Testing and Scaling Interventions to Improve the Tuberculosis Infection Care Cascade

Abstract

Tuberculosis (TB) preventive therapy (TPT) is increasingly recognized as the key to eliminating tuberculosis globally and is particularly critical for children with TB infection or who are in close contact with individuals with infectious TB. But many barriers currently impede successful scale-up to provide TPT to those at high risk of TB disease. The cascade of care in TB infection (and the related contact management cascade) is a conceptual framework to evaluate and improve the care of persons who are potential candidates for TPT. This review summarizes recent literature on barriers and solutions in the TB infection care cascade, focusing on children in both high- and low-burden settings, and drawing from studies on children and adults. Identifying and closing gaps in the care cascade will require the implementation of tools that are new (e.g. computer-assisted radiography) and old (e.g. efficient contact tracing), and will be aided by innovative implementation study designs, quality improvement methods, and shared clinical practice with primary care providers.

INTRODUCTION

In a follow-up to the Global Plan to Stop TB 2006–2015, the World Health Organization (WHO) launched the End TB Strategy in 2015 [1]. The targets for the Strategy were to reduce tuberculosis (TB) incidence by 90%, TB mortality by 95%, and catastrophic costs for TB patients by 100% by 2035. A key component was the treatment of TB infection, or TB preventive treatment (TPT). But, given that approximately one-quarter of the entire world’s population is estimated to carry latent TB infection [2], priority must be given to those at the highest risk of developing TB: persons living with HIV (PLHIV) and close contacts of persons with TB disease. In 2018, the United Nations General Assembly adopted a resolution to identify and treat 30 million persons with TB infection by the end of 2022 [3]. This resolution specified the treatment of 6 million PLHIV; this target has already been surpassed [4]. However, progress has been much slower toward the target of providing TPT to 4 million children under five years old. Despite the very high risk of disease in this age group [5–7], WHO estimated that by the end of 2021, only 1.2 million child contacts <5 years old had been treated, which is only 29% of the target with only one year remaining [4]. Progress towards the third target—of treating 20 million close contacts ages 5 and older (which includes many older children and adolescents)—has been abysmal: WHO estimated that less than 2% of this target had been achieved by the end of 2021 [4]. Successful interventions to improve the number of individuals receiving TPT have been found, but scaling those interventions to close the global gap in TB preventive services is a daunting challenge.

The Cascade of Care in TB Infection

The cascade of care concept was initially developed for HIV care and first proposed for the care of TB infection in a systematic review published by Alsdurf and colleagues in 2016 [8]. As shown in Figure 1, this cascade can be conceptualized as six steps: (1) identification of those eligible for latent TB treatment and initial entry to care, usually with testing for TB infection; (2) obtaining a test result and referral for further evaluation; (3) medical evaluation including chest radiography; (4) offering TPT; (5) accepting and starting TPT; and (6) completing TPT. Three subsequent systematic reviews have examined the cascade of TB infection care in children [9], child contacts in low- and middle-income countries [10], and PLHIV [11].

Figure 1.

Illustration of cumulative retention and losses at major steps in the cascade of care in TB infection. This figure is for illustration, and does not present real data.

Abbreviations: TPT—tuberculosis preventive therapy

All four systematic reviews found that losses occur at all steps of the cascade, particularly at the steps of entry to care and initial testing, medical evaluation including chest radiographs, and acceptance and starting TPT [8–11]. As summarized in Table 1, barriers to completing identification, diagnosis, and treatment steps in the cascade can be grouped into barriers at the health system, provider, and individual patient levels. All reviews have concluded that most attrition occurs prior to starting TPT. This means that more of the potential public health benefit of TPT is lost due to pre-treatment losses than due to non-adherence to TPT once started. However, pre-treatment losses are “invisible” to most health workers and researchers.

Table 1.

