Heart Failure Hospitalizations Rose 30-51% in Advanced CKD While Mortality Fell Slower Than in General Population
Swedish registry data show heart failure hospitalizations climbing sharply in advanced CKD while mortality improvements trailed the general population. Low and stable use of proven therapies coincided with the rise. Observational design limits causal inference; adoption of newer agents needs prospective evaluation.
The observational analysis included 54,705 patients with advanced nondialysis CKD or dialysis dependence and compared event rates to the general Swedish population. Heart failure hospitalizations rose 51% in stage G4, 36% in stage G5, and 30% in hemodialysis, reaching 3.7-6.8 times population rates. Renin-angiotensin inhibitor and beta-blocker use stayed flat and low while diuretic prescriptions declined in nondialysis groups, coinciding with the hospitalization increases. Mortality from heart failure declined up to 66% in hemodialysis but lagged general-population improvements.
These trends occurred against a backdrop of delayed uptake of SGLT2 inhibitors and mineralocorticoid receptor antagonists in advanced CKD, therapies shown in the DAPA-CKD and EMPEROR trials to reduce heart failure events. Stable RASi prescribing suggests therapeutic inertia rather than contraindication concerns drove the gap. The parallel rise in hospitalizations despite falling mortality points to longer survival with more decompensated episodes, a pattern also seen in U.S. Medicare data after 2015.
Registry linkage cannot establish causation or capture outpatient management changes. Future studies must track SGLT2i and finerenone adoption rates alongside hospitalization metrics and test whether protocolized cardio-renal clinics narrow the excess burden observed here.
Lankinen et al.: Heart failure hospitalization rates in Swedish CKD G4-G5 will decline at least 15% by 2029 once SGLT2i coverage exceeds 40%.
Sources (3)
- [1]Primary Source(https://www.kireports.org/article/S2468-0244(26)106695/fulltext)
- [2]Supporting Source(https://jamanetwork.com/journals/jama/fullarticle/2801234)
- [3]Supporting Source(https://www.nejm.org/doi/full/10.1056/NEJMoa2024816)