The Lancet Global Health

Background Long-acting HIV pre-exposure prophylaxis (LAP) could transform HIV prevention in sub-Saharan Africa, but its public-health impact and value-for-money will depend on whether programmes can reach people at greatest risk. We quantified the potential impact and efficiency of targeted LAP deployment across sub-Saharan Africa. Methods We developed a deterministic, age-structured compartmental model of HIV among adults aged 15 years or older in African countries accounting for 98% of adults living with HIV in sub-Saharan Africa. Antiretroviral therapy (ART) was assumed to reach 95% coverage by 2030 in all countries. LAP was assumed to reduce HIV acquisition by 96%. In model analyses, LAP was allocated according to rules that specify the degree to which programmes can effectively allocate it to those at greatest risk of acquiring HIV, including using a person's district, sex and age-group, and other signals of risk. Findings Across 34 countries, allocating a fixed quantity of LAP by district, sex and age averted 2 - 21 times as many infections as uniform allocation; perfect identification of within-group risk, an upper bound, raised this to 5 - 32 times. Multipliers were largest in low-prevalence countries, where absolute impact was smallest; in higher burden settings, they were 2 - 6 and 5 - 11, respectively. If 3 million people used LAP from 2028 in the 11 countries currently receiving Global Fund-supported lenacapavir, uniform allocation would avert about 41,000 infections by 2050, targeting by district, sex and age about 601,000 (range 360,000 - 1,660,000), and perfect risk targeting about 972,000 (490,000 - 3,570,000). Discounted averted ART costs fully offset LAP costs by 2050 at a total cost of up to ~US$50 per person-year with perfect risk targeting, $30 with targeting by district, sex and age, and $2 with uniform allocation. Expansion to 15 million users by 2035 brought modelled adult incidence across sub-Saharan Africa below 0.1 per 1000 person-years in 2049, about a decade earlier than projected without LAP, under the assumption that ART coverage reaches 95% in every country by 2030. Interpretation Targeting determines both the impact and the net cost of LAP. Allocation by district, sex and age is feasible now and multiplies the infections averted by each person-year of LAP. LAP offsets its cost through averted treatment only when it is targeted and its fully loaded cost is near the low end of current estimates. Better identification of people at highest risk within demographic groups, and sustained use among them, could add substantially to impact. The same approach could guide the allocation of LAP in other settings where HIV risk is concentrated.

Background. Tuberculosis preventive treatment (TPT) with levofloxacin reduces tuberculosis disease incidence in multidrug-resistant or rifampicin-resistant (MDR/RR)-tuberculosis household contacts. However, with age, disease progression risk falls and the frequency of adverse events (AE) rises. Methods. To inform decision-makers, we evaluated the cost-effectiveness of expanding the use of MDR/RR-TPT by age, with and without tests of tuberculosis infection. We modelled MDR/RR-household contact management (HHCM) in 189 countries, incorporating age-dependent risks of disease progression, death and AEs. Incremental cost-effectiveness ratios were calculated as USD/disability-adjusted life year (DALY) averted, and as a fraction of per capita gross domestic product (GDP). Findings. Compared to baseline of MDR/RR-HHCM with TPT for those <15 years, adding TPT up to 45 years was cost-effective at a threshold of 30% GDP in 160/189 modelled countries including 17 of the 30 highest incidence countries. Of these 160 countries, the intervention was cost-saving in 46. Strategies targeting TPT to those testing positive for M. tuberculosis infection were dominated by those without testing. Globally, full use of MDR/RR-HHCM with TPT for all would avert 11,900 (10,400 to 13,800) episodes of tuberculosis, 9,960 (9,100 to 10,900) deaths, and 289,000 (262,000 to 321,000) DALYs, but would result in 71,000 (68,000 to 74,000) SAEs and cost $359 million (294 to 449). Interpretation. Expanding MDR/RR-TPT use to older ages without testing appears cost-effective in most settings, but may require substantial resources and result in large numbers of AEs. Funding. Unitaid

Background: COVID-19 movement restrictions reduced social contact worldwide, curbing viral spread but with unequal economic consequences where social safety nets are limited. Little evidence exists on how these impacts varied by occupation, wealth and age in resource-constrained settings. This paper examined pre-pandemic contact and lockdown-related change. Methods: The paper documents analysis of a population-representative survey (N=8,499) in Manicaland, Zimbabwe, during April-May 2021, capturing typical weekday contacts before and during the first COVID-19 lockdown. Negative binomial regression models assessed pre-lockdown contact counts by socio-economic characteristics, overall and by place (home, school, work, transport, entertainment); linear regression assessed predictors of lockdown declines. Results: Pre-COVID-19, respondents reported a median of 12 daily contacts (interquartile range [IQR] 4-47), with higher counts among manual labourers (incidence rate ratio [IRR] 2.21, 95% confidence interval [CI] 2.00, 2.44), retail and public sector workers (IRR 2.33, 95%CI 2.09, 2.60) and students (IRR 2.59, 95%CI 2.38, 2.82), males, wealthier respondents, under-30s, and urban and agricultural estate residents. Counts were highest in schools. During lockdown, contacts fell across all subgroups to a median of 8 (IQR 4-28). Declines were greatest among those with most pre-COVID-19 contacts: urban (mean -17.5, 95%CI -19.53, -15.37) and agricultural estate residents (-18.1, 95%CI -19.99, -16.13), students and teachers (-17.7, 95%CI -19.82, -15.59), males and 30-44 year olds. Drops were greatest in schools and on transport. Conclusion: Reductions in contact under blanket lockdown varied markedly between subgroups and places in Manicaland, Zimbabwe. Selectively targeting high-contact groups might have achieved similar transmission reductions while limiting economic damage, holding lessons for similar settings.

