepted the treatment (95% confidence interval [CI], 46 to 52). With patient education support, 15 additional patients accepted treatment per 100 recommended (95% CI, 11 to 19).¹⁴

Treatment Adherence

Financial Incentives

In 1 cohort study identified by Herrmann et al.,²⁰ patients received US$15 in subway tokens per week for all doses ingested or a combination of this incentive and monthly bonuses of US$30 to US$60 for those who took more than 80% of the doses for the month. Patients who received US$15 in subway tokens per week for all doses ingested were 2.7 times more likely to take more than 80% of doses if they were given the larger incentive package. Herrmann et al.²⁰ also reported on an RCT that examined giving people who inject drugs either US$10 a month for adherence paid once a month for 6 months or US$10 per month for adherence paid out as a lump sum after 6 months. There was no difference in TB treatment adherence between the groups.

Non-Cash Incentives and Educational Support

In Alipanah et al.,¹⁶ the relative risk for treatment adherence for oral and written educational material compared with standard care was 1.83 (95% CI, 1.14 to 2.92) for 1 included RCT and 1.21 (95% CI, 1.05 to 1.40) for 1 included cohort study, indicating patients who received the oral and written education were 1.83 times more likely to adhere to treatment in the included RCT and 1.21 times more likely to adhere to treatment in the included cohort study.

Unclear Incentives

Heuvelings et al. reported 1 study comparing directly observed therapy alone with directly observed therapy in combination with incentives in hard-to-reach populations and found that incentives were beneficial for treatment adherence.²¹ It was unclear what these incentives were.

Treatment Completion, Success, and Failure

Pantoja et al.¹² and Wiysonge et al.¹³ both included the same SR and concluded that sustained material incentives made little or no difference in treatment completion for active TB compared with no incentives, although they did not provide numerical values.

Financial Incentives

Richterman et al.¹⁹ meta-analyzed 4 studies that examined cash incentives compared with usual care and found a significant benefit of cash incentives for treatment success (odds ratio [OR] = 1.77; 95% CI, 1.57 to 2.01). In the non–meta-analyzed studies, treatment success was more likely with a monthly cash incentive equivalent to a low civil service salary (OR = 1.19; 95% CI, 1.03 to 1.37), and treatment completion was more likely with cash incentives (OR = 3.28; 95% CI, 1.65 to 6.51).¹⁹

Alipanah et al.¹⁶ included 5 RCTs and 4 cohort studies examining financial incentives compared with standard of care (directly observed therapy or self-administered therapy) and reported that the pooled risk ratio (RR) for treatment completion in the 5 RCTs significantly favoured the intervention (RR = 1.23; 95% CI, 1.15 to 1.31), but the pooled RR for the 4 cohort studies was not significantly different for treatment completion. A similar effect occurred for treatment failure; the RR for 1 RCT was significant in favour of the intervention (RR = 0.66; 95% CI, 0.50 to 0.87), but the pooled RR for 2 cohort studies was not significant. The use of financial incentives was associated with significantly higher treatment success in both pooled RCTs (3 RCTs: RR = 1.07; 95% CI, 1.03 to 1.11) and cohort studies (4 cohort studies: RR = 1.25; 95% CI, 1.09 to 1.42).

For latent TB, Riquelme-Miralles et al.⁴ reported no benefit of cash incentives in people who are homeless compared with non-cash incentives, and no benefit in people who received monetary incentives compared with no incentives. Both these studies were also reported in Liu et al.¹⁷ and Herrmann et al.²⁰

In people who use drugs, an immediate incentive did not have a significant impact on treatment completion compared with a deferred incentive.¹⁷ When providing a monetary incentive or outreach alone, the monetary incentive was more effective for increasing treatment completion (OR = 45.5; 95% CI, 9.7 to 214.6). The primary study examining outreach and incentives was also reported in Herrmann et al.²⁰ and Stuurman et al.²²

In individuals who are incarcerated, the odds of treatment completion was higher with monetary incentives compared to usual care (OR = 1.07; 95% CI, 0.47 to 2.40).¹⁷ This study was also reported in Herrmann et al.²⁰

Food Supplements or Incentives

Riquelme-Miralles et al.⁴ reported a benefit of food incentives for treatment completion in patients with active TB compared with standard care (98% versus 82%) in 1 study, but no benefit in a second study (76% versus 78%). No statistical results were reported.⁴

One cohort study identified by Herrmann et al.²⁰ reported that a US$5 grocery gift card in addition to directly observed therapy made adults or children with medication nonadherence 5.7 times more likely to complete treatment (value not reported). This study was also reported in Alipanah et al.,¹⁶ van Hoorn et al.,²³ and Heuvelings et al.²¹

Non-Cash Incentives and Educational Support

Alipanah et al.¹⁶ included 1 RCT that examined education supports compared with no supports and reported a higher rate of treatment completion for the intervention group (RR = 1.71; 95% CI, 1.32 to 2.22). There was no significant effect on treatment success or failure.

Stuurman et al.,²² Liu et al.,¹⁷ Riquelme-Miralles et al.,⁴ and Herrmann et al.²⁰ all reported an RCT examining patients who were incarcerated with latent TB. The intervention was an informational or educational session combined with US$25 vouchers for food or transport if they attended a TB clinic within 1 month of release from incarceration compared with a control group (group receiving neither intervention). A second comparison was the informational session alone compared with a control group, which Liu et al. reported as significant (OR = 2.2; 95% CI, 1.04 to 4.72) in favour of the educational incentive. For the combination of non-cash incentive and education, the SRs^17,22 reported an OR of 1.07 (95% CI, 0.5 to 2.4) in favour of the incentives, but this was not statistically significant. However, Herrmann et al.²⁰ also reported this RCT, but reported opposite results: there was less treatment completion in the incentive group compared with controls. An examination of the original RCT²⁹ revealed that Herrmann et al. reported this incorrectly.

A study reported in Stuurman et al.²² reported that patients with TB who injected drugs and who received support through methadone treatment and counselling had higher odds of completing treatment compared with no incentive (OR = 14.5; 95% CI, 5.0 to 42), but this evidence was of very low quality.

Combined Incentives and Support

For active TB, Riquelme-Miralles et al.⁴ reported no benefit of financial incentives mixed with educational support compared with standard care over 6 to 9 months. In people with TB who use drugs, outreach mixed with incentives and incentives alone were more effective than outreach alone (52.8% versus 3.6% and 60% versus 3.6%, respectively). In adolescents, there was no benefit of peer counselling mixed with an incentive or incentives alone compared with standard of care.⁴ In a general population of patients with TB, there was a benefit of education and economic incentives compared with standard care over 9 to 12 months (63.8% versus 27.1%). No statistical results were reported.⁴

van Hoorn et al.²³ meta-analyzed studies providing socioeconomic supports (e.g., food supplementation and economic support), psycho-emotional supports (e.g., counselling, psychotherapy, and the organization of self-help groups), and combined supports for the outcome of treatment success. All the RRs significantly favoured the interventions over the control groups, with 4 studies examining socioeconomic supports (pooled RR = 1.08; 95% CI, 1.03 to 1.13) and 3 studies examining combined supports (pooled RR = 1.17; 95% CI, 1.12 to 1.22). When examining the outcome of unsuccessful treatment, the interventions were also significantly favoured, with 2 studies examining socioeconomic supports (pooled RR = 0.78; 95% CI, 0.69 to 0.88) and 4 studies examining combined supports (pooled RR = 0.42; 95% CI, 0.23 to 0.75).²³

