Muhimbili National Hospital is a 1259-bed hospital in Dar es Salaam, Tanzania. It is the national referral hospital as well as teaching and research facility for the Muhimbili University of Health and Allied Sciences. It offers speciality care across the spectrum of clinical medicine. About 40% of its beds are for private patients.
Introduction: Achieving viral suppression is central to HIV epidemic control; however, routine viral load (VL) testing in many low- and middle-income countries is constrained by laboratory capacity, logistics, and cost. In Tanzania, disparities in VL coverage persist across age groups and geographical regions, limiting the timely detection of treatment failure. Pharmacy refill adherence is a low-cost, routinely collected objective indicator of treatment behavior. This study assessed whether pharmacy refill adherence, enhanced using machine learning (ML) models, can reliably predict viral suppression among people living with HIV (PLHIV) in Tanzania. Methods: We conducted a retrospective analysis using nationally representative patient-level data from the Care and Treatment Center (CTC-2) database, collected between 2017 and 2021. A random sample of 40,000 records was drawn, of which 28,044 patients met the inclusion criteria. Pharmacy refill adherence was calculated as the proportion of days covered and capped at 100%. Viral suppression was defined as a VL of <1,000 copies/mL. Logistic regression, Random Forest, Gradient Boosting Machine (GBM), and XGBoost models were trained using an 80/20 training-testing split, and the model performance was evaluated using the area under the receiver operating characteristic curve (AUC). Youden's Index was used to determine the optimal adherence threshold. Results: Among the 28,044 patients included in the analysis, the median age at ART initiation was 38 years, and 64.9% were female. The median pharmacy refill adherence was 90.64% (mean, 87.37%). Viral load (VL) measurements were available for 21,572 patients, of whom 88.7% achieved viral suppression. Higher pharmacy refill adherence was strongly associated with viral suppression, whereas lower adherence was observed among adolescents, young adults, and individuals who were lost to follow-up. Marked geographic variation was observed, with higher adherence in regions such as Dar es Salaam and lower adherence in more remote regions, including Rukwa and Singida. Among machine learning models, XGBoost demonstrated the highest predictive performance (AUC >0.85), followed by Gradient Boosting Machines and Random Forest, while logistic regression provided stable baseline estimates. Pharmacy refill adherence, duration of follow-up, clinic visit frequency, and patient age were the strongest predictors of viral suppression. Conclusion: Pharmacy refill adherence is a strong predictor of viral suppression and provides a feasible and cost-effective tool for monitoring ART outcomes in settings with limited VL testing. Machine learning approaches further enhance the predictive value of routine program data and can support the early identification of patients at risk of virological failure. Integrating adherence-based predictive analytics into national HIV program monitoring systems may strengthen differentiated service delivery, improve treatment outcomes, and accelerate progress toward the UNAIDS 95-95-95 targets in Tanzania and similar resource-limited settings. ### Competing Interest Statement The authors have declared no competing interest. ### Funding Statement Meshack D. Lugoba received a small Sida-supported grant administered through Muhimbili University of Health and Allied Sciences (MUHAS), which contributed to capacity building but did not influence the study design, data analysis, interpretation of results, or the decision to publish. Data access and institutional support were provided by the National AIDS and Sexually Transmitted Infections Control Programme (NASHCOP) and MUHAS. ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Ethical clearance for this study was obtained from the Directorate of Research and Publication of the Muhimbili University of Health and Allied Sciences (MUHAS) with reference number DA.282/298/01.C/. The NASHCOP administration granted permission to collect the data. Strict privacy and confidentiality were maintained throughout the study. Only de-identified data were used, and patient IDs, names, or other personally identifiable information were not collected from the patients. Data collection adhered to national and international ethical standards, particularly regarding the handling of sensitive health information. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes The Government of Tanzania owns the data used in this study through NASHCOP, and they are not publicly available. The de-identified analytical code and derived output can be made available from the corresponding author upon reasonable request and subject to NASHCOP approval.
OBJECTIVE:Undiagnosed diabetes leads to delayed treatment and increased risk of complications, exacerbating global disease burdens. In this study, we estimated the prevalence and absolute numbers of individuals with undiagnosed diabetes globally and across regions and quantified gaps in national diabetes detection efforts. RESEARCH DESIGN AND METHODS:We systematically compiled estimates of biomarker-based diabetes prevalence and self-reported diabetes diagnosis using 2003-2024 data from all eligible population-based studies and gray literature. We calculated proportions of individuals with undiagnosed diabetes and case numbers among adults aged 20-79 years. For countries without data, we extrapolated estimates using available data within the same geographic region and country income group. Country-level estimates were benchmarked against the World Health Organization 80% diagnosis target. RESULTS:We identified 193 data sources on undiagnosed diabetes from 109 countries. Across all 215 countries/territories, 42.8% of individuals with diabetes were undiagnosed in 2024, equating to 251.7 million (95% uncertainty interval [UI] 250.4-253.0 million) adults. Proportions undiagnosed ranged from 16.2% in Colombia to 90.4% in Burkina Faso and 29.1% in North America and the Caribbean to 72.6% in Africa. A larger proportion of individuals were undiagnosed in low-income (58.7%) compared with high-income countries (28.9%). Middle-income countries accounted for 206.0 million (95% UI 202.3-209.7 million) adults with undiagnosed diabetes (81.8% of all individuals), including 127.1 million (95% UI 121.2-133.0 million [or 50.5%]) adults in China, India, and Indonesia alone. Less than 5% of all countries attained diabetes diagnosis levels ≥80%. CONCLUSIONS:Substantial global variability in undiagnosed diabetes indicates opportunities to close existing care gaps, likely requiring context-specific solutions and investments.