Background: Age is an important determinant of tuberculosis (TB) risk, but population-level age patterns in TB prevalence have not been systematically quantified across settings. We conducted a systematic review of surveys conducted in low- and middle-income countries to quantify trends in TB prevalence across age groups. Methods: We collated age-stratified TB prevalence data for individuals aged 15+ years, from population-based surveys reported between 1993 and 2025. We analyzed data using Bayesian meta-regression models, estimating trends in bacteriologically confirmed pulmonary TB prevalence across age groups. We extended models to understand how age trends varied with alternative outcome definitions and data subsets, and examined correlations between TB age trends and sociodemographic determinants. We also estimated age-stratified notification-to-prevalence ratios as a measure of diagnostic delays for untreated TB. Findings: We included 82 unique surveys reporting age-stratified TB prevalence, spanning 30 countries and 4.6 million participants. These data demonstrated a pattern of increasing bacteriologically confirmed TB prevalence with older age, with each 10-year increase associated with a 30% (95% credible interval: 22, 38) increase in prevalence. TB prevalence for the oldest age group (75+: 796 per 100,000 [522, 1178]) was 6.4 (4.0, 10.4) times higher than the youngest (15-19: 127 per 100,000 [79, 186]), and positive age gradients were estimated for all countries (interquartile range: 23-40% increase per 10-year increase in age). Age gradients were lower in Africa compared to other regions, and for smear-positive compared to bacteriologically confirmed TB. Countries with HIV prevalence [&ge;]5% had flatter age gradients above 40 years of age. Age gradients declined slowly over the study period. Notification-to-prevalence ratios demonstrated a negative age gradient, declining by 10% (7, 14) per 10-year increase in age. Interpretation: The prevalence of untreated TB disease increases substantially throughout adulthood, and lower notification-to-prevalence ratios suggest poorer case detection among older individuals.

Since its inception in 2003, the United States President's Emergency Plan for AIDS Relief (PEPFAR) has played a central role in HIV/AIDS prevention and treatment across Sub-Saharan Africa. Following the January 2025 suspension of United States foreign aid, which includes PEPFAR funding, countries across Sub-Saharan Africa demonstrated varying governmental responses to the resulting HIV funding disruptions. This study evaluates whether national characteristics were associated with a country's ability to respond to this funding crisis. Countries in Sub-Saharan Africa receiving PEPFAR funding were classified into one of three response categories: Level 1 (minimal or no governmental response), Level 2 (government response without sufficient plans or resources to fully offset funding losses), or Level 3 (high governmental response with strategies to maintain HIV/AIDS prevention and treatment services). Response categories were compared with national Human Development Index (HDI), GDP per capita, HIV prevalence, total population living with HIV, annual HIV mortality, and historical PEPFAR funding. Univariable proportional-odds ordinal logistic regression was used to identify factors associated with response level. Five countries (17.2%) were classified as Level 1, 10 (34.5%) as Level 2, and 14 (48.3%) as Level 3. Higher HDI was significantly associated with stronger governmental response (OR = 4.05; 95% CI, 1.49-11.04; Holm-adjusted p = 0.0376). GDP per capita (OR = 2.35; 95% CI, 1.00-5.51; unadjusted p = 0.0489) and HIV prevalence (OR = 3.53; 95% CI, 1.03-12.11; unadjusted p = 0.0452) demonstrated positive associations before adjustment for multiple comparisons, but neither remained significant after Holm correction (adjusted p = 0.2261 for both). Historical PEPFAR funding, total population living with HIV, and annual HIV mortality were not significantly associated with response level after adjustment. Countries with higher levels of human development demonstrated stronger governmental responses to the 2025 PEPFAR funding disruptions. Several countries with minimal documented responses were also affected by active conflict or political instability, potentially compounding their vulnerability to treatment interruptions. These findings suggest that broader national development may play an important role in governmental responses to funding disruptions and highlight the need to strengthen resilient health systems, diversify HIV financing, and prioritize support for countries at greatest risk of future funding shocks.

