Sliwa K, et al. Aetiology, management, and outcomes of acute heart failure in 17 African countries (THESUS-HF II): a prospective, multicentre, observational cohort study. Lancet. 2026;:S0140-6736(26)01657-0. PMID 42669305¶
One-paragraph summary¶
THESUS-HF II enrolled 1,578 adults presenting with acute heart failure at 50 hospitals across 17 African countries between 1 July 2024 and 31 July 2025, using seven 24-hour surveillance periods within an 8-week site-specific window. Acute HF was identified by acute dyspnoea and congestion supported by electrocardiography, chest radiography and point-of-care echocardiography. Median age was 56.0 years (IQR 40.0–69.0); 793 of 1,570 (50.5%) were female and 1,533 of 1,567 (97.8%) were of African descent. Hypertensive heart disease was the leading aetiology at 576 of 1,578 (36.5%), ahead of cardiomyopathy (370, 23.4%) and ischaemic heart disease (367, 23.3%), and 1,007 of 1,568 presentations (64.2%) were de novo. Discharge prescription of foundational therapy was high but target doses were reached in fewer than half of patients. In-hospital mortality was 8.7% (95% CI 7.3–10.2), 30-day all-cause mortality 11.1% (9.6–12.9) and 180-day mortality 20.6% (18.4–23.0). The study was registered with the Pan African Clinical Trial Registry (PACTR202410553810557).
Design¶
| Element | Detail |
|---|---|
| Design | Prospective, multicentre, observational cohort |
| Sites | 50 hospitals, 17 African countries |
| Enrolment | 1,741 assessed; 1,578 included; 1 Jul 2024 – 31 Jul 2025 |
| Case ascertainment | Acute dyspnoea and congestion plus ECG, chest radiography and point-of-care echocardiography |
| Sampling frame | Seven 24-hour surveillance periods within an 8-week site window |
| Outcomes | Presentation rate per catchment population, treatment patterns, readmission, all-cause mortality to 180 days |
| Funding | Cape Heart Institute (University of Cape Town); Hippocrate Foundation; The Heart Initiative; Pan African Society of Cardiology |
Key findings¶
- Hypertensive heart disease was the leading aetiology: 576 of 1,578 (36.5%); cardiomyopathy 370 (23.4%); ischaemic heart disease 367 (23.3%).
- De novo acute HF accounted for 1,007 of 1,568 presentations (64.2%).
- Weekly acute-HF presentation rate 0.03 per 10,000 adults (95% CI 0.02–0.05); 0.02 (0.01–0.04) for de novo presentations.
- Median ward stay 6.0 days (IQR 4.0–10.0); in-hospital mortality 8.7% (7.3–10.2; 134 of 1,542).
- Discharge therapy: RAAS inhibitors 1,014 of 1,373 (73.9%), beta-blockers 1,053 of 1,369 (76.9%), MRAs 982 of 1,366 (71.9%), SGLT2 inhibitors 743 of 1,359 (54.7%) — target doses reached in fewer than half of patients.
- All-cause mortality 11.1% (9.6–12.9) at 30 days and 20.6% (18.4–23.0) at 180 days, in a cohort of median age 56 years.
- Mortality did not differ across aetiological groups at 30 days (log-rank p=0.39) but did at 180 days (p=0.019); the authors conclude short-term outcomes were not driven by demographic, aetiological or echocardiographic factors.
Limitations¶
- Hospital-presentation series, not a population prevalence study: the denominator is people who reached a participating hospital, and access differs sharply across the 17 countries.
- 528 of 1,567 patients (33.7%) were lost to follow-up before 180 days, which bounds every time-to-event estimate.
- The apparent aetiological shift relative to the original THESUS-HF cohort may partly reflect changed ascertainment and diagnostic capability rather than changed disease distribution — the authors say so explicitly.
- Aetiology was assigned clinically in a cohort whose acute-HF ascertainment was supported by electrocardiography, chest radiography and point-of-care echocardiography. The PubMed record does not describe a reproducible HHD phenotype definition, so the 36.5% estimate should be treated as clinical attribution rather than imaging-defined prevalence.
- Surveillance-window sampling estimates presentation rates but is not continuous enumeration.
Why it matters¶
The HHD literature has a structural asymmetry: burden is often reported through modeled GBD estimates, while THESUS-HF II supplies a prospective hospital-cohort estimate based on clinically assigned aetiology. In the aggregate cohort, HHD was the single most common assigned cause of acute HF presentation. Discharge prescription of guideline-directed therapy was frequent, but target doses were reached in fewer than half of patients; estimated 180-day mortality was 20.6% in a cohort with a median age of 56, with 33.7% lost to follow-up before 180 days. These findings raise implementation questions about titration, follow-up and access alongside drug selection. The 36.5% figure should be quoted as a share of hospital presentations, never as population prevalence, and never merged with GBD's modeled HHD partition.
Cited by wiki pages¶
- epidemiology and burden
- heart-failure management