Barriers and Enablers/Solutions to Problems in the TB Infection Care Cascade

Level	Barriers	Possible Enablers/Solutions
Health System	Under-funded TB programmes [12]	Alternative funding mechanisms (e.g. public/government health insurance)
	Inadequate access to TB infection diagnostics [13–16]	Global drug facility for tuberculin and IGRA test kits
	Inadequate CXR facilities/interpretation [13, 14, 17]	Digital CXR;CXR interpretation software (e.g. CAD4TB)
	Inadequate access to TPT regimens [18, 19]	Global drug facility
	Poorly organized services/barriers in the referral process [12, 20, 21]	Cascade analyses
		Integrate TPT into primary care services
		Coordinate services and “One-stop TPT”
Providers	Knowledge gaps regarding testing and TPT (especially with changing recommendations) [22, 23]	Provider education
		On-line training and decision supports (e.g.TSTin3D, Periscope-TB)
	No incentives	Reimbursement for TB infection care services
	No tools for TB infection care	Quality improvement-based design and scale of tools (e.g. TB infection care registries)
Patients/Families	Knowledge gaps regarding testing and TPT [24–26]	Patient/family education material
		Social media to raise TB awareness and de-stigmatize TB
	Costs and structural barriers for care [26, 27]	Government payment for TB infection care costs
		Reimbursement for travel costs
		Integrate TB infection care into primary care services
	Time for care	Home visits, school-based care, integrate, and coordinate care (“One stop TPT”)
	Child/adolescent medication adherence [28]	Shorter TPT regimens; adherence support tools

Abbreviations: CXR—chest X-radiograph; IGRA—interferon gamma release assay; TPT—tuberculosis preventive therapy

The systematic reviews identified risk factors for losses at steps of the cascade prior to TPT initiation. Many of these barriers concentrate on the first step of the cascade: the identification of at-risk children who may benefit from TPT. Contact-tracing is the mainstay of pediatric TB prevention efforts globally [29]. However, well-documented health system barriers in contact identification and investigation contribute to substantial policy-practice gaps in contact identification and management [10, 30]. A recent modeling study estimated that full implementation of household contact case management would prevent approximately 160,000 cases of TB in children and adolescents annually, compared to no contact case management [31]. Recent data suggest that community transmission may be an important driver of pediatric TB [32]. However, given the current sub-optimal household contact management, we feel that the first priority is to optimize household contact tracing, and to initiate contact tracing outside of the home as a second step, as the latter will be resource-intensive.

For those children and adolescents who are identified as being at risk, additional health system barriers to diagnosis and referral include test stock-outs [13–16] and other inadequate diagnostic resources and facilities [13, 14, 17], limited clinical services leading to long travel distances, transitions in care setting [20], inconvenient hours of operation [12], and insufficient clinic staffing [21] leading to overcrowding and long waiting times. The most important barriers for providers are lack of experience and knowledge gaps of the benefits and risks of TB infection diagnosis and treatment, particularly for child contacts and other groups at high risk of TB infection and disease [22, 23]. These provider-level barriers are exacerbated by the de-prioritization of LTBI services [21, 33]. For patients and their families, the major barriers related to poor understanding of the risks of TB and benefits of prevention [24–26], and costs, resources, and time needed for TB infection testing and treatment [26, 27].

Patients’ age and development also have important effects on patients’ and caregivers’ engagement in care. For instance, recent studies have documented lower rates of appropriate TB testing among older children and adolescents than among younger children [26, 34, 35]. Age-based differences in the completion of cascade steps may in part be due to changing decision-making role of caregivers and the autonomy of children as children age [36]. For example, acceptance of testing and therapy differs between young children—for whom caregiver perceptions about TPT primarily affect TPT uptake [37]—and older children/adolescents—for whom patients’ autonomous decisions about their care have a larger influence over willingness to start treatment [38]. Further studies on how children’s age and changing autonomy facilitate or impede completion of cascade steps prior to TPT initiation will be important to tailor TB preventive efforts, including diagnostic testing for TB contacts (see below).