Background Stroke is a leading cause of death and disability globally, with sub-Saharan Africa, bearing the greatest burden. Tanzania has only one active stroke registry, limiting evidence-based care and policy development. We aimed to expand the registry into a multi-centre study across eight major tertiary hospitals to generate baseline data for a national stroke registry. Methods From January to August 2024, we analysed de-identified data from adults (≥18 years) admitted with a World Health Organisation defined stroke. Data collected included demographics, risk factors, imaging, and in-hospital mortality. Logistic regression identified predictors of mortality. Results A total of 1000 patients were registered with a mean age 60.2±15 years and 56.2% (562/1000) were females. Most strokes occurred in those aged 50-69 years 46.3% (463/1000). Hypertension was the most common risk factor 90.1% (901/1000), followed by diabetes 13.1% (131/1000), prior stroke 10.6% (106/1000) and HIV infection 3.5% (35/1000). Haemorrhagic and ischaemic strokes accounted for 57.9% (579/1000) and 38.3% (383/1000) of cases, respectively; and 5% (19/383) of ischaemic strokes presented within 4.5 h from symptom onset. In-hospital mortality was 31.5% (315/1000), highest among patients aged 50–59 years (23.2%). Independent predictors of mortality included previous cardiac disease (aOR 2.15; 95% CI: 1.18–3.94) and haemorrhagic stroke (aOR 1.38; 95% CI: 1.12–2.02). Conclusions Stroke imposes a high burden in Tanzania, with substantial mortality and delayed presentation. Strengthening hypertension control, early stroke recognition, and organized stroke unit care are critical priorities. These findings provide foundational data for the national stroke surveillance initiative and support evidence-based planning for stroke prevention, acute care, and system readiness across Tanzania.
Trauma remains a major cause of death in low-resource settings, yet prospective evidence on time-to-care, triage accuracy, and critical care allocation is scarce. We quantified care delays and predictors of mortality and evaluated whether a pragmatic, physiology- and injury-based “critical status” classification is robust and scalable. We conducted a prospective, multi-center observational study of 8,440 trauma patients presenting to emergency departments of four national referral hospitals in Tanzania (June 2023–June 2024) with 14-day follow-up. Critical status was defined by physiologic derangement—respiratory rate < 8 or > 30/min, oxygen saturation < 90
Introduction:Residual COVID-19 sequelae create a public health concern as they add to the already heavy burden of non-communicable diseases. We set out to investigate the magnitude of residual cardiopulmonary manifestations postacute COVID-19 and their associated factors in Tanzania. Methods:This was an ambispective study conducted between 26 March 2021 and 30 July 2021, among 712 hospitalised adults confirmed with SARS-CoV-2 in five tertiary-level hospitals in Tanzania. Retrospective data were analysed to determine baseline characteristics of the patients during acute COVID-19 admission. This was linked to prospective data that were collected 2 years postacute hospitalisation with COVID-19 to determine cardiopulmonary complications among these patients. Radiological pulmonary abnormalities were assessed by contrasted CT scan. Lung function tests were measured using a spirometer. Pulmonary hypertension and heart failure were confirmed by a transthoracic echocardiography. Generalised estimating equations were used to assess the associations between sociodemographic factors, clinical characteristics, treatment modalities and residual cardiopulmonary sequelae. Results:About half 317/712 (44.5%) were diagnosed with residual cardiopulmonary complications. Approximately 54% of participants were male. Median age (IQR) was 60 (48-69) years. Cardiopulmonary sequelae were significantly associated with body mass index (BMI) ≥25.0 kg/m2 compared with those with BMI <25.0 kg/m2 (adjusted OR (aOR) (95% CI)=3.4 (2.47 to 4.84), p<0.001), smokers compared with non-smokers (aOR (95% CI)=2.39 (1.09 to 5.23), p=0.030] and among participants who did not receive steroids during the acute COVID-19 (aOR (95% CI)= 1.62(1.16 to 1.92), p=0.042]. Conclusion:The independent predictors of residual cardiopulmonary manifestations due to COVID-19 were overweight, cigarette smoking, hypoxia and non-use of steroids during the acute phase of COVID-19 disease. Public health interventions addressing weight reduction and smoking cessation in conjunction with steroid use in acute COVID-19 may largely reduce the incidence of cardiopulmonary long COVID-19.
