Does motivational interviewing reduce health services use and mortality in heart failure patients? A secondary analysis of the MOTIVATE-HF trial

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): Center of Excellence for Nursing Scholarship, Rome, Italy Background. Health services use and mortality are high among patients with heart failure (HF). The MOTIVATE-HF was a randomized controlled trial in which we demonstrated that Motivational Interviewing (MI) can effectively improve HF patient self-care (primary outcome). However, it is still unclear if MI can reduce health services use and mortality. Purpose. To test the efficacy of MI in reducing health services use (emergency services and hospitalizations), and all-cause mortality in HF patients. Methods. The MOTIVATE-HF was a multicenter parallel trial (1:1:1) with randomization of patient and informal caregiver dyads across three arms: Arm 1, MI administered to patients; Arm 2, MI administered to patients and caregivers; Arm 3, standard of care. Participants were enrolled from three Italian HF specialty centers. Patients were included only if they had HF in NYHA functional class II-IV and excluded if they had a recent coronary event, lived in a residential setting, or had severe cognitive impairment. Informal caregivers were nominated by the patients as the person mainly responsible for their care. The intervention consisted of one face-to-face MI session followed by 3 telephone contacts within 2 months from enrollment. Follow-up and reassessment occurred after 3-, 6-, 9- and 12-months post-randomization. A longitudinal generalized mixed model and a Cox proportional-hazard model were fitted to compare rates of health care services use and mortality among the three study arms during follow-ups. Results. The sample of 510 HF patients (median age 74 years; IQR 65-82; 58% male) was distributed as follows: Arm 1, n = 155; Arm 2, n = 177; and Arm 3, n = 178. Within the 12 months" period, 25 patients (16.1%) in Arm 1, 30 patients (17%) in Arm 2, and 20 patients (11.2%) in Arm 3 used health services at least once (p = 0.25) without significant differences at any follow up. Regarding all-cause mortality, at 3 months, 3 patients (1.9%) in Arm 1, 1 patient (0.6%) in Arm 2, and 9 patients (5.1%) in Arm 3 had died (p = 0.02). At 6 months, 5 patients (3.2%) in Arm 1, 1 patient (1.7%) in Arm 2 and 12 patients (6.7%) in Arm 3 had died (p = 0.05). No significant group differences were observed at 9 or 12 months. Mortality was lower in Arm 2 compared to Arm 3 in the first 3 months (HR = 0.112, 95% CI: 0.014–0.882, p = 0.038), whereas no difference was detected in the following months (p = 0.699). Mortality was also lower in Arm 1 compared to Arm 3 in the first 3 months, but this reduction did not reach statistical significance (HR = 0.383, 95% CI:0.104–1.414, p = 0.155). Conclusions. This analysis indicates that MI was not effective in reducing health services use in our sample, but it was effective in reducing all-cause mortality, at least in the short-term period. MI could be an inexpensive intervention to reduce HF mortality but further studies are needed to see if more frequent MI interventions could produce greater sustained reduction in health services use and mortality.