There have been numerous observational studies describing rates and factors associated with TPT adherence and completion. Challenges with adherence to TPT are common among asymptomatic older children and adolescents who may be reluctant to take prolonged courses of medication, as reviewed elsewhere in this supplement [28]. A review of TPT adherence among adults in the US and Canada identified a limited number of factors relevant to children and adolescents, such as substance use and housing instability, that were consistently associated with non-adherence [39]. However, few personal or clinical characteristics were found to be consistently associated with treatment adherence or completion, although many of the studies included in the review were observational studies based on routinely collected data which limited their ability to assess all potentially relevant barriers [39]. In contrast, incentives and enablers have been consistently associated with better completion, including in studies among children and adolescents [40–42]. The major emphasis in TPT trials in children and adults over the last three decades has been the development of shorter regimens with the primary objective of improving completion. Several randomized controlled trials [43, 44], including two large open-label trials in children and adolescents [45, 46], have demonstrated that three to four months of rifamycin-based regimens yield significantly better completion rates than 6–12 months of isoniazid, with similar or better safety and comparable efficacy. Several of these shorter regimens are now recommended by WHO [47] and authoritative agencies in several countries [43, 48, 49]. Yet a better understanding of adherence and shorter regimens will have minimal public health benefit if most eligible patients do not even start TPT.

Controversies in the TB Infection Cascade of Care

The TB infection cascade is complex in part because of the need for tests for TB infection to identify who is eligible for TPT, and chest radiography to exclude TB disease before starting TPT. There are numerous barriers to testing for TB infection, whether using tuberculin skin tests (TST) or interferon gamma release assays (IGRA) [21]. There are also considerable difficulties obtaining chest radiographs, including very high equipment costs, shortages of materials and supplies, and limited personnel and technical expertise to perform and interpret radiographs [17, 50]. For decades the use of chest radiographs for the diagnosis of TB disease was discouraged as part of the DOTS strategy [51]. As a result, this service is not covered by TB programs in many low- and middle-income countries, meaning patients must pay. The high cost of chest radiography makes this test unaffordable and inaccessible for most TB patients and their contacts [52]. As a result of these barriers, the WHO has recommended that tests for TB infection and chest radiographs can be “skipped” in the TB infection care cascade so that more at risk individuals can be placed on TPT [53].

There is less debate about the risks and benefits of skipping these tests for PLHIV and child contacts under 5 years old in high-burden settings, given these patients’ high risks of development of disease if infected with TB and programmatic barriers associated with testing [54]. However, for other risk groups, particularly adults, in whom TPT may cause serious adverse events, there is much more debate regarding the need for TB infection testing and chest radiography. There is good evidence that the risk of disease is substantially higher in those with positive TST or IGRA than those with negative tests. A recent systematic review estimated that the relative risk of disease in PLHIV who had not received TPT was 11 times higher in those who were TST-positive or IGRA-positive compared to those with negative tests [55]. Similarly, the relative risk of disease was 6–7 times higher in contacts with positive than negative tests [55]. Two systematic reviews [56, 57] and subsequent large-scale randomized trials [58] found that there was a significant reduction of TB disease with isoniazid preventive therapy in persons with positive TB infection tests. But there was no significant reduction in those with negative tests in the same trials. Most of these trials involved PLHIV who were not on antiretroviral therapy. In summary, there is strong evidence that TB infection tests identify persons who will benefit most from TPT. Treatment of all persons at risk without testing for TB infection will subject many persons to the risk of TPT, without any benefit [59]. In child contacts, the risks of TPT are minimal, but in adults, particularly older adults, the risks of TPT are substantial. If there are no benefits, then TPT should be avoided, if possible.

