PEPFAR MER data are routinely collected by implementing partners and reported in aggregate with stratifications by age, sex, and ART status, among others. Some indicators assess repeated measures (e.g. TB screening and screening outcome), where many, but not all, individuals will be included from one reporting period to the next (Table 1). Others are one-time assessments (e.g. initiation and completion of TPT). The indicators included in this analysis were reported semi-annually by PEPFAR-supported treatment sites [25] in 16 countries in central/west, east, and southern sub-Saharan Africa during a 4-year period: 1 October 2018 to 30 September 2022 (corresponding with FY2019–FY2022) (Table 1). All 16 countries are considered high-burden for HIV-associated TB (high TB/HIV burden) [1]. Countries were categorized as follows, consistent with standard sub-regional categorizations: central/west Africa (Cameroon, the Democratic Republic of the Congo, and Nigeria); east Africa (Ethiopia, Kenya, Tanzania, and Uganda); and southern Africa (Botswana, Eswatini, Lesotho, Malawi, Mozambique, Namibia, South Africa, Zambia, and Zimbabwe). We analyzed MER data describing TB services provided to CLHIV on ART aged under 15 years served at PEPFAR-supported HIV clinics.

We describe six TB indicators (adapted from PEPFAR MER guidance [22]) to describe the pediatric TB/HIV clinical cascade (Table 1 and Fig. 1) including TB screening, treatment and prevention indicators among CLHIV on ART. Table 1 details reporting frequency, numerator, and denominator for each indicator described.

We reviewed and described the performance of each of these MER TB indicators in FY2022 and across the 4-year analytic period (FY2019–FY2022) in semi-annual intervals for the 16 included countries. For FY2022, we stratified variables by geographic region, sex, and ART status wherever possible. Across the 4-year analytic period, we describe performance over time and by region. Because PWH should receive at least one TB screening during each semi-annual reporting period, most of the same individuals screened for TB in one semiannual reporting period would be screened again the following period. In addition, the symptoms used for TB symptom screening are common to other childhood illnesses and opportunistic infections affecting PWH [21], and it is possible some individuals screen positive in multiple reporting periods. Therefore, rates for screening coverage and yield, as well as TB treatment initiation among those screening positive, are presented as the annual mean value of two semi-annual reporting periods. Conversely, TPT initiations and completions were summed across semiannual reporting periods, as these values represent unique individuals (i.e. CLHIV newly starting TPT). Group stratifications included sex, geographic region (central or west Africa, east Africa, or southern Africa), and ART status (newly initiated or already on ART), wherever applicable. In addition, we conducted a supplementary analysis reviewing performance of the six MER TB indicators during FY2022 between CLHIV aged younger than 15 years and adolescents and adults living with HIV aged at least 15 years.

Sites reporting illogical values for any indicator, such as a numerator greater than or equal to 110%, were excluded from the analysis for that time period. Data from Kenya were excluded for the TB screening, screening-positivity, and TB treatment initiation indicators for FY2022 due to a known data quality issue. TPT initiation data are not included for the period October 2018–March 2019 because the denominator for this indicator is calculated from data collected before October 2018, which was not included in this analysis.

All analyses were conducted using R software and Microsoft Excel. This analysis is purely descriptive and includes no tests of statistical significance. This activity was reviewed by the US Centers for Disease Control and Prevention (CDC), deemed not research, and was conducted consistent with applicable federal law and CDC policy. (See 45 C.F.R. part 46, 21 C.F.R. part 56; 42 U.S.C. §241(d); 5 U.S.C. §552a; 44 U.S.C. §3501 et seq.)

In FY2022/Quarter Four (Q4), 524543 CLHIV were in care in the 16 countries in PEPFAR-supported programs. An average 80% of CLHIV in FY2022 were screened for TB using the standard symptom screen, of whom 3.6% screened positive. Among those screening positive, 20% initiated TB treatment (Table 2). Screening coverages in central and west Africa (93%) and east Africa (95%) were higher than in southern Africa (71%). East Africa had the highest proportion of CLHIV screening positive (4.3%). The proportion of CLHIV who were initiated on TB treatment after screening positive for TB varied across regions (east Africa 25%, southern Africa 18%, and central/west Africa 14%). Among CLHIV newly initiated on ART and screened for TB, 11.5% screened positive and 40% of those screening positive were initiated on TB treatment. Comparatively, 3.1% of those already on ART screened positive, and 15% of those screening positive initiated TB treatment. Screening coverage rates were similar among PWH (≥15 years) and CLHIV (82 and 80%, respectively; Table 3). Compared to older PWH, CLHIV had higher screening positivity (3.6 vs. 2.5%) and lower TB treatment initiation rates among those screening positive (20 vs. 23%).

