Efficacy of Blended Collaborative Care for Patients With Heart Failure and Comorbid Depression: A Randomized Clinical Trial

Abstract

Importance: Depression is often comorbid in patients with heart failure (HF) and is associated with worse clinical outcomes. However, depression generally goes unrecognized and untreated in this population.

Objective: To determine whether a blended collaborative care program for treating both HF and depression can improve clinical outcomes more than collaborative care for HF only and physicians' usual care (UC).

Design, setting, and participants: This 3-arm, single-blind, randomized effectiveness trial recruited 756 participants with HF with reduced left ventricular ejection fraction (<45%) from 8 university-based and community hospitals in southwestern Pennsylvania between March 2014 and October 2017 and observed them until November 2018. Participants included 629 who screened positive for depression during hospitalization and 2 weeks postdischarge and 127 randomly sampled participants without depression to facilitate further comparisons. Key analyses were performed November 2018 to March 2019.

Interventions: Separate physician-supervised nurse teams provided either 12 months of collaborative care for HF and depression ("blended" care) or collaborative care for HF only (enhanced UC [eUC]).

Main outcomes and measures: The primary outcome was mental health-related quality of life (mHRQOL) as measured by the Mental Component Summary of the 12-item Short Form Health Survey (MCS-12). Secondary outcomes included mood, physical function, HF pharmacotherapy use, rehospitalizations, and mortality.

Results: Of the 756 participants (mean [SD] age, 64.0 [13.0] years; 425 [56%] male), those with depression reported worse mHRQOL, mood, and physical function but were otherwise similar to those without depression (eg, mean left ventricular ejection fraction, 28%). At 12 months, blended care participants reported a 4.47-point improvement on the MCS-12 vs UC (95% CI, 1.65 to 7.28; P = .002), but similar scores as the eUC arm (1.12; 95% CI, -1.15 to 3.40; P = .33). Blended care participants also reported better mood than UC participants (Patient-Reported Outcomes Measurement Information System-Depression effect size, 0.47; 95% CI, 0.28 to 0.67) and eUC participants (0.24; 95% CI, 0.07 to 0.41), but physical function, HF pharmacotherapy use, rehospitalizations, and mortality were similar by both baseline depression and randomization status.

Conclusions and relevance: In this randomized clinical trial of patients with HF and depression, telephone-delivered blended collaborative care produced modest improvements in mHRQOL, the primary outcome, on the MCS-12 vs UC but not eUC. Although blended care did not differentially affect rehospitalization and mortality, it improved mood better than eUC and UC and thus may enable organized health care systems to provide effective first-line depression care to medically complex patients.

Trial registration: ClinicalTrials.gov Identifier: NCT02044211.

Conflict of interest statement

Conflict of Interest Disclosures: Drs Rollman, Anderson, and Muldoon reported receiving grants from the National Heart, Lung, and Blood Institute during the conduct of the study. Dr Jakicic reported receiving grants from the National Institutes of Health during the conduct of the study; and personal fees from WW International, Inc (scientific advisory board) and Naturally Slim (advisory board) outside the submitted work. Dr Karp reported serving as an adviser for Aifred Health and NightWare; receiving medication supplies for investigator-initiated studies from Pfizer and Indivior; receiving compensation for work on the editorial boards of the Journal of Clinical Psychiatry and American Journal of Geriatric Psychiatry; and preparing and delivering a webinar for Otsuka. No other disclosures were reported.

Figures

Figure 1.. Flowchart of Screening, Enrollment, Randomization, and Follow-upa

aStudy participants who missed a 3- or 6-month follow-up assessment remained eligible to complete a later assessment. We display the cumulative number of withdrawn and deceased patients in parentheses. There were no differences in follow-up rates between randomized arms or by baseline depression status (eTable 1 in Supplement 3). HIPAA indicates Health Insurance Portability and Accountability Act; NYHA, New York Heart Association; PHQ, Patient Health Questionnaire; Tx, treatment; UC, usual care.

Figure 2.. Twelve-Month Adjusted Differences in Mental Component Summary of the 12-Item Short Form Health Survey Scores for All Randomized Patients With Depression and by Sex

Subgroup analysis by sex revealed no significant subgroup effect (P = .57 for 3-way interaction). eUC indicates enhanced usual care; UC, usual care.

Figure 3.. Estimated Mental Component Summary of the 12-Item Short Form Health Survey (MCS-12) Scores by Time Point for All Patients, Men, and Women

Error bars indicate 95% CIs. eUC indicates enhanced usual care; ND, no depression; UC, usual care.