Home | About SPRM


The Seattle Proportional Risk Model (SPRM) is a calculator that provides the estimated proportion of sudden death (that is sudden death/all cause mortality) in patients with heart failure with reduced ejection fraction (primarily EF≤35%). The SPRM developers believe this approach is superior to predicting the annual rate of sudden death due to the highly variable competing risk of non-sudden death. Non sudden death is highly variable and is not accounted for in models that predict sudden death and heart failure death as these do not include non-cardiac modes of death, which are increased in older patients with comorbidities.

The Seattle Proportional Risk Model was developed in 9,885 heart failure patients with 2,552 deaths. About 48% of the deaths were classified as sudden death. The model identified 10 clinical variables that were associated with a higher proportion of sudden death; younger age, lower NYHA class, lower ejection fraction, male sex, higher body mass index, digoxin use, systolic blood pressure closer to 140 mmHg, lack of diabetes and normal serum sodium and creatinine. The proportion of sudden death is estimated from the 10 clinical variables.

We have demonstrated that the benefit of a primary prevention ICD varies with the proportion of sudden death with the greatest benefit in those patients with ≥50% estimated proportion of sudden death. Among these patients in the Sudden Cardiac Death Heart Failure Trial (~75% of all patients in the trial), the benefit of the ICD varied from a 57% to 95% reduction in sudden death and a 28% to 66% reduction in all-cause mortality. The variable ICD benefit from a Cox model among all patients in the SCD-HEeFT is used to create the variable ICD hazard ratio that is presented in this website.

If a user enters an annual mortality with medical therapy (and potentially with CRT-P) as provided from the Seattle Heart Failure Model or potentially other models, the SPRM will provide an annual mortality with and without an ICD and the projected life expectancy (Gompertz Model) with and without an ICD. It is likely the benefit of an ICD will diminish over longer periods of time as the patient ages and as heart failure progresses. Any future changes in the medical course of the patient are not taken into account by the SPRM. Estimates of ICD benefit beyond the 5 years in SCD-HeFT are less certain. For longer life expectancies, the patient may require more than one ICD.

Use of this model may provide a patient level estimate of a primary prevention ICD and may be used in shared decision making as required by CMS. However, CMS has not approved the use of this model for this purpose.

This information is used to estimate the variable benefit of a primary prevention ICD based on the application of this model in the Sudden Cardiac Death - Heart Failure Trial. The variable ICD benefit is has also been demonstrated in the randomized ICS in the DANISH trial in non-ischemic heart failure patients and with observational ICDs in HF-ACTION and NCDR-ICD cohorts. The SPRM may also provide the incremental benefit of adding an ICD to a CRT device (i.e. CRT-D vs. CRT-P) as was demonstrated in the DANISH trial. SPRM is not designed to provide the overall benefit of CRT-P vs. no device or CRT-D vs. no device.

The Javascript webapp to calculate the score was designed by Drs. Linker and Levy and programmed by Dr. David T. Linker.
Terms of Use | Copyright 2018 University of Washington. All rights reserved.