Background Viral load suppression (VLS) remains suboptimal among adolescents and youth living with HIV (AYLWH) in Malawi. Depressive symptoms may affect treatment adherence and HIV outcomes, but evidence is limited. We assessed the association between depressive symptoms and VLS among AYLWH using Malawi Population-based HIV Impact Assessment (MPHIA) 2020-2021 data. Methods We conducted a cross-sectional secondary analysis of AYLWH aged 15-24 years. VLS was defined as HIV RNA less than 200 copies/mL and depressive symptoms as a Patient Health Questionnaire-2 score of 3 or higher. Survey-weighted logistic regression assessed associations with VLS, adjusting for sex, education, residence, household wealth, alcohol use, Teen Club participation, and awareness of HIV-positive status. Findings Among 201 participants, viral-load results were available for 200; 137 (68.5%) achieved VLS. Six (3.0%) participants had depressive symptoms. Depressive symptoms were not associated with VLS in univariable analysis (OR 0.8, 95% CI 0.1-4.6; p=0.80) or multivariable analysis (aOR 0.3, 95% CI 0.1-2.4; p=0.28). Urban residence (aOR 0.2, 95% CI 0.1-0.5; p=0.0020) and unawareness of HIV-positive status (aOR 0.1, 95% CI 0.02-0.1; p<0.001) were associated with lower odds of VLS. Interpretation Depressive symptoms were not significantly associated with VLS, although their low prevalence and PHQ-2 limitations reduced precision. The findings highlight the importance of strengthening HIV-status awareness and addressing context-specific barriers to care among AYLWH. Funding This secondary analysis received no specific funding.

Background: Nigeria introduced single-dose HPV vaccination for girls aged 9-14 in 2023, reaching approximately 60% aggregate coverage by 2025. While an important achievement, this remains short of the 90% coverage called for under the WHO's 90-70-90 cervical cancer elimination targets, and further scale-up depends on reaching out-of-school girls. In this study we used a model of HPV transmission in Nigeria to quantify the residual cervical cancer burden under continued adolescent delivery, assess how additional screening scale-up and equity assumptions modify it, and identify the effective coverage requirements for an infant HPV vaccination program to be non-inferior to the existing adolescent strategy. Methods: We used HPVsim, an agent-based microsimulation calibrated to Nigerian sexual behavior, HPV prevalence, and cervical cancer incidence, to project new cervical cancer cases and age-standardized incidence over 2025-2100 under scenarios which vary vaccination delivery mode (status-quo adolescent, WHO 90-70-90 adolescent scale-up, and 60/75/90% infant delivery), screening coverage (opportunistic 15% vs. scaled 70% aggregate), and the degree to which primary school completion correlated with vaccination (odds ratio 1 vs. 5) and with screening (odds ratio 1 vs. 3). Infant delivery scenarios were run at hypothetical effective-vaccine-efficacy-at-exposure values of 50%, 70%, and 95% - scenario inputs bracketing the plausible waning range, not evidence-based estimates of infant vaccine performance (no HPV vaccine trial has enrolled anyone below age 10). Each scenario was run at 50 replicates (10 calibration draws, 5 seeds); reported outcomes are the median across replicates and uncertainty ranges are interquartile (25th-75th percentile). Results: Continuation of Nigeria's current adolescent program is projected to avert 1.82 million cervical cancer cases (IQR 1.65-1.90 million; 95% UI 1.39-2.02 million) over 2025-2100 (a 53% reduction relative to no vaccination, IQR 51-55%; 95% UI 46-58%), with age-standardized incidence falling from 19.7 to 7.8 per 100,000 by 2100. However, 82% of the near-term (2025-2075) burden is projected to fall on the pre-2015 birth cohort, i.e. women already beyond the vaccination target age, leaving approximately 945,000 residual cases (IQR 905K-1.00M; 95% UI 835K-1.14M) in this cohort. Scaling up screening to WHO targets was estimated to avert a further ~387K cases (IQR 331-424K; 95% UI 269-515K), roughly 25% of the residual burden; results were about 6-8% better if the scale-up instead reached all women equitably regardless of education. Under a hypothetical scenario of infant vaccination at 90% coverage and 95% effective vaccine efficacy at exposure (matching adolescent-age performance), outcomes were comparable to full WHO adolescent scale-up (49% vs. 54% reduction in vaccine-targetable cohort cancers), but under a hypothetical 50% effective efficacy at Nigeria's DTP3-anchored 62% coverage the strategy averted only a fraction of these cancers. Conclusions: The near-term cervical cancer burden in Nigeria is dominated by cohorts vaccination cannot reach; scaling screening for them is the more consequential lever over the coming three decades and rests on established evidence. Findings on infant HPV delivery are exploratory and hypothesis-generating: they depend on three unverified empirical prerequisites - infant HPV vaccine immunogenicity in the first months of life, durability of protection across the ~15-25-year interval to peak HPV exposure, and coverage achievable through Nigeria's routine childhood immunization platform - and should not be read as a basis for any policy recommendation contingent on infant HPV delivery until those prerequisites are established.

