Implementation of routine vs selective cardiac magnetic resonance (CMR) imaging was not found to improve etiological classification of heart failure (HF) in patients with nonischemic HF, according to a study published in Circulation.

Although CMR imaging is often recommended for patients with HF, it remains unclear whether the technology offers added benefit compared with transthoracic echocardiography. In a first of its kind study exploring the use of CMR for nonischemic HF diagnosis, investigators sought to compare routine vs selective CMR imaging, hypothesizing that routine imaging would yield more specific diagnoses/etiologies and would lead to better patient outcomes.

In this multicenter prospective randomized controlled OUTSMART-HF study ( Identifier: NCT01281384), 500 consecutive patients with new or worsening HF (mean age, 59 years; 69% men) were enrolled between 2011 and 2016. Participants were randomly assigned to receive routine CMR (echocardiography plus CMR in all patients; n=248) or selective CMR (echocardiography with or without CMR, at the physician’s discretion; n=252). The patients’ cardiovascular profile and the Meta-Analysis Global Group in Chronic Heart Failure (MAGGIC) risk score — used to estimate HF burden — were assessed at baseline. Imaging was interpreted, and the most likely HF etiology was assigned to each patient. Follow-up clinical assessments occurred at 3 and 12 months, at which point HF etiologies were recorded by practitioners.

The study’s primary outcome was the determination by the treating physician of a specific HF etiology, if possible at the 3-month follow-up. A secondary outcome were the 3- and 12-month clinical adverse events (death or cardiovascular hospitalization).

There was no significant difference between the routine and selective CMR approaches in terms of determining a specific HF etiology at the 3-month clinical follow-up (44% vs 50%, respectively; P =.22). The rates of HF etiological classification based on imaging interpretation were comparable in the routine vs selective group (34% vs 30%, respectively; P =.34). A nonprotocol CMR imaging was conducted on 24% of patients in the selective CMR group.

Patients with a specific vs nonspecific HF etiology had more clinical adverse events when the etiology was determined based on imaging (19% vs 12%, respectively; P =.02) but not when it was determined based on clinical assessment (15% vs 14%, respectively; P =.49).

Routine CMR was found to be superior to selective CMR in identifying specific HF etiologies (36% vs 20%, respectively; P <.001). The only factor found to predict the assessment of an etiology for HF at the 3-month follow-up was baseline HF etiology (relative risk, 4.36; 95% CI, 3.48-5.46; P <.0001).

Study limitations include a lack of prespecified HF etiology definitions and a lack of use of the most contemporary CMR techniques due to unavailability.

Related Articles

“This study confirmed that CMR should not be used routinely for the diagnosis of undifferentiated, non-ischemic heart failure,” noted the authors. They recommended that future research continue to evaluate the role of CMR imaging in nonischemic patient populations and explore potential clinical decision-making biases.

Conflicts of Interest Disclosures

RB is a consultant for and receives research funding from Lantheus Medical Imaging, Jubilant

DraxImage, and GE. JE reports study funding from Novartis and Servier and grants from Merck, Bayer, Trevena. and Amgen. BC receives research support from CV Diagnostix and Ausculsciences and educational support from TeraRecon Inc and has equity interest in GE. None of the other authors have competing interests to disclose.


Paterson DI, Wells G, Erthal F, et al. OUTSMART HF: a randomized controlled trial of routine versus selective cardiac magnetic resonance for patients with non-ischemic heart failure (IMAGE-HF 1B). Circulation. 2020. doi: 10.1161/circulationaha.119.043964