Heart failure clinic follow-up and outcomes after hospitalisation: a propensity-weighted cohort study
J Rolland, E Stenhede, T Ovrebotten, H Nahoui, H Schirmer, K Broch, P L Myhre, K BergeAbstract
Background
Guideline-directed medical therapy (GDMT) for heart failure (HF) reduces mortality and HF hospitalisations and improves quality of life. However, structured follow-up with up-titration of GDMT remains suboptimal, and only 20% of patients hospitalised for acute HF receive follow-up at HF outpatient clinics in Norway.
Purpose
To assess differences in patient characteristics and outcomes after hospitalisation for HF between patients who receive follow-up at an HF outpatient clinic and those who do not.
Methods
Adults discharged with either a primary or secondary diagnosis of HF between 2020 and 2024 at a large tertiary care hospital in Norway were identified using a predefined list of ICD-10 codes. HF clinic attendance was identified using a dedicated ICD-10 reimbursement code. The first HF clinic visit following hospital discharge was used as the index date, and the corresponding number of days from discharge was used for patients who did not attend the HF clinic. Groups were balanced using inverse probability of treatment weighting (IPTW) with overlap weights (average treatment effect in the overlap population, [ATO]) based on 26 predefined baseline variables.
Results
In total, 5,013 patients hospitalised for HF were included, of whom 793 (16 %) attended the HF clinic, median 62 days (IQR 40-106) after discharge (Table). The mean age was 75 ± 12 years and 40% were female. Patients attending the HF clinic were younger (68±13 vs 77±12 years), more often male (71% vs 48%), and had lower left ventricular ejection fraction (LVEF; 40 % [32-49] vs 51 % [43-58]) than those not attending the HF clinic. They had a lower prevalence of non-cardiac comorbidities, but a higher prevalence of ischemic heart disease and prior myocardial infarction. At discharge from the index hospitalisation, the HF clinic patients were more frequently treated with GDMT and were less often prescribed loop diuretics. Independent predictors of HF clinic follow-up included younger age, lower LVEF, larger left ventricular volumes, higher NT-proBNP levels, lower potassium levels, and atrial fibrillation. After overlap weighting, all baseline characteristics were well-balanced between the patients who attended the HF clinic and those who did not. HF clinic attendance was associated with a 44% lower risk of mortality compared with non-attendance (HR 0.56, 95% CI 0.45–0.68; p<0.001).
Conclusion
Patients receiving follow-up at HF clinics after hospitalisations were younger, had lower LVEF, fewer non-cardiac comorbidities, and were more likely to be treated with GDMT. In the propensity-weighted cohort, HF clinic follow-up was associated with lower mortality. These findings suggest that structured HF follow-up may improve outcomes and should be considered for more patients.Table:Baseline characteristicsFigure:Kaplan-Meier Curves, ATO