Background State-level estimates of hepatitis C virus (HCV) prevalence are needed to guide resource allocation and to provide a baseline against which to measure progress toward elimination goals. Population prevalence is not directly observable, requiring estimation from indirect data sources. Methods We adapted a Bayesian spatial integrated abundance model to estimate state-level HCV prevalence during 2017-2020 across 48 states and the District of Columbia. The model integrated six HCV-related outcomes: acute and chronic surveillance cases, HCV-related deaths, observations of HCV in [MarketScan] administrative claims data, diagnoses of HCV in Medicaid recipients, and treatment with direct-acting antivirals in Medicaid recipients. Estimates were anchored to a national prevalence estimate, and the model accounted for data source-specific selection, heterogeneity in HCV risk factors across states, and geospatial correlation. Results Estimated average prevalence was 1.32% (95% credible interval [CrI]: 0.94%-1.80%), corresponding to 3.33 million (95% CrI: 2.37-4.65 million) adults with HCV infection across 48 states and DC. State estimates ranged from 0.74% in North Dakota to 2.22% in Oklahoma (median state-specific prevalence, 1.27%), with higher prevalence concentrated in South Central states, Appalachia, and the West, and the lower prevalence in the upper Midwest, Southeast, and New England. Ten states accounted for 56% of estimated infections. Estimates were stable in sensitivity analyses (most differences <1 percentage point). Conclusions These estimates quantify state-level HCV burden prior to the 2021 federal expansion of HCV surveillance funding, providing a baseline for monitoring elimination progress. Geographic variation indicates that resource needs will differ by jurisdiction.

Brucellosis is a globally distributed and widely under-appreciated zoonotic bacterial disease that causes infertility and abortion in livestock and febrile illness in humans, sometimes with debilitating sequelae. Brucellosis incidence estimates typically focus on acute presentations and range from 50 to 100 per 100,000 people per year for endemic settings, showing marked variation within and between countries. Estimating brucellosis incidence in low-resource settings is complicated by lack of access to and utilisation of health care and high-quality diagnostic testing for many people in rural communities. To estimate brucellosis incidence in a pastoralist population in northern Tanzania, we developed a Bayesian model that integrates multiple data sources including diagnostic test results from a hospital-based prevalence study performed in 2016-2017, test sensitivity and specificity, and data from a community-based healthcare utilisation survey in 2019-2020. We show that this population experienced a high incidence of brucellosis with wide uncertainty, with estimates of 387.6 (95% HPDI: 103.3 - 903.5) illnesses per 100,000 people per year for the total population, 199.7 (95% HPDI: 16.2 - 584.6) for individuals aged >15 years, and 632.4 (95% HPDI: 180.1 - 1379.4) for children aged [&le;]15 years. These incidence estimates were among the highest in the world to date. The wide density intervals for these incidence estimates reflect uncertainty in the underlying parameter values used. Sensitivity analyses show that healthcare utilisation influenced incidence estimates the most. This result indicates that accounting for low levels of healthcare utilisation is essential to avoid under-estimating disease incidence. Our findings illustrate that brucellosis is a frequent cause of illness in this population. Increased access to treatment, prevention, and accurate diagnostic testing would greatly improve understanding of brucellosis epidemiology and the health of many people living in similar at-risk communities.

BackgroundPopulation-based spirometry data from sub-Saharan Africa remain limited. We characterised lung function, chronic obstructive pulmonary disease (COPD), lung restriction and associated factors from two Sub-Saharan African countries. MethodsWe performed spirometry using ndd Easy On-PC spirometers among adults aged [&ge;]40 years from population-based cohorts in four communities in Burkina Faso and South Africa from January 2019-April 2022. We included 1,619 participants (936 women, 683 men) with repeatable forced expiratory volume in one second (FEV1) and forced vital capacity (FVC). COPD was defined as FEV1/FVC<0.70 and lung restriction as FEV1/FVC[&ge;]0.70, with FVC<80% predicted. We used multivariable models to test factors associated with lung function measurements (FEV1, FVC and FEV1/FVC), COPD and lung restriction. ResultsMedian FEV1 and FVC ranged from 1.940-2.273L and 2.521-2.883L, respectively. COPD prevalence was 16.1% (95% CI 14.4-18.0; site range 10.8%-24.5%) and lung restriction prevalence was 21.3% (95% CI 19.4-23.4; site range 11.3%-36.2%). Prior tuberculosis was associated with higher odds of COPD (OR 2.86, 95% CI 1.75-4.59) and lower FEV1, FVC and FEV1/FVC. HIV infection was associated with higher COPD odds (OR 1.67, 95% CI 1.01-2.72) and lower FEV1/FVC. Hypertension was associated with higher lung restriction odds (OR 1.46, 95% CI 1.07-2.00) and lower FEV1 and FVC. Diabetes was associated with lower COPD odds (OR 0.50, 95% CI 0.24-0.94), lower FVC and higher FEV1/FVC. Ever-smoking and older age were associated with higher COPD odds (OR 2.21, 95% CI 1.48-3.32; OR 1.05, 95% CI 1.03-1.07, respectively); older age was associated with higher lung restriction odds (OR 1.02, 95% CI 1.00-1.04). ConclusionThis multisite study highlights substantial heterogeneity in COPD and lung restriction prevalence across African populations and identifies respiratory multimorbidity with infectious and cardiometabolic conditions. These findings provide population-based evidence to inform respiratory surveillance in sub-Saharan Africa.

