Impact of HF Medication Adherence on Hospitalization and Mortality Outcomes

Study Questions:

Is the association between heart failure (HF) symptoms and the composite endpoint of hospitalization/death explained by medication adherence?


A secondary analysis was conducted of 2 tandem, prospective studies with parallel methods on medication adherence using the Medication Event Monitoring System (MEMS), with up to 3.5 years of follow-up in patients with HF. Patients were enrolled in outpatient clinics and inpatient hospital stays for HF. Demographic, clinical, symptom, medication adherence, and outcomes data were analyzed for 219 patients enrolled in the parent study. Patient data were analyzed according to those who had symptoms and those who did not have symptoms during the data collection period. Two medications, beta-blockers and angiotensin-converting enzyme inhibitors, were used to measure HF medication adherence. The cutoff of 88% was used to determine adherence.


Patients from the 2 prospective studies who reported having symptoms during the previous 3 months were younger, had greater body mass index, and were less likely to take angiotensin-converting enzyme inhibitors than those who reported not having symptoms. Ninety patients (41%) were non-adherent to HF medications: 66 (46%) with symptoms during the previous 3 months and 24 (32%) without symptoms. In all, 4 cardiac deaths and 46 cardiac-related hospitalizations occurred during the follow-up period: 38 events for patients who had symptoms during the previous 3 months and 12 events for patients without symptoms. Kaplan-Meier survival estimates of time-to-event plots demonstrated that cardiac event-free survival was worse among patients with symptoms than those without symptoms (p = 0.039). After modelling the data using medication adherence, HF symptoms were not significant predictors of event-free survival (p = 0.091), indicating mediation.


In this study, patients who adhered to their prescribed HF medications had fewer HF symptoms, and medication adherence mediated the relationship between HF symptoms and cardiac event-free survival.


In this secondary analysis, sample data from two medication adherence studies were combined to determine the influence of medication adherence on symptoms and the health outcomes of hospital readmission and death. Although the MEMS is a valid and reliable measure of medication adherence, individual differences in the samples might have produced bias in the parent observational study. For example, left ventricular ejection fraction, HF etiology, and diabetes were controlled in the analysis, but patients with HF often have multiple co-morbid conditions that might affect HF outcomes. Patients who were recruited from the outpatient setting were younger and might have been healthier than those enrolled in the in-patient hospital. Nonetheless, this analysis is an important step in better understanding variability in medication adherence and potential impact on worsening symptoms and outcomes.

Keywords: Heart Failure, Medication Adherence, Patient Readmission, Outpatients, Inpatients, Hospitalization, Adrenergic beta-Antagonists, Angiotensin-Converting Enzyme Inhibitors

< Back to Listings