Between FY2019 and FY2022, semiannual TB screening coverage among CLHIV fluctuated between 70 and 81% and TB screening positivity between 2.5 and 4.1% (Fig. 2a). TB treatment initiation among all CLHIV screening positive fluctuated substantially over the 4-year time period, ranging from 19 to 43%. Performance over time for TB screening positivity, screening coverage, and treatment initiation stratified by region is available in Fig. 2b–d.

In FY2022, 37% of CLHIV newly initiated on ART started TPT within 6 months of initiating ART (Table 4). CLHIV in east African countries had lower TPT initiation rates (23%) than those in southern (40%) and central/west African countries (44%). In FY2022, approximately 86% of all CLHIV on ART who initiated TPT in the previous 6-month reporting period completed it. TPT completion among all CLHIV on ART varied by geographic region, from 83% in southern Africa to 92% in central/west Africa. TPT completion rates among CLHIV who were newly initiating both ART and TPT (84%) were similar to TPT completion among all CLHIV on ART (86%). The TPT initiation rate among CLHIV newly on ART was 37% in FY2022 compared to 59% among older PWH newly on ART (Table 3). TPT completion for both PWH and CLHIV newly on ART was 84%.

Between FY2019 and FY2022, TPT initiation among CLHIV newly initiating ART steadily increased over time, from 13% in FY19/Q4 to 37% in FY22/Q4 (Fig. 3a). During FY2019–FY2022, TPT completion rates among all CLHIV on ART ranged from 77 to 86%. Cumulatively, from FY2019 to the end of FY2022, 344 177 CLHIV (newly and already on ART) initiated TPT, of whom 82% (281 875) completed TPT. Performance over time stratified by region for these three indicators is available in Fig. 3b–d.

TB screening and evaluation for TB among CLHIV was overall sub-optimal: while east and central/west Africa reported very high screening coverage, the proportion screening positive was below expected thresholds across time periods and regions. A study reviewing records from 2014 to 2017 found that when TB symptom screening is applied rigorously and consistently, 15% of all CLHIV (regardless of ART status) screen positive for TB in clinic settings similar to those included in our analysis [21]. A separate, multisite study from 2019 to 2021 reported 6% screen-positive rate among CLHIV (already on ART/established in care) being routinely screened in healthcare settings in the Catalyzing Pediatric TB Innovations (CaPTB) project, which also included settings outside of sub-Saharan Africa [26,27]. Lower-than-expected proportion screening positive in our analysis (3.1% (already on ART) – 11.5% (newly initiating ART)) may indicate low-quality, inconsistent, or poorly reported TB symptom screening.

All regions had high proportions of TB diagnoses/treatment initiations among those who screened positive for TB, regardless of ART status. Because symptoms used to screen for TB are also common in other childhood illnesses and opportunistic infections affecting CLHIV, most children screening positive are not expected to be diagnosed with TB. The studies mentioned above found that 3.5–14% of CLHIV with a positive TB symptom screen were ultimately diagnosed with TB [21,26,27]. Of note, in these two studies, a relationship between screening positive and being diagnosed with TB was evident: a higher screen-positive rate (15%) coincided with lower TB diagnosis among those screen-positive (4%) [21], while a lower screen-positive rate (6%) coincided with higher diagnosis among those screen-positive (14%) [21,26]. A more rigorously applied symptom screen will refer more children for evaluation than ultimately have disease, but it will also ensure those with less severe or obvious TB disease are appropriately evaluated [18]. Both studies demonstrated higher screen-positive rates and lower diagnosis/treatment among those screening positive than in PEPFAR-supported settings included in this analysis.

The proportion of CLHIV newly initiating ART who were initiated on TB treatment following a positive TB symptom screen (40%) was high compared with previous findings [21,25]. This may indicate poor screening or reporting practices and missed opportunities to detect TB among this population. This finding may also indicate truly high incidence of TB among CLHIV newly initiating ART, but higher proportions of children would still be expected to screen positive if this were the case. Especially for southern and west/central Africa, trends over the analytic period show the proportion of children initiated on TB treatment appeared lower when screening positivity was higher while the proportion initiated was higher when screening positivity was low, This may be an effect of seasonal peaks in other childhood illnesses with overlapping clinical presentation or suggest that during low screen-positive periods, only the sickest children were being referred for diagnostic evaluation. Further study is warranted to better understand the quality and fidelity of TB screening, expected yield in various settings among CLHIV newly initiating and already on ART, as well as the diagnostic tools employed to make a treatment decision among CLHIV in PEPFAR-supported programs. Globally, new screening and diagnostic approaches for children and adolescents have been recommended, including use of child-friendly specimens, rapid diagnostic tests, and integrated treatment decision algorithms to improve pediatric TB diagnosis [19,28].