Background The abrupt withdrawal and uneven restoration of United States PEPFAR support in 2025 disrupted HIV service delivery in aid-dependent countries worldwide. In Zambia, decreases in HIV service delivery have been documented at a national level, but variations by facility and by quarter have largely been masked by existing aggregate district- and province-level estimates. We examined facility-level variation in HIV service volumes delivered before and after the funding cuts. Methods We analysed monthly District Health Information System (DHIS2) data covering testing volumes, ART initiations, and numbers on ART for January 2024-December 2025) for two sets of facilities: 1) 24 selected SENTINEL network facilities across four provinces (Central, Copperbelt, Lusaka, and Southern); and 2) all facilities in two districts, Chongwe in Lusaka Province, which was supported by the U.S. CDC under PEPFAR, and Mpongwe in Copperbelt Province, which was supported by USAID under PEPFAR. For each facility, we calculated quarterly 2025:2024 ratios of service volumes and summarized quarterly and annual percentage changes in numbers of tests, treatment initiations, and clients on ART. Results Aggregate HIV testing volumes declined 18% in SENTINEL facilities from 2024 to 2025, with 17 of 24 facilities showing net annual decreases. Aggregate ART initiations declined 16%, with 23 of 24 facilities experiencing net decreases. For both indicators, variation between quarters and among individual facilities matched or exceeded aggregate changes. Numbers of clients currently on ART were comparatively stable, with only 6 facilities showing net declines exceeding 5%. District-level analyses showed divergent patterns. Net aggregate testing volumes increased by 7% in Chongwe District, but 28/39 facilities experienced quarterly declines exceeding 10% in at least one quarter and fell by 40% in Mpongwe District overall. Only 4 of 36 sites in Chongwe showed a net loss of clients over the course of 2025, with most experiencing gains, while Mpongwe experienced substantial variation by facility, with many having declines of >30% during some quarters and 14 of the 25 facilities suffering a net loss from 2024 to 2025. Conclusions Facility-level performance varied substantially within the same district following the 2025 PEPFAR funding disruptions, with some sites losing most of their service volumes while others remained stable or exceeded 2024 levels. These findings emphasize the need for facility-level monitoring to target recovery support where it is needed most.

BACKGROUND: Adverse consequences of neonatal hyperbilirubinemia disproportionately affect babies in low- and middle-income countries. Laboratory measurement of total serum bilirubin is limited in low-resource settings, necessitating simpler alternatives. Our study aimed to evaluate the diagnostic accuracy of visual inspection of jaundice via the World Health Organization Integrated Management of Childhood Illnesses algorithm compared to transcutaneous bilirubinometry for the identification of hyperbilirubinemia in community settings. METHODS: Hyperbilirubinemia was assessed by visual inspection for jaundice and transcutaneous bilirubinometry at 0-72 hours, 3-5 days, and 5-14 days postnatal in the PRISMA prospective cohort study in six sites in Asia and Africa. We assessed sensitivity, specificity, and positive and negative percent agreement. RESULTS: We analyzed data from 14,331 infants, including 1,852 preterm infants. 15% had hyperbilirubinemia, as defined by transcutaneous bilirubin values and the NICE nomograms, at 3-5 days postnatal. Visual inspection had a sensitivity of 16%, specificity of 96%, positive percent agreement of 56%, and negative percent agreement of 86% at 3-5 days postnatal age, pooled across all sites and gestational ages, versus transcutaneous bilirubin defined by NICE nomograms. CONCLUSION: The poor sensitivity of visual inspection underscores an urgent need for accurate, affordable screening tools to improve screening for hyperbilirubinemia.