Losses to Follow-up

Financial Support

Law et al.¹⁵ conducted a meta-analysis of studies examining financial support (reimbursement of rent and travel expenses, and compensation of lost wages) compared with no support, and nutritional support (e.g., food baskets, provisions of basic foods, hot meals) compared with no support. In patients who received travel expenses (10 cohorts pooled together), the proportion of patients who were lost to follow-up was 0.15 (95% CI, 0.10 to 0.24); in patients who received rent and travel expenses (4 cohorts pooled together), the proportion of patients lost to follow-up was 0.08 (95% CI, 0.06 to 0.10); and in patients who received supplemental income (3 cohorts pooled together), the proportion of patients lost to follow-up was 0.06 (95% CI, 0.00 to 0.61). For patients who received no financial support, the proportion of patient lost to follow-up was 0.24 (95% CI, 0.17 to 0.34).¹⁵ There was a significant difference in loss to follow-up between the subgroups of travel reimbursement, travel and rent reimbursement, compensation of lost wages, and no treatment (P < 0.01), but there were no direct statistical comparisons provided.

In Alipanah et al.,¹⁶ financial incentives were associated with lower loss to follow-up (1 RCT: RR = 0.74; 95% CI, 0.60 to 0.90; 5 cohort studies: pooled RR = 0.48; 95% CI, 0.28 to 0.81). There did not appear to be any primary study overlap between Alipanah et al. and Law et al. for this intervention.

Food Packages

For food support, the proportion of patients (14 pooled cohorts) who received food packages and were lost to follow-up was 0.15 (95% CI, 0.10 to 0.22). The proportion of patients (17 pooled cohorts) who received no support and were lost to follow-up was 0.18 (95% CI, 0.06 to 0.27).¹⁵ There was no significant difference between these groups.¹⁵

Cure Rate and Microbiologic Changes

Material or Financial Incentives

Pantoja et al.¹² and Wiysonge et al.¹³ included the same SR and concluded that sustained material incentives had little to no difference in cure rates for active TB compared with no incentives, but numerical values were not provided.

Alipanah et al. found financial incentives were associated with higher rates of cure in pooled cohort studies but not RCTs (4 cohort studies: RR = 1.13; 95% CI, 1.02 to 1.26; 1 RCT: RR = 0.92; 95% CI, 0.85 to 1.01) and sputum conversion (1 RCT: RR = 1.21; 95% CI, 1.02 to 1.43).¹⁶

In Muller et al.,¹⁸ cure rates pooled from 2 RCTs were not significantly different between patients who received financial incentives and patients who received no financial incentives. One of the included RCTs in the pooled estimate from Muller et al.¹⁸ was the RR from the RCT included in Alipanah et al.¹⁶

In Richterman et al.,¹⁹ monthly cash transfer and travel reimbursement interventions and monthly cash incentives to households (through the Bolsa Família program) had higher odds of microbiologic cure compared with usual care (OR = 79.08; 95% CI, 4.42 to 1,413.33 and OR = 1.07; 95% CI, 1.04 to 1.11, respectively).

Food Supplementation

Visser et al.¹¹ included 1 SR of relevance to this report. This SR examined cure rates and sputum conversion in patients with TB, with or without co-occurring HIV, who received food supplementation or no supplementation. The cure rate was reported in 1 primary study and it was not significantly different for patients receiving interventions compared with control groups. Sputum conversion was also not different between the groups in 3 primary studies, but it was significantly better for patients receiving supplementation in 1 small study. The authors noted that the studies were underpowered for these outcomes.¹¹

In Muller et al.,¹⁸ cure rates pooled from 3 RCTs found no significant difference between patients who received food incentives and patients who did not.

Educational Supports

In Muller et al.,¹⁸ cure rates pooled from 2 RCTs were higher in patients who received education or counselling (RR = 1.16; 95% CI, 1.05 to 1.29).

Mortality

Visser et al.¹¹ included 1 SR that examined mortality outcomes of patients who received food supplementation compared with no supplementation at 1 year of follow-up. There was no significant difference between the groups, and no subgroup differences in patients who also had HIV.¹¹ Alipanah et al.¹⁶ reported that in pooling 3 cohort studies, there was a significant benefit of material or financial incentives on mortality for patients with TB (RR = 0.51; 95% CI, 0.37 to 0.71); however, in pooling 2 RCTs, there was no significant effect. Muller et al.¹⁸ pooled 2 RCTs and found no significant difference between patients who received financial incentives or none. One RCT included in the pooled estimate for Muller et al. was also included in the pooled estimate for Alipanah et al.¹⁶

Quality of Life

One SR identified in Visser et al.¹¹ examined QoL of patients receiving food supplementation compared with no supplementation. The supplementation may have improved QoL in the first 2 months of treatment, but the authors noted the evidence was of low certainty and narratively described (no statistical comparisons).¹¹

Guidelines Regarding Support Programs for Treatment of Tuberculosis

The guidelines from the WHO²⁵ recommended that health education and counselling for treatment adherence should be provided to patients who are on TB treatment (strong recommendation, moderate certainty in the evidence). It is also recommended providing a package of treatment adherence interventions to patients in conjunction with treatment administration (conditional recommendation, low certainty in the evidence), which could include material support to patients (conditional recommendation, moderate certainty in the evidence).²⁵

The guidelines from NICE²⁶ recommended that the care plan identify the reasons why a patient may not attend treatment, including determining any enablers or incentives to help patients overcome barriers to treatment. This plan should also define the supports needed to address needs, such as those to acquire housing. These guidelines also recommended that multidisciplinary teams implement strategies that encourage following treatment plans, including health education counselling, tailored health education booklets, and incentives and enablers to help people follow their treatment regime.²⁶ During treatment, TB teams should assess living situations of patients and work with agencies to provide accommodation for those that need it.²⁶ Housing should be funded by local government and clinical commissioning groups for individuals who are homeless and ineligible for state-funded accommodations. All these recommendations were deemed to do more good than harm for the vast majority of patients.²⁶

The Canadian Tuberculosis Standards chapter 5³ conditionally recommended that a comprehensive, patient-centred treatment program be provided for patients initiating treatment, although this was based on weak evidence. The key elements of a program such as this would include incentives and enablers, social service support, housing support, and provision of transportation.³ The Canadian Tuberculosis Standards chapter 13²⁴ also conditionally recommended — based on weak evidence — that individuals who are homeless and with medical conditions associated with high risks of reactivation should be considered for special measures such as incentives and enablers. Those who are at the highest risk in general should also be considered for incentives and enablers.²⁴

Limitations

There are limitations associated with the body of evidence and overall conclusions presented in this report.

Two of the overviews of reviews focused on low-income countries,^12,13 and therefore may not be generalizable to the Canadian context or to contexts in higher income countries with a low TB incidence. Additionally, there was a lot of heterogeneity in the studies that were included — the results from primary studies reported in the included SRs fell under the umbrella of types of supports but may have consisted of a variety of different interventions and comparators. Due to the limited reporting of many of the SRs, it was not clear exactly what were the interventions and the comparators.

