OpiniãoPICO questions:
a. In children and adolescents with signs and symptoms of TB, should decentralization of child and adolescent TB services versus centralized child and adolescent TB services (at referral or tertiary hospital level) be used?
b. In children and adolescents exposed to TB, should decentralization of child and adolescent TB prevention and care services versus centralized prevention and care services (at referral or tertiary hospital level) be used to increase coverage of TPT in eligible children and adolescents?
c. In children and adolescents with signs and symptoms of TB, should family-centred, integrated services versus standard, non-family-centred, non-integrated services be used?
d. In children and adolescents exposed to TB, should family-centred, integrated services versus standard, non-family-centred, non-integrated services be used to increase TPT coverage in eligible children and adolescents?
For definitions of decentralization and family-centred, integrated services, please refer to the section with definitions from page xi.
Evidence: A systematic review of studies assessing the impact of decentralized, integrated or family-centred care models on TB diagnosis, treatment or prevention outcomes in children and adolescents with TB between 0 and 19 years old, comprising both children (0-9 years old) and adolescents (1019 years old), was conducted to answer this group of PICO questions. The PubMed, Embase, Web of Science, Global Index Medicus, Global Health and Cochrane Central databases were searched in February 2021, as well as the references of 17 related reviews. 3265 abstracts from databases and 129 additional references from related reviews were identified and assessed. 516 full-text articles were assessed for eligibility, from which 25 comparative studies (7 randomized, 18 observational) were identified; one unpublished observational study was added making a total of 26 studies. Four studies (one randomized, three observational) were excluded after review because the care model described was community-based treatment support, for which a WHO recommendation already exists (100). Of the remaining included studies, 16 had elements of decentralization, five had elements of integration, and three had elements of family-centred care; four studies had elements of more than one care model of interest, but were only included based on their main model, such as either decentralization or family-centred, integrated care. Most studies focused on the 0-14-year age group.
Studies where the primary intervention was decentralization mostly assessed diagnosis or case notification outcomes (n = 16) (48, 101-115), with fewer assessing TPT outcomes (n = 3) (106, 116, 117). In general, interventions that included both strengthening of diagnostic capacity in primary care settings as well as strengthening linkages between communities and facilities consistently showed an increase in case notifications and TPT initiations, while interventions that involved only community-based activities did not.
Two studies of service integration were identified (118, 119), which showed limited impact on case notifications of screening in IMCI clinics or co-location of TB and ART services. The two studies of family-centred care that were identified (120, 121) showed that provision of socioeconomic support packages to families affected by TB was associated with increased TPT initiation and completion.
The reviewers noted that, while substantial wider literature on integration and family-centred care is available, evidence for the specific impact on child and adolescent TB outcomes is limited. Some overlap was noted between integration of TB services into non-specialized settings, such as general outpatient or primary care services, or decentralization. For the evidence review this was a slightly artificial separation, while in practice decentralization and integration into PHC may happen together.
GDG considerations: Regarding the evidence reviewed on the impact of decentralization on TB case detection, the GDG observed that two trials (109, 111) and one observational study of home-based screening (without facility-based strengthening) (114) had fewer diagnoses or notifications among children aged below 15 years in the intervention group compared to the control group, but that none of these differences were statistically significant. The GDG discussed that while there may be a reduction in case notifications at higher levels of care, TB detection may improve if children are seen by a competent clinician at the first point of access (such as at PHC level). The evidence overall was recognized as uncertain. The benefit of increased case finding and increased number of children with TB who are initiated on TB treatment was considered to outweigh the concern for overtreatment. Therefore, undesirable effects for case detection were considered trivial. The GDG discussed potential risks of provision and management of TPT at the peripheral level, including undetected drug-related adverse events such as hepatotoxicity and insufficient capacity to manage these. In addition, there may be a risk of TB disease being treated with a course of TPT rather than with a complete treatment regimen. All these undesirable events can potentially happen but were considered rare and not of major concern. Therefore, undesirable effects for TPT provision were considered trivial as well. Overall, the GDG agreed that the balance of desirable and undesirable effects probably favours decentralized TB services for case detection and provision of TPT to children and adolescents. The panel noted that differences in setting and availability of adequate resources are important considerations.
The GDG discussed that family-centred, integrated care includes interventions at the household level to identify members of the household requiring evaluation for TB disease, TPT, treatment support, etc. Some overlap between integration of TB services into non-specialized settings such as general outpatient or primary care services, and decentralization was noted. For the evidence review, this was a slightly artificial separation, while in practice decentralization and integration into PHC may happen together. Overall, despite a lack of evidence on undesirable effects and low quality of the data, the panel agreed that there is evidence of positive effects of family-centred integrated care. It was suggested that family-centred, integrated care could be an addition to the standard of care as well as to specialized services which do not have an integration component. Family-centred care in the sense of family involvement was highlighted as a core principle of child health care.
The GDG discussed that setting specific factors related to TB burden or the organization of health services may impact feasibility, acceptability and equity considerations. They also discussed that the initial health system costs to establish decentralized and family-centred, integrated services may be relatively high (such as infrastructure, human resources, training, equipment, community engagement), but that costs are likely to decrease over time, assuming that people with TB are effectively managed and TPT is provided at the peripheral level, leading to a reduction in TB incidence. Decentralized and family-centred, integrated services may result in important savings for affected families. Equity was considered an important cross-cutting issue impacting cost as well. The GDG highlighted that TPT implementation can be very challenging with high levels of loss-to-follow-up in programmes implemented at higher levels of the health system, considering that children who are eligible for TPT are not sick. The panel agreed that decentralization and integration of services can potentially increase equity and enhance the success of the programme and judged that cost-effectiveness probably favours decentralized and family-centred, integrated approaches to both case finding and provision of TPT.
While the GDG stressed the importance of taking into consideration the potential impact of stigma when decentralizing TB services for children and adolescents to lower levels, the panel judged that decentralized approaches are probably acceptable to key stakeholders. Overall decentralized and family-centred, integrated approaches were judged feasible to implement, although feasibility may vary depending on infrastructure, available funding and the structure of the NTP, among others. However, adequate investment is critical to enable the acceptability, equity and feasibility of decentralized approaches.