HFSA ePoster Library

Introduction: Poor medication adherence is associated with increased hospitalization and mortality rates in patients with heart failure with reduced ejection fraction (HFrEF), but its association with health status is unknown.
Methods: CHAMP-HF is a multicenter observational study of outpatients with HFrEF (EF ≤40%). Medication adherence was assessed using the Morisky Medication Adherence Scale 4-Item© (MMAS-4) and any “yes” response was taken as a positive screen for non-adherence at enrollment. Health status was assessed by the Kansas City Cardiomyopathy Questionnaire Overall Summary score (KCCQ-OS) at baseline and 12-months. Robust linear regression was used to assess differences in health status at baseline, as well as change from baseline to 12 months. Models were serially adjusted for sociodemographic characteristics and comorbidities (Model 2), and HF medications (Model 3).
Results: Of 4,780 patients, 3,245 (67.9%) were classified as adherent and were more likely to be older (69 vs 65 years), white (75.1% vs 71.1%), more educated, and on fewer HF medications (3[2-4] vs 3[3-4]). They were also less likely to smoke (18.4% vs 23%) and have depression (24.9% vs 29%). Non-adherence to HF medications was associated with lower mean baseline KCCQ-OS (60.2 vs 66.8, p<0.001) that persisted after adjusting for baseline characteristics, comorbidities, and HF medications (Table 1A). Non-adherent patients also had a lower mean KCCQ-OS 12-months later (71.4 vs 77.1, p<0.001), with less improvement over 12-months in fully-adjusted analyses (Table 1B).
Conclusions: Poor medication adherence was associated with both worse baseline health status and less health status improvement over 12-months. Interventions that improve medication adherence should be tested to improve patients’ health status.

Table 1A: Association between Baseline Morisky Medication Adherence Scale 4-item© status and Baseline Kansas City Cardiomyopathy Questionnaire Overall Summary scores
Model	Difference in mean baseline KCCQ-OS Score	P-Value
Model 1
Non-adherent vs Adherent	-6.63 (-8.24, -5.02)	<0.001
Model 2*
Non-adherent vs Adherent	-5.90 (-7.42, -4.37)	<0.001
Model 3†
Non-adherent vs Adherent	-5.85 (-7.29, -4.40)	<0.001
Table 1B: Association between Baseline Morisky Medication Adherence Scale 4-item© status and Change in Kansas City Cardiomyopathy Questionnaire Overall Summary scores from Baseline to 12-Months
Model	Difference in Mean Change in KCCQ Score ∆	P-Value
Model 1⁋
∆ Non-adherent vs ∆ Adherent	-1.39 (-2.53, -0.24)	0.017
Model 2*
∆ Non-adherent vs ∆ Adherent	-1.64 (-2.76, -0.51)	0.004
Model 3†
∆ Non-adherent vs ∆ Adherent	-1.70 (-2.81, -0.59)	0.003
⁋ Adjusted for baseline KCCQ-OS
* Model 2: Model 1 + age, sex, ethnicity/race, body mass index, insurance status, level of education, household income, employment status, hypertension, dyslipidemia, chronic obstructive pulmonary disease, chronic kidney disease, depression, atrial fibrillation, ventricular arrhythmias, history of HF hospitalization in the year before enrollment
† Model 3: Model 2 + HF medications
∆ is the mean change in KCCQ-OS (12-month score - baseline score). Negative estimated mean difference in ∆ correspond to unfavorable changes on average for ∆ non-adherent versus ∆ adherent. The MMAS (4-item) content, name, and trademarks are protected by US copyright and trademark laws.  Permission for use of the scale and its coding is required.  A license agreement is available from Donald E. Morisky, ScD, ScM, MSPH, President, MMAR LLC, 294 Lindura Ct., Las Vegas, NV 89138, donald.morisky@moriskyscale.com