This report only includes SRs and overviews of reviews; therefore, it is limited because the last search date of the SRs is the true cut-off for information from relevant primary studies. It is likely that some primary studies were missed in the analysis that were published between 2018 — the latest search date in the SRs — and 2020. Therefore, there may be key information missing from the analysis, and it is unknown if any primary studies have been published that would vastly change the clinical findings of this report.

The guidelines that were developed for the Canadian context^3,24 were of low methodological quality because of a lack of reporting of methods, and they were not specific in what incentives and enablers were recommended for individuals with TB. The other included guidelines^25,26 were of higher quality, but were not specific to the Canadian context, and therefore may have limited applicability in Canada.

Conclusions and Implications for Decision- or Policy-Making

Three overviews of reviews^11-13 and 11 SRs^4,11-23 were identified regarding the clinical effectiveness adherence incentives for those who require assistance to complete their tuberculosis treatment. Additionally, 4 guidelines^3,24-26 were identified that provided recommendations regarding the use of adherence incentives for those who require assistance completing their tuberculosis treatment. Two of these guidelines were separate chapters of an overarching guideline (the Canadian Tuberculosis Standards).^3,24

Overall, the results were neutral to positive for financial incentives and support, food incentives and support, educational incentives and support, non-cash incentives and support, and mixed supports. No studies found a detrimental clinical effect of provision of adherence incentives. The most-reported outcomes were treatment adherence and treatment completion, and the identified populations ranged from people with active or latent TB in the general public, people in low-income countries, people who use drugs, people who were homeless, and people who were incarcerated or newly released. There was significant overlap between the identified studies. The identified overviews of reviews concluded that sustained material incentives had little to no impact on cure rates or treatment completion for active TB,^12,13 and food incentives did not significantly affect cure rates, mortality, or sputum conversion.¹¹ However, there were some reported benefits of incentives, as reported in some included SRs, such as benefits of financial incentives on mortality, cure rates, and treatment success. However, these benefits were not sustained in every SR or across every primary study; therefore, it is not possible to conclusively determine a benefit of adherence incentives for treatment of TB.

The SRs were of variable quality. The SRs had comprehensive search strategies and frequently employed sound methodology, such as duplicate screening, duplicate data extraction, and assessments of risk of bias and primary study quality. However, many of the SRs had search dates that were significantly earlier than the publication dates of the report, and 5 of the SRs imposed language limitations on their searches which may have led to potentially missed studies.

The identified guidelines generally recommend the use of incentives and enablers in the provision and initiation of treatment but did not provide specific recommendations regarding which incentives should be given and to whom. Additionally, the evidence on which these recommendations were based was generally weak or of low certainty.

Implementation of policies and programs that provide incentives to individuals require suitable clinical evidence to justify the costs of operation. This report did not include questions regarding cost-effectiveness of these programs because it was not within the scope of the report; therefore, it is unclear whether the economic impact of these programs would be supported by a positive clinical impact to patients. It is necessary to consider the different needs of different populations when determining which incentives will be both appropriate and useful for patients.

References

1.Canadian Tuberculosis Standards 7th Edition - Chapter 2 - Pathogenesis and Transmission of Tuberculosis. Public Health Agency of Canada; The Lung Association; 2014: https://www.canada.ca/en/public-health/services/infectious-diseases/canadian-tuberculosis-standards-7th-edition/edition-14.html. Accessed 09/12/2020.

2.Global Tuberculosis Report 2020. World Health Organization; 2020: https://www.who.int/tb/publications/global_report/en/. Accessed 09/12/2020.

3.Canadian Tuberculosis Standards 7th Edition - Chapter 5: Treatment of Tuberculosis disease. Public Health Agency of Canada; The Lung Association; 2014: https://www.canada.ca/en/public-health/services/infectious-diseases/canadian-tuberculosis-standards-7th-edition/edition-17.html. Accessed 08/12/2020.

4.Riquelme-Miralles D, Palazon-Bru A, Sepehri A, Gil-Guillen VF. A systematic review of non-pharmacological interventions to improve therapeutic adherence in tuberculosis. Heart Lung. 2019;48(5):452-461. Medline

5.Official American Thoracic Society/Centers for Disease Control and Prevention/Infectious Diseases Society of America Clinical Practice Guidelines: Treatment of Drug-Susceptible Tuberculosis. American Thoracic Society/Centers for Disease Control; 2016: https://www.thoracic.org/statements/resources/tb-opi/treatment-of-drug-susceptible-tuberculosis.pdf. Accessed 09/12/2020.

6.Lutge EE, Wiysonge CS, Knight SE, Sinclair D, Volmink J. Incentives and enablers to improve adherence in tuberculosis. Cochrane Database Syst Rev. 2015(9):CD007952. Medline

7.Hill S, Gunn H, Picheca L. Support Programs for Tuberculosis Treatment: Clinical Utility and Guidelines. CADTH rapid response report: summary of abstracts. Ottawa: CADTH 2020: https://cadth.ca/sites/default/files/pdf/htis/2020/RB1514%20TB%20Support%20Programs%20Final.pdf. Accessed 08/12/2020.

8.Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ. 2017;358:j4008. Medline

9.Agree Next Steps C. The AGREE II Instrument. [Hamilton, ON]: AGREE Enterprise; 2017: https://www.agreetrust.org/wp-content/uploads/2017/12/AGREE-II-Users-Manual-and-23-item-Instrument-2009-Update-2017.pdf. Accessed YYYY Mmm DD.

10.Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. J Clin Epidemiol. 2009;62(10):e1-e34. Medline

11.Visser J, McLachlan MH, Maayan N, Garner P. Community-based supplementary feeding for food insecure, vulnerable and malnourished populations - an overview of systematic reviews. Cochrane Database Syst Rev. 2018;11:CD010578. Medline

12.Pantoja T, Opiyo N, Lewin S, et al. Implementation strategies for health systems in low-income countries: an overview of systematic reviews. Cochrane Database Syst Rev. 2017;9:CD011086. Medline

13.Wiysonge CS, Paulsen E, Lewin S, et al. Financial arrangements for health systems in low-income countries: an overview of systematic reviews. Cochrane Database Syst Rev. 2017;9:CD011084. Medline

14.Barss L, Moayedi-Nia S, Campbell JR, Oxlade O, Menzies D. Interventions to reduce losses in the cascade of care for latent tuberculosis: a systematic review and meta-analysis. Int J Tuberc Lung Dis. 2020;24(1):100-109. Medline

15.Law S, Daftary A, O'Donnell M, Padayatchi N, Calzavara L, Menzies D. Interventions to improve retention-in-care and treatment adherence among patients with drug-resistant tuberculosis: a systematic review. Eur Respir J. 2019;53(1):01.