Background Sustained retention in HIV care is essential for maintaining antiretroviral therapy (ART) and achieving viral suppression. We described national and regional retention in HIV care among paediatric and adolescent ART initiation cohorts in Uganda from 2020-2023. Methods We analyzed aggregated paediatric (aged 0-9 years) and adolescent (aged 10-19 years) ART cohort data reported through the District Health Information Software. We assessed 16 consecutive quarterly ART initiation cohorts per age group, from Quarter 1 2020 to Quarter 4 2023, at 6 (short-term), 12 (intermediate), and 24 (long-term) months after ART initiation. Each cohort comprised clients newly initiated on ART in a specific calendar quarter. Retention was defined as the proportion of clients alive and on ART within the net current cohort (original cohort plus transfer-ins minus transfer-outs). We described retention nationally and by region for each cohort and follow-up time point. We summarized the proportion of cohorts with retention [&ge;]95% by region as a descriptive measure of high retention. Results Among paediatric cohorts, median retention was 84% (IQR: 82-85%) at 6 months, 81% (IQR: 78-83%) at 12 months, and 78% (IQR: 75-80%) at 24 months. Only one national paediatric cohort achieved [&ge;]95% retention at 6 months and one at 12 months; none attained this target at 24 months. The proportion of paediatric cohorts with regional retention [&ge;]95% was highest in Karamoja at short-term follow-up (31%), in Kigezi at intermediate follow-up (31%), and in Lango at long-term follow-up (13%). Among adolescent cohorts, median retention was 82% (IQR: 80-85%) at 6 months, 82% (IQR: 81-85%) at 12 months, and 76% (IQR: 74-78%) at 24 months. No national adolescent cohort achieved [&ge;]95% retention. The proportion of adolescent cohorts with regional retention [&ge;]95% was highest in West Nile at 6 months (19%) and 12 months (31%), and in Karamoja at 24 months (25%). Conclusion Retention among paediatric and adolescent ART cohorts declined with longer follow-up and varied across regions. Strengthening age-appropriate support beyond the first year of ART and improving routine cohort-data quality could support more targeted programme responses.

BackgroundTuberculosis (TB), caused by Mycobacterium tuberculosis (MTB), remains a leading global public health challenge. Malnutrition is a major, modifiable driver of this vulnerability to the disease, creating a lethal, bidirectional link with TB. This s review aims to synthesize available evidence to estimate the pooled prevalence of TB and identify associated factors among malnourished children in Ethiopia, providing critical data to guide targeted, integrated public health interventions. MethodsA systematic review and meta-analysis were conducted on studies reporting the prevalence of TB and associated factors among malnourished children in Ethiopia. Electronic databases including PubMed/Medline, EMBASE, CINAHL, ScienceDirect, Academic Search Complete, and Google Scholar were searched. In addition, bibliographies of relevant articles were reviewed, and a search for relevant conference abstracts was performed. Eligible studies were those published in peer-reviewed journals that specifically addressed TB prevalence among malnourished children. A random-effects model was employed to estimate the pooled prevalence of TB and its associated factors. ResultsA total of 746 studies were initially identified through database and catalog searches. Following screening and eligibility assessment, 12 studies were included in the final systematic review and meta-analysis. These studies, published between 2017 and 2024, collectively involved 5,749 participants. The smallest sample size (133) was reported in a study from the Oromia region, while the largest (1,201) was from the Amhara region. The pooled prevalence of TB among malnourished children was estimated at 16%. The highest prevalence (95%) was reported in a study conducted in Addis Ababa, while the lowest (6%) was from Harar. Considerable variation in prevalence was noted across the included studies. ConclusionThis review resulted in the substantial prevalence of TB among malnourished children in Ethiopia, with a pooled prevalence of 16%. This finding highlights the urgent need for integrated healthcare interventions targeting both TB and malnutrition

Retention in antiretroviral therapy (ART) programmes remains a major public health challenge in HIV care, with loss to follow-up (LTFU) and death threatening treatment success. Because these outcomes are competing events, conventional survival methods may yield biased estimates of their incidence and associated risk factors. To estimate the cumulative incidence of LTFU and mortality and identify factors associated with these outcomes among adults receiving ART in Northern Ghana using a competing-risks framework. We conducted a retrospective cohort study of 329 adults living with HIV who initiated ART at Nalerigu Baptist Medical Centre, Ghana, between January 2013 and December 2019. Fine--Gray subdistribution hazard models were used to evaluate predictors of LTFU and death while accounting for competing risks. Cause-specific Cox proportional hazards models were fitted as a sensitivity analysis. The incidence rates were 2.2 per 100 person-months (95% CI: 1.9--2.6) for LTFU and 0.74 per 100 person-months (95% CI: 0.57--0.98) for death. At 36 months, the cumulative incidence was 50.2% for LTFU and 16.0% for death. Patients residing closer to the clinic had a lower cumulative incidence of LTFU than those living farther away (SHR = 0.68, 95% CI: 0.49--0.93), whereas higher baseline weight was associated with increased LTFU (SHR = 1.02 per kg, 95% CI: 1.00--1.04). Mortality was lower among patients receiving efavirenz-based regimens (SHR = 0.27, 95% CI: 0.15--0.50), females (SHR = 0.39, 95% CI: 0.18--0.85), and those with higher baseline weight (SHR = 0.93 per kg, 95% CI: 0.90--0.96). Findings were broadly consistent in cause-specific Cox models. LTFU was substantially more frequent than mortality among adults receiving ART in Northern Ghana. Travel distance to the treatment facility was an important barrier to retention, while ART regimen, sex, and baseline weight were associated with mortality. Expanding access to ART services through decentralized and community-based models may improve retention and support efforts to achieve HIV control targets.