Prior to initiating TPT, TB disease must be excluded. WHO has suggested that symptom screening is adequate to exclude TB disease in high-risk persons [60]. Among PLHIV who are not on antiretroviral therapy, the sensitivity of symptom screening to detect TB disease is about 80% [61], but the sensitivity of symptoms screening is as low as 25% in non-HIV infected adults with TB infection [62], and only 50% in PLHIV on ARVs [60]. A recent Cochrane review of TB screening strategies in children found that pooled sensitivity estimates of three different symptom-based screening tools varied between 61% and 89% using a clinical–microbiological composite reference standard; specificity varied between 37% and 94% [63]. The authors of this review noted likely overestimation of these estimates and identified a lack of high-quality prospective evidence supporting these tools’ use. Because these tools lead to frequent false-positive tests, they do not obviate the need for additional diagnostic testing capacity. Additionally, given ongoing uncertainty about the sensitivity of symptom screens, particularly in children, reliance on symptom screening alone may result in individuals with TB disease receiving TPT; this increases the risk of development of drug resistance [64]. Two systematic reviews of trials in adults have concluded that isoniazid [65] and rifamycin-based [66] TPT do not create drug resistance. This is reassuring, but it is important to remember that in all the randomized trials included in these two reviews, all participants underwent clinical evaluation and chest radiographic screening prior to beginning TPT. Hence, the reassurance that TPT does not create drug resistance can only be given in populations who have undergone clinical evaluation and chest radiographic screening prior to beginning TPT. There is no doubt that pragmatic issues continue to hinder the use of chest radiography [67]; however, access to high-quality radiography and clinical evaluation is needed to optimize the selection of patients for preventive treatment.

Does testing for TB infection and performance of chest radiography substantially alter the cascade of TB infection care? A recent systematic review of TPT given to PLHIV found no difference in the proportion of PLHIV who initiated or completed TPT in cohorts that underwent testing for TB infection and chest radiography compared to cohorts in which these tests were not performed [11]. From a programmatic perspective, this finding suggests that requirements for testing did not result in lower proportions of persons successfully treated with TPT. Notably, this review included only 3 studies that focused exclusively on children; requirements for chest radiography and associated transportation costs and logistical barriers have been found to impede the initiation of TPT in children in some studies [68, 69]. While findings from the systematic review add weight to policies requiring radiography prior to TPT initiation, cost, convenience, and quality barriers to radiography in children must continue to be identified and addressed.

A randomized trial is ongoing to assess the need for and barriers created by TST and chest radiography [70]. In this trial, household contacts ages 5–50 years old are randomized to investigation with TST and chest radiography (control or standard arm), to an arm in which chest radiography is replaced by GeneXpert, or a third arm in which TST is not done, so all contacts undergo chest radiography only. This study will help determine the patient and health system benefits, risks, and costs of these different TB infection care strategies for household contacts.

Solutions to Improve Retention in the TB Infection Care Cascade

The most commonly assessed interventions to improve retention in the care cascade are shorter TPT regimens, which have been consistently associated with better TPT completion [44, 71]. Additional interventions targeting TPT adherence and completion in children and adolescents have included incentives [40, 41, 72] and counseling and coaching behavioral interventions [42, 73, 74]. However, these interventions directed to TPT completion will not ameliorate barriers and losses that occur before starting treatment.

Relatively few interventional studies among children have investigated strategies to mitigate losses in the cascade before TPT initiation. In high-income settings, these have involved evaluations of targeted testing strategies [75, 76] and behavioral interventions to minimize loss to follow-up after TST [27]. More recently, implementation studies have employed multilevel interventions and a cascade framework to mitigate losses prior to and during TPT, targeting both children and adults [12, 69, 77–79]. For instance, a study in Peru attempted to improve medical evaluation and TPT completion in contacts through interventions designed to improve testing uptake, treatment accessibility, and physician knowledge; completion of the cascade was higher in children and adolescents than in adults, though overall completion was low [77]. Importantly, this study and other recent implementation programs have been informed by detailed input from stakeholders both to design interventions and to understand acceptability, often through qualitative or mixed methods research [79–82].