16.Alipanah N, Jarlsberg L, Miller C, et al. Adherence interventions and outcomes of tuberculosis treatment: A systematic review and meta-analysis of trials and observational studies. PLoS Med. 2018;15(7):e1002595. Medline

17.Liu Y, Birch S, Newbold KB, Essue BM. Barriers to treatment adherence for individuals with latent tuberculosis infection: A systematic search and narrative synthesis of the literature. Int J Health Plann Manage. 2018;33(2):e416-e433. Medline

18.Muller AM, Osorio CS, Silva DR, Sbruzzi G, de Tarso P, Dalcin R. Interventions to improve adherence to tuberculosis treatment: systematic review and meta-analysis. Int J Tuberc Lung Dis. 2018;22(7):731-740. Medline

19.Richterman A, Steer-Massaro J, Jarolimova J, Luong Nguyen LB, Werdenberg J, Ivers LC. Cash interventions to improve clinical outcomes for pulmonary tuberculosis: systematic review and meta-analysis. Bull World Health Organ. 2018;96(7):471-483. Medline

20.Herrmann ES, Matusiewicz AK, Stitzer ML, Higgins ST, Sigmon SC, Heil SH. Contingency Management Interventions for HIV, Tuberculosis, and Hepatitis Control Among Individuals With Substance Use Disorders: A Systematized Review. J Subst Abuse Treat. 2017;72:117-125. Medline

21.Heuvelings CC, de Vries SG, Greve PF, et al. Effectiveness of interventions for diagnosis and treatment of tuberculosis in hard-to-reach populations in countries of low and medium tuberculosis incidence: a systematic review. Lancet Infect Dis. 2017;17(5):e144-e158. Medline

22.Stuurman AL, Vonk Noordegraaf-Schouten M, van Kessel F, Oordt-Speets AM, Sandgren A, van der Werf MJ. Interventions for improving adherence to treatment for latent tuberculosis infection: a systematic review. BMC Infect Dis. 2016;16:257. Medline

23.van Hoorn R, Jaramillo E, Collins D, Gebhard A, van den Hof S. The Effects of Psycho-Emotional and Socio-Economic Support for Tuberculosis Patients on Treatment Adherence and Treatment Outcomes - A Systematic Review and Meta-Analysis. PLoS ONE. 2016;11(4):e0154095. Medline

24.Canadian Tuberculosis Standards 7th Edition - Chapter 13: Tuberculosis surveillance and screening in selected high-risk populations. Public Health Agency of Canada; The Lung Association; 2014: https://www.canada.ca/en/public-health/services/infectious-diseases/canadian-tuberculosis-standards-7th-edition/edition-9.html. Accessed 08/12/2020.

25.Guidelines for treatment of drug-susceptible tuberculosis and patient care, 2017 update. Geneva: World Health Organization; 2017: https://www.who.int/tb/publications/2017/dstb_guidance_2017/en/. Accessed 08/12/2020.

26.Tuberculosis - NICE guideline [NG33]. National Institute for Health and Care Excellence; 2016: https://www.nice.org.uk/guidance/ng33. Accessed 08/12/2020.

27.Brett KD, C; Severn, M. Treatment of Tuberculosis: A Review of Guidelines. CADTH; 2020: https://cadth.ca/sites/default/files/pdf/htis/2020/RC1237%20TB%20treatment%20guidelines%20Final.pdf. Accessed 08/12/2020.

28.Brett KD, C; Severn, M. Identification of Tuberculosis: A Review of the Guidelines. CADTH; 2020: https://cadth.ca/sites/default/files/pdf/htis/2020/RC1236%20TB%20identification%20Final.pdf. Accessed 08/12/2020.

29.White MC, Tulsky JP, Goldenson J, Portillo CJ, Kawamura M, Menendez E. Randomized Controlled Trial of Interventions to Improve Follow-up for Latent Tuberculosis Infection After Release From Jail. Archives of Internal Medicine. 2002;162(9):1044-1050. Medline

30.Developing NICE guidelines: the manual. London: National Institute for Health and Care Excellence; 2014; updated 2018: https://www.nice.org.uk/process/pmg20/resources/developing-nice-guidelines-the-manual-pdf-72286708700869. Accessed 29/11/2020.

31.Preface: Canadian Tuberculosis Standards 7th Edition: 2014. Public Health Agency of Canada; 2014: https://www.canada.ca/en/public-health/services/infectious-diseases/canadian-tuberculosis-standards-7th-edition/edition-22.html. Accessed 29/11/2020.

32.V RMMAC-M. Evidence review on the effectiveness and cost-effectiveness of interventions aimed at identifying people with tuberculosis and/or raising awareness of tuberculosis among hard-to-reach groups. NICE; 2011: https://www.nice.org.uk/guidance/ng33/evidence/appendix-g8. Accessed 08/12/2020.

Appendix 1: Selection of Included Studies

Figure 1: Selection of Included Studies

Table 4: Rating System for Included Guidelines

Table 7: Strengths and Limitations of Guidelines Using AGREE II⁹

AGREE II = Appraisal of Guidelines for Research and Evaluation II; NICE = National Institute for Care and Health Excellence; PHAC = Public Health Agency of Canada; WHO = WHO.

Appendix 4: Main Study Findings and Authors’ Conclusions

Summary of Findings From Included Overviews of Reviews

Visser et al. (2018)¹¹

Main Study Findings

• One included SR

  • Patients with TB with or without HIV

  • Included interventions of oral nutritional supplement given for more than 4 weeks

  • Intervention: 60 days to 6 months long

  • Follow-up: 8 weeks to 1 year

• Quality of life

  • 2 primary studies with 134 participants

  • Conclusion: “supplementation may have increased QoL scores during the first 2 months of treatment (low-certainty evidence) (p. 17).”

• Nutritional status of adults

  • Weight gain (kg) increase at

    • 6 weeks: mean difference = 1.73 (95% CI, 0.81 to 2.65) (1 primary study with 34 participants)

    • 12 weeks: mean difference = 2.60 (95% CI, 1.74 to 3.46) (1 primary study with 100 participants)

    • 32 weeks: mean difference = 2.60 (95% CI, 0.52 to 4.68) (1 primary study with 265 participants)

  • Weight gain did not change in 1 primary study with patients co-infected with HIV

  • Conclusion: probably a modest increase in weight, but not consistent (moderate-certainty evidence)

• Illness (or disease-related outcomes)

  • No difference in disease outcomes, but studies underpowered for free food or high-energy supplements versus no supplementary feeding

    • Cure rate: RR = 0.91 (95% CI, 0.59 to 1.41) (1 primary study with 102 participants)

    • Sputum negative at 8 weeks results: RR = 1.08 (95% CI, 0.86 to 1.37) (3 primary studies with total of 222 participants)

  • One primary study found significant benefit for treatment completion and sputum conversion

  • Conclusion: very low-certainty evidence for all

• Mortality

  • Death at 1-year follow-up

    • RR = 0.34 (95% CI, 0.10 to 1.20)

    • No subgroup differences for patient with or without HIV

    • Conclusion: very low-certainty evidence

Authors’ Conclusion

“Mortality data were limited and underpowered in meta-analysis in all populations (children with MAM, in children with HIV, and in adults with tuberculosis) … In adults with tuberculosis, one small trial found a significant benefit on treatment completion and sputum conversion rate. There were also significant but modest benefits in terms of weight gain (up to 2.60 kg) during active tuberculosis (p. 2).”

Pantoja et al. (2017)¹² and Wiysonge et al. (2017)¹³

Pantoja et al. (2017)¹²and Wiysonge et al. (2017)¹³ included the same SR (Lutge et al. [2015]⁶). They also provided identical results for the review.

Main Study Findings

• One included SR

  • Patients receiving drug therapy to cure TB

  • Included 11 primary studies; 2 studies were on adherence to treatment

  • Comparisons included:

    • Immediate versus deferred incentives

    • Cash versus non-cash incentives

    • Different levels of cash incentives

    • Incentives versus other interventions

“Sustained material incentives may lead to little or no difference in cure or completion of treatment for active TB, compared to no incentive (p. 42).”