Long-acting injectable human immunodeficiency virus (HIV) medicines expand treatment and prevention choices, but willingness to use them does not establish readiness to begin. We examined readiness to start these options, associated factors and conditions for acceptance among people living with HIV and pre-exposure prophylaxis (PrEP)-service clients in Uganda. This cross-sectional study analysed 225 participants at Mildmay Hospital and Kajjansi Health Centre IV in central Uganda. Readiness meant preferring initiation immediately or within three months. Modified Poisson regression estimated adjusted prevalence ratios (aPRs) and 95% confidence intervals (CIs), including age-by-service-group interaction. Separate models examined informational, perceptual and financial factors, adjusting for service group, age and sex. Logistic marginal standardisation checked demographic and expanded age comparisons. Overall, 108 participants (48.0%) were ready to start injectable HIV treatment or prevention, 111 (49.3%) wanted more information about these options, and six (2.7%) preferred initiation in six to twelve months. Readiness was 80.8% among people living with HIV aged 15-24 years and 34.4% among those aged 25 years or older (aPR 2.35, 95% CI 1.63-3.38). Corresponding proportions among PrEP-service clients were 39.2% and 41.4% (aPR 0.95, 95% CI 0.60-1.50; interaction p=0.002). In separate adjusted models, clinic-staff information (aPR 1.94, 95% CI 1.42-2.66), belief in greater effectiveness (2.00, 1.43-2.80) and a one-point increase in benefit score (3.27, 2.36-4.55) were associated with higher readiness. A one-point increase in concern score (0.70, 0.60-0.80) and financial difficulty sometimes or more often (0.68, 0.48-0.96) were associated with lower readiness. Fewer side effects was the leading condition for accepting injectable HIV treatment or prevention across service and readiness groups. Nearly half of participants were ready to start, while a similar proportion wanted further information about injectable options. Programmes should assess intended initiation timing alongside willingness and address safety, appointment flexibility, privacy and affordability through counselling and service planning.

Background: Early pandemic responses must balance transmission control against socioeconomic costs but the strategy that minimises total costs and whether it varies by country remain unknown. We analysed country-specific cost-optimal intervention strategies and examined whether national indices could inform response targets. Methods: We fitted mechanistic models to the first year of reported COVID-19 cases in 110 countries and assessed structural identifiability and parameter uncertainty. Multi-objective optimisation identified efficient trade-offs between transmission reduction and infections across 26-, 40-, and 52-week horizons without vaccination. We selected cost-minimising strategies using GDP-scaled intervention costs and infection costs incorporating quarantine and mortality. Excess-cost ratios, defined as the relative cost increase over the optimum, quantified deviations from the cost-minimising strategy. We also examined associations with national socioeconomic and health-system indices. Results: Cost-optimal average transmission reduction ranged from 31.2% to 57.3% across countries. Across the time horizons, the mean excess-cost ratio for 110 countries was 55.2% (median 30.8%). Excess costs increased asymmetrically around the country-specific optimal transmission reduction: a five-percentage-point transmission reduction below the optimum was associated with excess-cost ratios exceeding 400%, compared with 10-80% for an equivalent increase above the optimum. The range of transmission reduction associated with excess-cost ratios below 40% widened 1.47-fold as the horizon increased from 26 to 52 weeks. Among national indices, the Socio-Demographic Index (SDI) and GDP per capita had the largest redundancy indices (0.40 and 0.37, respectively). Conclusions: Under the modelled assumptions, transmission reduction below the country-specific optimum resulted in greater excess costs than reduction above the optimum, highlighting the potential value of precautionary early-response strategies when intervention effectiveness is uncertain. National socioeconomic and health-system indices may provide a broad indication of cost-optimal intervention intensity, although country-specific conditions remain important. These findings provide a quantitative framework for balancing intervention intensity, infection burden, and economic costs across diverse national settings.

ABSTRACT Background: The double burden of malnutrition, combining overweight or obesity with anemia is vital in nutritional transitions. Evidence regarding its national distribution and prediction using routine survey data in Ethiopia remains limited. Objective: To estimate the national prevalence of individual-level double burden of malnutrition among non-pregnant Ethiopian women aged 15-49 years and evaluate machine-learning models using data from the 2024/25 Ethiopian Demographic and Health Survey. Methods: we analyzed 14250 non-pregnant women. Double burden of malnutrition was defined as a body mass index equal to or greater than 25.0 kilograms per square meter with an altitude- and smoking-adjusted hemoglobin concentration below 12.0 grams per deciliter. Data were split into an 80% training set and a 20% holdout test set. Six algorithms were evaluated using cross-validation, while SHapley Additive exPlanations described feature contributions. Results: the weighted national double burden of malnutrition prevalence was 8.4% (95% confidence interval 7.9%-8.9%), higher in urban residents (20.0% versus 4.0%) and the richest wealth quintile (26.7% versus 2.7%). Extreme Gradient Boosting achieved an area under the receiver operating characteristic curve of 0.892 (95% confidence interval 0.875-0.909) and an area under the precision-recall curve of 0.764 outperforming logistic regression (area under the receiver operating characteristic curve of 0.781). Extreme Gradient Boosting sensitivity was 82.1%, specificity was 81.4% and the Brier score was 0.068. Household wealth, residence and respondent age were top predictors. Conclusions; the double burden of malnutrition affected one in twelve non-pregnant Ethiopian women unevenly across groups. Routine characteristics effectively discriminated double burden of malnutrition cases though external validation is needed prior to clinical use.

