Abstract
<title>Abstract</title> <p> <bold>Background</bold> Stroke is the second leading cause of death and the main cause of acquired adult disability worldwide, with a disproportionate share of its burden concentrated in low- and middle-income countries. In Angola, hospital-based evidence remains fragmented, and a systematic description of local stroke cohorts is a prerequisite for targeted prevention, organised acute-care pathways and appropriately resourced specialist services. <bold>Methods</bold> This single-centre, retrospective observational study used the administrative database of Hospital do Prenda (Luanda, Angola), covering 483 stroke admissions recorded between 1 March 2023 and 31 December 2025. Before analysis the database underwent an explicit curation step: harmonisation of column names and diagnostic labels across annual sheets, conversion of age to a numeric scale, recalculation of length of stay, reconciliation of the discharge disposition using adjacent fields when possible, and a light consistency check for duplicates and diagnosis/sub-diagnosis concordance. Quantitative results are reported as absolute and relative frequencies for categorical variables and as measures of central tendency and dispersion for age and length ofstay after exclusion ofbiologically implausible values. <bold>Results</bold> Across the 33-month study window, admissions rose from 120 in 2023 to 142 in 2024 and 221 in 2025 (approximately 84% growth vs 2023). Female patients predominated (308 of 483; 63.8%) over male patients (172; 35.7%); three records lacked sex. Mean age was 54.6 years (SD 15.0; median 53; IQR 45.0–64.5; range 12–100). Cranial computed tomography was performed in 95.2% of episodes (460/483). The subtype distribution was haemorrhagic stroke in 48.7% (235/483), ischaemic stroke in 45.5% (220/483) and unspecified stroke in 5.8% (28/483). Median length of stay after curation was 6 days (IQR 3–8). Hospital outcomes were discharge in 83.6% (404/483), death in15.5% (75/483) and transfer in 0.8% (4/483). Among the 479 episodes with a comparable clinical outcome, overall in-hospital mortality was 15.7% (Wilson exact onclusions This series confirms a relatively young stroke profile with female predominance and a slight excess of haemorrhagic over ischaemic stroke at a tertiary Angolan centre. The high rate of cranial CT use and the observed mortality fall within the range expected for African inpatient cohorts and argue for a structured stroke pathway, better recording of comorbidities and multi-centre studies in the country. </p>