“Compared to a non-cash incentive, cash incentives may slightly increase the number of people who return to a clinic for reading of their tuberculin skin test and may increase the number of people who complete TB prophylaxis (p. 42).”

“Compared to counselling or education interventions, material incentives may increase the number of people who return to a clinic for reading of their tuberculin skin test (p. 42).”

“Compared to counselling or education interventions, material incentives may lead to little or no difference in the number of people who return to a clinic to start or continue TB prophylaxis or in the number of people who complete TB prophylaxis (p. 43).”

“Higher cash incentives may slightly improve the number of people who return to a clinic for reading of their tuberculin skin test, compared to lower cash incentives (p. 43).”

Authors’ Conclusion

“Compared to routine care, cash-and non-cash incentives probably increase health service utilization (return visits for tuberculin skin test reading, start or continuation of treatment) (low- to moderate-certainty evidence). They may not improve completion of TB prophylaxis (low-certainty evidence), and it is uncertain if they improve completion of treatment for active TB (very low-certainty evidence). Cash incentives may slightly improve patient return for tuberculin skin test reading and completion of TB prophylaxis compared to non-cash incentives (low-certainty evidence). Immediate (compared to deferred) incentives may not improve adherence to anti-tuberculosis treatment (low-certainty evidence) (p. 14).”

Summary of Findings From Included Systematic Reviews

Barss et al. (2020)¹⁴

Main Study Findings

N = 9 total primary studies

• Completion of initial assessments

  • Improved with financial and non-financial incentives: risk difference = 42 (95% CI, 34 to 51) additional patients completing initial assessment per 100 starting, 9 studies or cohorts, I² = 89%

  • Did not significantly improve with patient education: risk difference = 22 (95% CI, 6 to 49 per 100 people), 5 studies or cohorts, I² = 97%

• Completion of medical evaluations

  • improved significantly with patient incentives (risk difference = 48; 95% CI, 15 to 81 additional patients completing medical evaluation per 100 starting), 2 studies or cohorts, I² = 93%

• Acceptance of treatment

  • improved with patient incentives (risk difference = 49; 95% CI, 46 to 52) additional patients accepting treatment per 100 recommended), 1 study or cohort

  • improved with patient education (risk difference = 15; 95% CI, 11 to 19) additional per 100), 1 study or cohort

• One study included patient incentives but the results for incentives were not reported

Authors’ Conclusion

“Step 7: In single studies, patient incentives (49 [95% CI, 46–52] additional patients accepting treatment per 100 recommended) and patient education (15 [95% CI, 11–19] additional per 100) improved the rates of patient acceptance of LTBI [latent tuberculosis infection] treatment (p. 105).”

Law et al. (2019)¹⁵

Main Study Findings

N = 19

Losses to follow-up

• Financial support

  • Financial support offered in 12 studies, and included reimbursement of rent and travel expenses, and compensation of lost wages

  • Financial support associated with fewer losses to follow-up

  • Pooled proportions of patients lost to follow-up

    • Covering travel expenses (10 cohorts) was associated with less loss to follow-up: 0.15 (95% CI, 0.10 to 0.24; I² = 94%)

    • Covering rent and travel expenses (4 cohorts) associated with less loss to follow-up: 0.08 (95% CI, 0.06 to 0.10; I² = 0%)

    • Providing supplemental income (3 cohorts) associated with less loss to follow-up: 0.06 (95% CI, 0.00 to 0.61; I² = 96%)

    • No financial support (14 cohorts): 0.24 (95% CI, 0.17 to 0.34; I² = 95%)

    • Cochran Q test: P < 0.01

  • In patients who received twice daily or daily direct observed therapy (DOT) (proportion of patients lost to follow-up)

    • Covering travel expenses (8 cohorts) was associated with less loss to follow-up: 0.14 (95% CI, 0.07 to 0.24; I² = 94%)

    • Covering rent or travel expenses (4 cohorts) associated with less loss to follow-up: 0.08 (95% CI, 0.06 to 0.10; I² = 0%)

    • Providing supplemental (2 cohorts) associated with less loss to follow-up: 0.01 (95% CI, 0.00 to 0.68; I² = 88%)

    • No financial support (10 cohorts): 0.21 (95% CI, 0.14 to 0.30; I² = 94%)

    • Cochran Q test: P < 0.01

• Food packages

  • Food packages provided in 15 studies

  • Food package included food baskets, provisions of basic foods, hot meals, and more

    • weak evidence of an association with lower losses to follow-up

  • Pooled proportions of patients lost to follow-up

    • Food packages (14 cohorts) associated with less loss to follow-up: 0.15 (95% CI, 0.10 to 0.22I² = 95%)

    • No food packages: 0.18 (95% CI, 0.11 to 0.29; I² = 96%)

    • Cochran Q test: P = 0.51

  • In patients who received twice daily or daily DOT (proportion of patients lost to follow-up)

    • Food packages (13 cohorts) associated with less loss to follow-up: 0.13 (95% CI, 0.09 to 0.19; I² = 94%)

    • No food packages: 0.13 (95% CI, 0.06 to 0.27; I² = 97%)

    • Cochran Q test: P = 1.00

Authors’ Conclusion

“Additionally, provision of financial support to reimburse rent or travel expenses, as well as to compensate lost wages during treatment, was associated with fewer losses to follow-up. There was weak evidence of any association between providing food packages, group counselling or counselling to family members and losses to follow-up (p. 7).”

Riquelme-Miralles et al. (2019)⁴

Main Study Findings

N = 8

Completion (%)

• Food supplements and incentives

  • Active TB

    • 1 study (food supplements versus standard care) found benefit in adherence (98% versus 82%)

    • 1 study (food incentives versus standard care) found no benefit in adherence over 8 months (76% versus 78%)

• Education and economic incentives

  • Active TB

    • 1 study (economic and education incentive versus standard care) found no benefit over 6 months to 9 months (97.7% versus 91.1%)

  • Latent TB, patients who were homeless

    • 1 study (cash incentives versus non-cash incentives) found no benefit over 6 months (89.2% versus 81.5%)

    • 1 study (monetary incentives versus standard care) found benefit over 6 months (44% versus 26%, P not reported, but reported as not significant in other SRs)

  • Latent TB, inmates

    • 1 study (non-cash incentive versus standard care) found benefit over 6 months (23% versus 12%), but no benefit over 6 months for education versus standard care (12% versus 12%)

  • Latent TB, people who use drugs

    • 1 study (outreach plus incentive versus outreach alone and incentive alone versus outreach alone) found benefit for incentives and outreach and incentives alone over 6 to 12 months (52.8% versus 3.6% and 60% versus 3.6%, respectively)

  • Latent TB, adolescents

    • 1 study (peer counselling and incentive versus standard care and incentive contract versus standard care) found no benefit for either peer counselling plus incentive or incentives alone over 6 months (84.8% versus 77.8% and 76.4% versus 77.8%, respectively)

  • Latent TB, general

    • 1 study (education and economic incentives versus standard care) found benefit for incentives and outreach and incentives alone over 9 months to 12 months (63.8% versus 27.1%)

Authors’ Conclusion

“The studies found are in reality very different from each other. There is too much variability in studies on therapeutic adherence, both in the active tuberculosis and in the latent infection treatment groups, to be able to compare strategies for identifying interventions, objectives and effects. In addition, the designs generally have methodological flaws, preventing us from accurately determining which interventions we could apply in clinical practice for our patients. Accordingly, we encourage other authors to continue researching in this line, by developing new clinical trials, following the current recommendations that minimise the risk of bias, and all of this with a sample size that is adequate for its objective. Once several studies of this nature have been carried out, we will be in a position to reassess the clinical question posed in this systematic review (p. 459).”