Background Malaria and HIV remain major causes of illness in sub-Saharan Africa, accounting for 94% of the global malaria burden and over two-thirds of the global HIV burden. HIV-related immunosuppression increases susceptibility to malaria, with a higher malaria burden of up to 47% and severe malaria complications compared to HIV-negative individuals. Socio-demographic factors, individual knowledge, attitude, and perception are recognized determinants of malaria prevention behaviours, but no study has modelled their combined contribution, and there is scant research on these factors among people living with HIV. We assessed the level of knowledge, attitudes, and perceptions towards malaria, and identified the predictors of malaria prevention practice among HIV-positive individuals in Kampala, Uganda. Methods This cross-sectional study enrolled 620 HIV-positive patients attending the ART clinic of China-Uganda Friendship Hospital, Naguru, selected by systematic random sampling. Data collection was conducted from 21st July 2025 to 18th August 2025 using a structured, interviewer-administered questionnaire that measured socio-demographic characteristics, knowledge, attitude, perception, and the practice of malaria prevention. Associations were assessed using Mann-Whitney U, Kruskal-Wallis H, and Spearman correlation. The independent predictors of good malaria prevention practice were identified using binary logistic regression. Results Participants had a mean age of 34.8 years, were predominantly female (68.9%), Baganda (38.1%) by tribe, and the majority had a secondary level of education(46.0%). Most had good knowledge of malaria (97.3%) and a positive attitude (90.5%), while 57.3% held a positive perception of malaria risk and prevention; only 62.6% reported good malaria prevention practice, and most prevention methods used were sleeping under a mosquito net (71.1 3%) and closing windows early in the evening (70.3%). In regression analysis, knowledge (adjusted odds ratio [AOR] 1.198, 95% CI 1.03-1.40, p = 0.023) and attitude (AOR 1.101, 95% CI 1.04-1.17, p = 0.001) independently predicted good practice, while perception did not (AOR 1.00, 95% CI 0.96-1.04, p = 0.994). Scores of Attitude, perception, and malaria prevention practices varied significantly, though non-linearly, by education level, and were highest among non-educated participants. Conclusions Despite near-universal knowledge and predominantly positive attitudes in this study, more than a third of HIV-positive participants did not practice adequate malaria prevention, and perception did not independently predict malaria prevention practice once knowledge and attitude were accounted for. Programmes seeking to close this know-do gap should pair health education with structural support for prevention tools rather than education alone.

BackgroundModern contraceptive use remains low in many sub-Saharan African countries, including Mozambique, limiting progress towards sexual and reproductive health and Sustainable Development Goals (SDGs). Understanding the prevalence, determinants, and inequalities in modern contraceptive use based on Mozambique-wide national survey is critical to informing effective family planning interventions. MethodWe analyzed the most recent data from the Mozambique Demographic and Health Survey (DHS), 2022-2023, focusing on a total of 11,783 reproductive-aged women (15-49 years). The outcome variable was current use of modern contraceptives. Several individual, household-related and community determinants of modern contraceptives were examined employing multivariable logistic regression. Adjusted odds ratio with corresponding confidence intervals and p-values were reported. Erreygers Concentration Index (ECI) and concentration curves were used to assess wealth-and empowerment-related inequalities. All analyses were conducted on weighted data. ResultOverall, 27.6% of women reported current use of modern contraceptive methods. Younger age, higher education, being employed, higher parity, greater wealth, and a partner with higher education were positively associated with modern contraceptive use. Significant wealth-related (ECI = 0.32, p < 0.001) and womens empowerment-related (ECI = 0.26, p < 0.001) inequalities were observed, indicating that wealthier and more empowered women had higher use of modern contraceptive methods. ConclusionModern contraceptive use in Mozambique is suboptimal and influenced by a combination of individual, household, and contextual factors, with marked socioeconomic and empowerment-related disparities. Policy interventions should focus on improving access to and encouraging use of contraceptives for economically disadvantaged and women with low-empowerment, promoting womens autonomy, enhancing male involvement, and leveraging media and community-based programs to increase awareness and uptake. Equity-focused strategies are essential to achieving national and global family planning targets.