A major cluster-randomized intervention trial conducted in 6 countries and enrolling children and adults demonstrated the importance of locally tailored strategies to address losses at earlier steps of the cascade [79]. In this trial, the development of locally adapted interventions began with a detailed cascade analysis at participating health care facilities. Simultaneously, key stakeholder input was solicited to understand why losses occurred and identify locally appropriate and feasible solutions, which varied by setting. For example, in Ghana, community meetings were held to reduce stigma, and improved organization of TB preventive services reduced clinic waiting time [12]; in Indonesia, text message reminders were introduced for parents, and toys were given as incentives for children; in Vietnam, the National Health Insurance plan agreed to reimburse providers for patient care activities related to TB infection diagnosis and TPT. Based on the principles of quality assurance cycles, in which problems in care are identified, solutions implemented, then care is re-assessed [83], cascade analyses were repeated frequently to monitor progress and identify new or unresolved problems. The trial’s complex public health intervention was associated with significant increases in the number of household contacts identified and initiated on TPT in all participating low- and middle-income countries. Figure 2 provides an example of results in one country, in which high losses from the cascade occurred at baseline (checkered bars), but completion substantially improved after interventions were implemented (solid bars). Overall, this trial demonstrated how locally and iterative-adapted interventions can reduce losses from the care cascade—including losses at steps prior to TPT initiation—and can be scaled in disparate settings [79].

Figure 2.

Cascade analysis before and after multiple interventions to improve TB infection care. Data courtesy of Dr J. Obeng.

Abbreviations: TPT—tuberculosis preventive therapy

Scale-up in Different Countries and Settings

TB programmes in most countries are chronically underfunded and understaffed [84], a situation that has been exacerbated by the COVID-19 pandemic [53]. Hence, we do not judge it feasible for TB programmes to undertake the massive scale-up of TB infection diagnosis and treatment that is needed to achieve EndTB and United Nations General Assembly targets. Successful scaleup will require shared responsibility for TB infection care with primary care providers in all countries. For instance, “flexible” frameworks have been proposed for children and adolescents that involve patient/family-centered decision-making to guide selecting optimal settings and strategies for testing and TPT [85]. This would encourage greater integration of care; non-integration of care has been identified as a barrier to TB infection care in many settings [8, 9, 11]. Integrating care in primary care settings in turn will require simplified algorithms for testing, improved access to testing for TB infection and disease, and support for medical decision-making (including in-line test interpretation tools/risk calculators, such as TSTin3D [for adults; available at http://www.tstin3d.com/] and Periskope-TB [for adults and children; available at http://periskope.org/] [86]), as well as access to TPT regimens that are simple, short, and safe. Safety of TPT is critical to scale-up because safe regimens will enhance acceptance by providers, patients, and caregivers, and will simplify follow-up by obviating the need for close surveillance. Self-administered regimens will be preferred if primary care providers do not have access to the resources needed for directly observed TPT.

The most important aspect of successful scale-up is financing because massive scale-up will be costly, even if cost-effective and potentially cost-saving in the long term [87]. For TB infection care to be transferred to primary care, there must be mechanisms for reimbursement for diagnosis and treatment services. Reimbursement by the National Health Insurance plan for providing services including the performance of TST, chest radiography, TPT initiation, and TPT completion was integral to the successful scale-up of household contact management in a recent study in Vietnam [79]. Since many low- and middle-income countries are now adopting similar health insurance schemes, reimbursement changes may allow TB infection management to be scaled up dramatically without providing direct funding to TB programmes. Importantly this will further integrate TB prevention efforts as part of comprehensive care and bring most pediatric TB prevention into the mainstream of primary care, where it belongs.

Notes

Financial support. Dr. Menzies is supported by a Canada Research Chair - Tier 1. Dr. Campbell was supported by Agency for Healthcare Research and Quality grant number T32 HS000063 as part of the Harvard-wide Pediatric Health Services Research Fellowship Program.

Supplement sponsorship. This article appears as part of the supplement “What’s New in Childhood Tuberculosis?” sponsored by the Stop TB Partnership.

Potential conflicts of interest. The authors have no conflicts of interest to disclose.