Alipanah et al. (2018)¹⁶

Main Study Findings

N = 15 total

• Patient education (oral and written educational material) plus standard care versus standard care alone

  • 4 RCTs, 1 cohort study

  • Associated with a higher rate of treatment completion (1 RCT: RR = 1.71; 95% CI, 1.32 to 2.22)

  • Associated with a higher rate of treatment adherence (1 RCT: RR = 1.83, 95% CI, 1.14 to 2.92; 1 cohort study: RR = 1.21; 95% CI, 1.05 to 1.40)

  • Associated with a higher rate of cure (1 RCT: RR = 2.15, 95% CI, 1.58 to 2.92)

  • No significant effect on rates of mortality, treatment success, failure, or loss to follow-up

• Patient incentives and enablers plus standard care versus standard care alone

  • 4 RCTs, 11 cohort studies

  • Associated with lower rates of mortality (3 cohort studies: RR = 0.51; 95% CI, 0.37 to 0.71; 2 RCTs: RR = 0.93; 95% CI, 0.41 to 2.09)

  • Associated with lower rates of treatment failure (1 RCT: RR = 0.66; 95% CI, 0.50 to 0.87; 2 cohort studies: RR = 0.18; 95% CI, 0.02 to 2.10)

  • Associated with lower loss to follow-up (1 RCT: RR = 0.74; 95% CI, 0.60 to 0.90; 5 cohort studies: RR = 0.48; 95% CI, 0.28 to 0.81)

  • Associated with higher rate of treatment success (3 RCTs: RR = 1.07; 95% CI, 1.03 to 1.11; 4 cohort studies: RR = 1.25; 95% CI, 1.09 to 1.42)

  • Associated with higher rates of treatment completion (5 RCTs: RR = 1.23; 95% CI, 1.15 to 1.31; 4 cohort studies: RR = 1.18; 95% CI, 0.97 to 1.43)

  • Associated with higher rates of cure (4 cohort studies: RR = 1.13; 95% CI, 1.02 to 1.26; 1 RCT: RR = 0.92; 95% CI, 0.85 to 1.01)

  • Associated with higher rate of sputum conversion at 2 months (1 RCT: RR = 1.21, 95% CI, 1.02 to 1.43)

Authors’ Conclusion

“The addition of other adherence interventions to DOT, such as education (for staff or patients), material or psychological support, or reminder systems (including SMS technology and phone reminders), correlated with reduced rates of mortality and loss to follow-up and higher rates of treatment success and cure (p. 22).”

Liu et al. (2018)¹⁷

Main Study Findings

N = 7

• Incentive strategies

  • Monetary or cash incentives and non-cash incentives significantly increased completion of treatment in some studies

  • Homeless adults (2 studies):

    • Cash incentive versus non-cash: 89.2% versus 81.5%

    • Cash incentive versus none: adjusted OR = 1.94 (95% CI, 0.65 to 5.83; P = 0.24)

    • Completion with monetary incentive versus none: OR = 2.57 (95% CI, 1.11 to 5.94)

    • Monetary incentive versus peer health advisor only versus usual care: 44% versus 19% versus 26%; P = 0.02 for monetary incentive versus peer health and P = 0.11 for monetary incentive versus usual care

  • People who use drugs (2 studies):

    • Immediate incentive versus deferred incentive: 83% versus 75% (P = 0.09)

    • Monetary incentive versus outreach alone: adjusted OR = 45.5 (95% CI, 9.7 to 214.6); monetary incentive plus outreach: adjusted OR = 29.7 (95% CI, 6.4 to 137.5)

  • Inmates (1 study):

    • Incentive versus none: OR = 1.07 (95% CI, 0.47 to 2.40)

• Education

  • Inmates (1 study):

    • Incentive versus none: OR = 2.2 (95% CI, 1.04 to 4.72)

• Results from 4 studies regarding incentives (cash or non-cash) were not reported

Authors’ Conclusion

“Incentive strategies, including cash or monetary incentives and noncash incentives (e.g., toys for children, free lunch, grocery store coupons, and phone cards or bus tokens), significantly increased completion rates among LTBI patients. For example, Tulsky et al. found an 18% increase in completion among homeless adults in the United States. Chaisson et al. found that patients receiving immediate incentives had higher completion rates than patients receiving deferred incentives among drug users (83% vs 75%) (p. e427).”

Muller et al. (2018)¹⁸

Main Study Findings

N = 7

• Education and counselling

  • Patient education/counselling versus no education/counselling

    • Cure rates: 2 RCTs (1,106 patients), patient education or counselling led to better cure rates (RR = 1.16; 95% CI, 1.05 to 1.29; P = 0.004; I² = 0%)

    • Default rate: 2 RCTs, decreased in patients receiving education by 13% (RR = 0.87; 95% CI, 0.77 to 0.98; P = 0.03; I² = 0%)

• Food incentives

  • Food incentives versus no food incentives

    • Cure rates: 3 RCTs found no significant difference (434 patients: RR = 1.07; 95% CI, 0.95 to 1.21; P = 0.27; I² = 50%)

• Financial incentives

  • Financial incentives versus no financial incentives

    • Cure rates: 2 RCTs, no significant difference between the use of financial incentives and no financial incentives (4,214 patients, RR = 1.00; 95% CI, 0.81 to 1.23; P = 0.99; I² = 67%)

    • Mortality: 2 RCTs, patients receiving financial incentive versus none (RR = 1.02; 95% CI, 0.82 to 1.27; P = 0.85) (Note: RR was written as “1.2” in text, but “1.02” in forest plots)

    • Default rate: 2 RCTs, decreased by 26% for patients receiving financial incentive (RR = 0.74; 95% CI, 0.61 to 0.90; P = 0.002; I² = 0%)

• The quality of evidence for cure rates, default rates, and mortality was low

Authors’ Conclusion

“In addition, the default rate decreased by respectively 49%, 26% and 13% with DOTS, financial incentives and patient education and counselling. There was no significant reduction in mortality rates with the use of these interventions. Assuming an appropriate drug regimen is prescribed, treatment success depends largely on the patient’s adherence to the regimen. Without adequate support, a significant proportion of patients with TB discontinue treatment before the end of the planned period or take medication irregularly (p. 737).”

