Dallas, TX - A prediction model invented to stratify patients with heart failure according to all-cause mortality risk also seems to predict their mode of death, whether from sudden cardiac death (SCD) or from progressive pump failure, suggests an analysis [1]. The Seattle Heart Failure Model (SHFM) [2], which showed signs of being more discriminating for HF mortality outcomes than the traditional New York Heart Association functional class, could potentially be used to guide treatment decisions, according to the authors.

In any group with the same level of SHFM-defined risk, there may be patients at any level of NYHA class, "but conversely, within each NYHA class, there are dramatically different risks according to the Seattle HF model score," lead author Dr Dariush Mozaffarian (Harvard School of Public Health, Boston, MA) told heartwire. One of the study's messages for clinicians, he said, is that "if they want to figure out a patient's risk, they should probably consider this model in addition to just looking at NYHA class."

The analysis encompassed 10 538 patients with NYHA class 2-4 "predominantly systolic" heart failure enrolled in six prospective trials and registries, for whom all the necessary data were on record and who had not received an implantable cardioverter-defibrillator (ICD). Over a follow-up averaging 1.6 years, the annual rate of death from any cause was 12%, which included 6.1% for SCD and 4.1% for pump failure. The patients had been enrolled in the PRAISE, ELITE-2, Val-HeFT, or RENAISSANCE trials or the IN-CHF registry or were members of a University of Washington prospective heart-failure cohort.

The effect of SHFM score on mode of death was attenuated somewhat among the 30% of the patients on beta blockers compared with the rest. Still, Mozaffarian observed, the score's predictive power remained strongly significant for either mode of death regardless of beta-blocker use.

Importantly, according to the researcher, knowing a patient's SHFM-based risks for both SCD and pump failure could perhaps one day sharpen management decisions. The scores, Mozaffarian said, "might help identify treatments that are good for a specific cause of death but where there's not a high competing risk from another cause of death."

As an example, ICDs clearly wouldn't be very cost effective for patients with a low projected risk of sudden death. On the other hand, Mozaffarian observed, a patient could have SHFM scores suggesting a high risk for SCD but an even higher risk for pump failure, such that an ICD would still not be very cost-effective strategy. "You want a high enough absolute risk to justify the treatment, but [also] low enough competing risks to justify the treatment."

But this application of the SHFM, he cautioned, would have to be validated in prospective studies before it could be put into use.