TPT initiation early in the course of ART, once TB disease is ruled out, is important to prevent TB disease morbidity and mortality. This is especially true for younger CLHIV, who are often severely immunosuppressed at the time of HIV diagnosis and ART initiation [8,9,29]. Less than half of CLHIV newly initiating ART during FY2019–FY2022 were initiated on TPT within the same reporting period of initiating ART. Timing of ART and/or TPT initiation may be appropriate but may fall in different semiannual reporting periods, leading to artificial underreporting of some TPT initiations. Nevertheless, our findings are consistent with previously published findings, which demonstrated that from 2015 to 2019, 43% of CLHIV on ART in western Kenya initiated TPT within the first 6 months of ART, but in total, 68% of CLHIV on ART initiated TPT regardless of the time on ART [30]. A delay in TPT initiation following ART initiation may occur if there are supply chain issues, or if a clinician has concerns regarding high pill burden, side effects, presumptive TB disease, or treatment for an acute illness [30]. Our findings demonstrated that PWH 15 years of age and older also experienced suboptimal TPT initiation rates during that period, but rates were still considerably higher (by approximately 20%) than among CLHIV, highlighting further the global gap in TPT uptake among children compared to adults [31,32].

Approximately 85% of CLHIV who initiated TPT in FY2022 completed the course, and completion rates between CLHIV and PWH (≥ 15 years) were similar. Completion rates increased over time but did not exceed 86%. Literature describing factors associated with TPT completion in CLHIV is sparse. One study assessed factors among adolescents and young adults living with HIV in Kenya and determined those on ART longer (1–2 years) were more likely to complete TPT than those new on ART (<1 year) [33]. We did not find such a difference in our analysis, though ‘newly initiated on ART’ is defined by MER guidance as within 6 months instead of 1 year. TPT completion rates may improve as shorter and more tolerable regimens (e.g. 3HP) become more widely available for CLHIV [34]. Nevertheless, individualized preferences and challenges will need to be accounted for to assure TPT adherence and completion [35,36]. Overall, our analysis shows that, from FY2019 through FY2022, more than 280 000 CLHIV completed TPT in the 16 high TB and TB/HIV-burden countries.

Several limitations related to MER data collection and reporting hindered the interpretation of our findings. First, variabilities in national guidance, interpretation of reporting requirements, and available tools and diagnostics may impact program function and recording and reporting practices across countries. We reported findings based on established PEFPAR guidance and data available in MER. Second, limitations of MER prevent full assessment of the pediatric TB/HIV cascade, including lack of age-stratification of diagnostic evaluation indicators (e.g. specimen sent for lab testing and test result returned positive) and lack of more finely delineated age stratifications, such as by 5-year increments, for TB screening, treatment, and prevention variables. Third, underreporting and inconsistencies in reporting of TB MER data may have occurred and vary across PEPFAR-supported sites, which may impact our findings. Exclusion criteria were applied to mitigate the impact of illogical values, which resulted in datapoints being censored, particularly in the TB prevention cascade. While retaining these values may have artificially inflated performance, censoring them may also have impacted findings. Correlation may exist between quality of reported data and service delivery; facilities excluded due to data quality issues may have been more likely to also exhibit suboptimal service delivery. Fourth, patient-level data are not reported to PEPFAR, and MER data are based on aggregate values. Therefore, it is not possible to determine whether the same patient cohort is represented at every stage of the TB treatment or prevention cascades. Fifth, our findings are not generalizable to CLHIV or children with TB outside the PEPFAR-supported sites from the 16 countries reviewed in this analysis. Sixth, no causal or associative inferences can be made from this purely descriptive analysis for which no tests of significance were performed. Mitigation measures taken during the COVID-19 pandemic, which began in mid-FY2020 in sub-Saharan Africa, may have resulted in interruptions to care impacting performance during that time.

In conclusion, in this analysis of PEPFAR data from 16 high-burden countries, we reviewed implementation performance of TB/HIV services for CLHIV. While gradual progress is evident, an achievement amidst the disruptions of the COVID pandemic, substantial gaps persist in the delivery of TB screening and prevention services for CLHIV across the 4-year period. Closing these gaps and ensuring CLHIV receive regular, high-quality TB screening and diagnosis, and prompt prevention services are key to reduce their risk of TB-related morbidity and mortality. Lastly, TB reporting gaps persist, particularly with regard to age-specific data and diagnostic evaluation. The limited data available for children under 5 are notable given their heightened risk. Enhancing implementation of pediatric TB/HIV services can help ensure children affected by HIV and TB are not left behind.