Richterman et al. (2018)¹⁹

Main Study Findings

N = 8

• Meta-analysis

  • Likelihood of treatment success with cash incentives versus usual care

    • 4 studies meta-analyzed: OR = 1.77 (95% CI, 1.57 to 2.01; I² = 0%)

• Non–meta-analyzed studies

  • Treatment success

    • Monthly cash equivalent to low civil service salary versus usual care: OR = 1.19 (95% CI, 1.03 to 1.37)

  • Treatment completion

    • Cash transfer for transport, poverty reduction, and other costs versus usual care: OR = 3.28 (95% CI, 1.65 to 6.51)

  • Microbiologic cure

    • Monthly cash transfer and travel reimbursement versus usual care: OR = 79.08 (95% CI, 4.42 to 1,413.33)

    • Monthly cash to female head of household versus usual care: OR = 1.07 (95% CI, 1.04 to 1.11)

  • Weight gain

    • Monthly cash transfer and travel reimbursement versus usual care: 10.4 lbs versus 1.7 lbs (P not reported)

  • Return to work after 1 year

    • Monthly cash transfer and travel reimbursement versus usual care: 93% versus 47% (P not reported)

  • Mortality

    • Monthly cash transfer and travel reimbursement versus usual care: 0% versus 10% (P not reported)

    • Monthly cash transfer equivalent to median direct cost for tuberculosis care versus usual care: 7% versus 6% (P not reported)

    • Monthly cash, cash at treatment completion, transport reimbursement versus usual care: 5% versus 6%, (P not reported)

    • Cash transfers throughout treatment, approximately 10% household income versus usual care: 4% versus 4% (P not reported)

  • Treatment failure

    • Monthly cash, cash at treatment completion, transport reimbursement versus usual care: 2% vs 5% (P not reported)

  • Loss to follow-up

    • Monthly cash, cash at treatment completion, transport reimbursement versus usual care: 5% vs 10% (P not reported)

    • Monthly cash transfer equivalent to median direct cost for tuberculosis care versus usual care: 5% vs 20% (P not reported)

  • Sputum positivity after 6 months

    • Monthly cash transfer and travel reimbursement versus usual care: 0% versus 13% (P not reported)

  • Negative smear at 2 months

    • Monthly cash transfer equivalent to median direct cost for tuberculosis care versus usual care: 88% versus 92% (P not reported)

Authors’ Conclusion

“In conclusion, we found some evidence that cash transfer interventions improve treatment outcomes in patients with active pulmonary tuberculosis in low- and middle-income countries, although the overall quality of this evidence is low. These findings support calls by WHO and others to incorporate cash transfer interventions into social protection schemes within tuberculosis treatment programmes (p. 480).”

Herrmann et al. (2017)²⁰

Main Study Findings

N = 8

• Education and financial incentives (≥ 33% population are people who use drugs)

  • Inmates:

    • 1 RCT, education plus US$5 for visit versus education alone: 26% attendance versus 23% attendance (P = 0.82)

    • 1 RCT, informational session plus US$25 vouchers for food or transport if attending TB clinic within 1 month of release versus informational session alone versus control: 37% (incentive group) versus 37% (education group) versus 24% (control group) attendance, 12% (incentive group) versus 23% (education group) versus 12% (control group) treatment completion (P = NR) (Note: Herrmann et al²⁰ originally wrote this as 23% completion in the control group and 12% completion in incentive group. This was erroneously reported upon review of the original primary study)

  • People who are homeless:

    • 1 RCT, US$5 per dose (observed) versus no money (self-administered): 44% versus 26% (P = 0.11)

    • 1 RCT, US$5 per dose versus US$5 non-cash incentive per dose: 89% completion versus 81% completion (P = 0.23)

  • General population:

    • 1 cohort study, US$15 in subway tokens per week for all doses ingested versus US$15 in subway tokens per week for all doses ingested and monthly bonuses of US$30 to US$60 for 80% or more of doses taken in the month: 2.7 times more likely to take 80% or more of doses with bigger incentive package

  • People who inject drugs:

    • 1 RCT, US$10 a month for adherence paid once a month for 6 months versus US$10 per month for adherence paid out as a lump sum after 6 months: 84% completion versus 75% (P = 0.09)

    • 1 RCT, DOT plus US$5 per dose versus DOT alone: 53% completion versus 4% (P = NR)

  • Adults (56% with substance use disorder) or children with medication nonadherence:

    • 1 cohort study, US$5 grocery gift card plus DOT or physician visit versus DOT alone: 5.7 times more likely to complete treatment with incentive, 60% completion versus 19% (P = NR)

Authors’ Conclusion

“In summary, the present review demonstrates that there is compelling evidence that incentive-based interventions improve adherence to vaccinations, diagnostic tests and pharmacotherapies critical for the control of hepatitis, HIV and TB among individuals with SUDs [substance use disorders]. The parameters that moderate the efficacy of these interventions appear consistent with those shown to influence outcomes of CM [contingency management] for the treatment of SUDs. Incentives are a valuable tool that can be used to improve public health outcomes related to infectious disease (p. 10-11).”

Heuvelings et al. (2017)²¹

Main Study Findings

N = 3

• Screening adherence

  • People who are homeless:

    • 1 RCT, incentives versus usual care: “improves tuberculosis… screening (p. e177)”

    • From original NICE review³²: 84% versus 53% (OR = 4.7; 95% CI, 2.2 to 9.8)

• Tuberculosis management

  • People who use drugs:

    • DOT plus incentives versus DOT alone: “DOT increases successful treatment outcomes and improves treatment adherence among several hard-to-reach populations, especially when combined with incentives (p. 3155)”

    • From original NICE review³²: at 32 weeks, 60% versus 19% (OR = 5.73; 95% CI, 2.25 to 14.84)

    • From original NICE review³²: at 52 weeks, 89% versus 52% (OR = 7.29; 95% CI, 2.45 to 22.73)

  • Migrants, homeless people, people who use drugs, people with HIV:

    • DOT plus incentives versus DOT alone: “DOT increases successful treatment outcomes and improves treatment adherence among several hard-to-reach populations, especially when combined with incentives (p. e155)”

    • From original NICE review³²: 75.2% versus 26.7% (RR = 3.069; 95% CI, 2.133 to 4.414; P < 0.0001)

Authors’ Conclusion

“The NICE review found that DOT increases successful treatment outcomes and improves treatment adherence among several hard-to-reach populations, especially when combined with incentives (p. e155).”

Stuurman et al. (2016)²²

Main Study Findings

N = 4

• People who inject drugs:

  • Higher completion rates for patients receiving monetary incentives versus no incentive: adjusted OR = 32.0 (95% CI, 7.1 to 145); moderate quality

  • Higher completion rates for patients receiving methadone treatment and substance use disorder counselling versus no incentive: OR = 14.5 (95% CI, 5.0 to 42); very low quality

• Inmates:

  • Food or transport vouchers if attending clinic 1 month after release versus no vouchers: OR = 1.07 (95% CI, 0.5 to 2.4), moderate quality

• People who are homeless

  • Cash incentives versus non-cash incentives: OR = 1.7 (95% CI, 0.7 to 4.3); low quality

Authors’ Conclusion

“Overall, however, the evidence was inconclusive and recommendations on the best interventions to improve uptake of LTBI medication are hampered by the heterogeneity of the studies. The benefit of interventions to improve treatment completion, such as incentives and DOT, appears to be population and setting dependent. Specific needs of the different populations with LTBI should be addressed taking into consideration the local context, specific settings and conditions in which the LTBI treatment programme is implemented (p. 15).”

