A RANDOMIZED CONTROLLED CLINICAL TRIAL OF A MULTI-DISCIPLINARY INPATIENT TEAM INTERVENTION IN PATIENTS WITH HEART FAILURE

BACKGROUND: Patients with congestive heart failure have poor outcomes with a high rate of hospital readmission, death and functional compromise. OBJECTIVE: Whether a non-pharmacologic intervention consisting of compliance enhancing interventions by a multi-disciplinary heath care team improves medication compliance, readmissions, and mortality. DESIGN: Randomized controlled trial. SETTING: Acute medical and surgical units at a teaching hospital in London, Canada. PARTICIPANTS: 134 patients 18 years of age or older with a clinical diagnosis and cardiac ejection fraction of <40%. METHODS: Patients stratified by whether or not heart failure was the primary reason for admission and randomized to receive either usual care (Control) or an enhanced educational intervention (Intervention). OUTCOME MEASURES: Non-compliance defined as compliance less than 0.8 for each heart failure medication. We also measured compliance using, the total number of days that a patient missed taking that drug, the “cumulative gap in therapy”. For both measures of compliance we calculated the relative risk (RR) for each medication. Knowledge gain was measured using a validated knowledge questionnaire. We measured health-related quality of life (HRQoL) using the Minnesota Living with Heart Failure Questionnaire (MLHFQ) and the SF-36, and conducted a time to first event analysis of a composite endpoint including all cause mortality, and hospital visits or readmissions. RESULTS: The RR of noncompliance for all HF medications in Intervention ranged from 0.78 [0.33 – 1.89] for ACE-I to 1.14 [0.18 – 7.08] for Spironolactone. The RR of a cumulative gap in the Intervention arm ranged from 0.76 [0.27 – 2.16 ] for spironolactone to 1.19 [0.68 – 2.09] for digoxin. The proportion of patients who had gaps in therapy ranged from 28.9% I, 37.6% C for spironolactone to 48.2% I, 55.9% C for ACE-I. Compliance decreased over time for both groups (p=0.051). Intervention had higher knowledge scores compared to Control at discharge (p=0.05). The intervention had no significant impact on SF-36 function summary scales. The MLHFQ showed differences between the groups over time at certain time points (p=0.039) that were not sustained over the duration of the trial. The hazard ratio of the composite endpoint was 0.85 [0.55 – 1.30]. CONCLUSIONS: A patient-centred multi-facted intervention demonstrated a trend toward lower RR of non-compliance and gaps in therapy in the Intervention patients, improved knowledge acquisition, showed transient effects in some aspects of HRQoL, and showed a trend in reducing a composite endpoint.