Real-world drug prescription patterns and mortality risk in patients hospitalized with acute heart hailure: impact of multimorbidity and heart failure phenotype

Abstract Background Real-world data on the impact of multimorbidity and of heart failure (HF) phenotype on longitudinal drug prescription patterns and dosing of guideline-recommended HF drugs and on mortality risk following acute HF (AHF) is sparse. Purpose We studied prescription rates (PR) and dosing of renin-angiotensin-system inhibitors (RASi), betablockers (BB), and mineralocorticoid-receptor antagonists (MRA) in patients hospitalized with AHF by comorbidity burden (CB) and HF phenotype on admission, at discharge, and at 6-month follow-up (6MFUP), and determined 12-month mortality risk. Methods Consecutive AHF patients enrolled on a 24/7 basis in a single-center prospective cohort study from 8/2014 to 2/2020, who had their left ventricular ejection fraction (LVEF) measured while in hospital, were eligible. Proportions of patients taking each HF drug class on admission, at discharge and at 6MFUP were determined in subgroups according to LVEF (<50% versus ≥50%) and CB (number present of: hypertension, chronic kidney disease, diabetes, anemia, chronic lung disease, psychiatric disorder, malignancy, atrial fibrillation, coronary disease, peripheral vascular disease, hypercholesterolemia). Results Of 943 patients, 47% had a LVEF <50%, and 53% had a LVEF of ≥50% (Table). CB was similar in both groups (p = 0.197), but patients with LVEF ≥50% were older and more often female (Table). Average PR of the 3 HF drug classes were higher in patients with LVEF <50% compared with ≥50% (mean count 2.2±0.8 vs. 1.7±0.8, respectively, p<0.001), but PR trajectories were similar: In patients with fewer comorbidities, PR were lower on admission but showed the greatest increase during the hospital stay (in particular for BB and RASi). PR improved in all patients irrespective of CB and reached a maximum at discharge (Figure). Still, many patients remained on suboptimal drug doses (irrespective of CB), and PR tended to decrease again at 6MFUP, except for MRA, which were often not prescribed at all even in patients with LVEF <50%. Absolute 12-month mortality risk was similar in the patients with LVEF <50% and ≥50%, but CB and number of HF drugs prescribed at discharge impacted significantly on 12-month mortality risk only in patients with LVEF <50% (Table). Conclusion In this study, PR of HF drugs increased during the in-hospital period, irrespective of HF phenotype. Neither PR, nor daily doses improved further after discharge. Prescription patterns were similar across spectra of LVEF and CB, but only in patients with LVEF <50% the number of comorbidities and of prescribed HF drug classes impacted mortality risk significantly. Given that timely implementation of foundational HF therapies (which today include also SGLT2 inhibitors) may improve outcomes, reasons for underprescription (e.g., suspected intolerability of individual drug classes or structural deficits of post-discharge HF management) in real world warrant further study and better control.TableFigure