Van Hoorn et al. (2016)²³

Main Study Findings

N = 21

• Meta-analysis: treatment success (9 RCTs)

  • Psycho-emotional supports

    • 3 studies: pooled RR = 1.37 (95% CI, 1.08 to 1.73; I² = 78%)

    • Omission of 1 study with high risk of bias: RR = 1.20 (95% CI, 1.07 to 1.35; I² = 0%)

  • Socioeconomic supports

    • 4 studies: pooled RR = 1.08 (95% CI, 1.03 to 1.13; I² = 14.4%)

  • Combined supports

    • 3 studies: pooled RR = 1.17 (95% CI, 1.12 to 1.22; I² = 0%)

  • Overall

    • Pooled RR = 1.17 (95% CI, 1.09 to 1.25; I² = 72.8%)

  • All favoured intervention over control

• Meta-analysis: unsuccessful treatment outcomes (9 studies)

  • Psycho-emotional supports

    • 4 studies: pooled RR = 0.46 (95% CI, 0.22 to 0.96; I² = 85%)

    • Omission of 1 study with high risk of bias: RR = 0.33 (95% CI, 0.22 to 0.50; I² = 0%)

  • Socioeconomic supports

    • 2 studies: pooled RR = 0.78 (95% CI, 0.69 to 0.88; I² = 0%)

  • Combined supports

    • 4 studies: pooled RR = 0.42 (95% CI, 0.23 to 0.75; I² = 64%)

  • Overall

    • Pooled RR = 0.53 (95% CI, 0.41 to 0.70; I² = 80.2%)

• Treatment adherence (3 RCTs)

  • Socioeconomic supports

    • 1 study: RR = 1.01 (95% CI, 0.85 to 1.33)

  • Combined supports

    • 1 study: RR = 1.11 (95% CI, 0.92 to 1.33)

• Non–meta-analyzed data

  • Studies excluded from meta-analysis

    • Treatment adherence

      • Historically controlled study, indirect economic support versus usual care: 32 weeks or less OR = 5.73 (95% CI, 2.25 to14.84); 52 weeks or less OR = 7.29 (95% CI, 2.45 to 22.73)

      • Case-control study, indirect economic support versus usual care: The odds patient will adhere is 2.7 times as great as person receiving the basic incentive package

      • Before-and-after study, direct economic support versus usual care: default rates reduced (11% versus 1%; P = 0.03) in favour of intervention

  • Non-randomized studies

    • 7 non-randomized studies reported some effect of socioeconomic supports on treatment success (RR range from 1.03 to 2.51; 95% CI, 0.96 to 2.99); 5 of 7 studies reported significant RR in favour of intervention

    • 6 non-randomized studies reported some effect of socioeconomic supports on treatment failure (RR range from 0.32 to 0.96; 95% CI, 0.18 to 3.49); 5 of 6 studies reported significant RR in favour of intervention

    • 2 case-control studies reported significant beneficial effects of socioeconomic supports on treatment failure (RR = 0.51; 95% CI, 0.37 to 0.70 and RR = 0.10; 95% CI, 0.05 to 0.20)

Authors’ Conclusion

“This review found that PE [psycho-emotional] and SE [socioeconomic] support did improve treatment outcomes across a variety of settings and patient populations, with a tendency towards better outcomes with PE interventions or a combined approach. However, the quality of evidence was classified as “very low” under the GRADE approach. Food supplementation and counselling were commonly included in the package of support. PE, SE and combined interventions improved treatment outcomes; only for interventions including SE support exclusively there was no significant improvement in treatment success. Overall, support interventions were associated with significantly higher treatment success (overall RR 1.08; CI 1.03 to 1.13) and reductions in unsuccessful treatment outcomes (overall RR 0.53; CI 0.41 to 0.70) (p. 21).”

Appendix 5: Overlap Between Included Systematic Reviews

Table 9: Overlap in Relevant Systematic Reviews Between Included Overviews of Reviews

Appendix 6: Further Information

Systematic Reviews Included Within Overviews of Reviews

Grobler L, Nagpal S, Sudarsanam TD, Sinclair D. Nutritional supplements for people being treated for active tuberculosis. Cochrane Database Syst Rev. 2016(6):CD006086. Medline

Lutge EE, Wiysonge CS, Knight SE, Sinclair D, Volmink J. Incentives and enablers to improve adherence in tuberculosis. Cochrane Database Syst Rev. 2015(9):CD007952. Medline

Excluded Primary Studies (Due to Volume of Literature)

Cocozza AM, Linh NN, Nathavitharana RR, et al. An assessment of current tuberculosis patient care and support policies in high-burden countries. Int J Tuberc Lung Dis. 2020;24(1):36-42. Medline

Rohit A, Kumar AMV, Thekkur P, et al. Does provision of cash incentive to HIV-infected tuberculosis patients improve the treatment success in programme settings? A cohort study from South India. J. 2020;9(8):3955-3964.

Watthananukul T, Liabsuetrakul T, Pungrassami P, Chongsuvivatwong V. Effect of Global Fund financial support for patients with multidrug-resistant tuberculosis. Int J Tuberc Lung Dis. 2020;24(7):686-693. Medline

Benzekri NA, Sambou JF, Tamba IT, et al. Nutrition support for HIV-TB co-infected adults in Senegal, West Africa: A randomized pilot implementation study. PLoS ONE. 2019;14(7):e0219118. Medline

Bhatt R, Chopra K, Vashisht R. Impact of integrated psycho-socio-economic support on treatment outcome in drug resistant tuberculosis - A retrospective cohort study. Indian J. 2019;66(1):105-110. Medline

Carter DJ, Daniel R, Torrens AW, et al. The impact of a cash transfer programme on tuberculosis treatment success rate: a quasi-experimental study in Brazil. BMJ glob. 2019;4(1):e001029.

Kim H, Choi H, Yu S, et al. Impact of Housing Provision Package on Treatment Outcome Among Homeless Tuberculosis Patients in South Korea. Asia Pac J Public Health. 2019;31(7):603-611. Medline

Klein K, Bernachea MP, Irribarren S, Gibbons L, Chirico C, Rubinstein F. Evaluation of a social protection policy on tuberculosis treatment outcomes: A prospective cohort study. PLoS Med. 2019;16(4):e1002788. Medline

Yuen CM, Millones AK, Contreras CC, Lecca L, Becerra MC, Keshavjee S. Tuberculosis household accompaniment to improve the contact management cascade: A prospective cohort study. PLoS ONE. 2019;14(5):e0217104. Medline

Mansour O, Masini EO, Kim BJ, Kamene M, Githiomi MM, Hanson CL. Impact of a national nutritional support programme on loss to follow-up after tuberculosis diagnosis in Kenya. Int J Tuberc Lung Dis. 2018;22(6):649-654. Medline

Priedeman Skiles M, Curtis SL, Angeles G, Mullen S, Senik T. Evaluating the impact of social support services on tuberculosis treatment default in Ukraine. PLoS ONE. 2018;13(8):e0199513. Medline

Verdecchia M, Keus K, Blankley S, et al. Model of care and risk factors for poor outcomes in patients on multi-drug resistant tuberculosis treatment at two facilities in eSwatini (formerly Swaziland), 2011-2013. PLoS ONE. 2018;13(10):e0205601. Medline

Samuel B, Volkmann T, Cornelius S, et al. Relationship between Nutritional Support and Tuberculosis Treatment Outcomes in West Bengal, India. J. 2016;4(4):213-219.

Kliner M, Canaan M, Ndwandwe SZ, et al. Effects of financial incentives for treatment supporters on tuberculosis treatment outcomes in Swaziland: a pragmatic interventional study. Infect. 2015;4:29. Medline