Predicting heart failure survival: Value of a mortality risk score

Congestive heart failure does not necessarily "go away" in the elderly. It is not that heart failure patients are dying in middle age before they become elderly; quite the contrary. The growing prevalence of heart failure in the elderly is due in part to our doing a reasonably good job of treating the syndrome—at least in its milder forms—in middle-aged patients, who then age. In addition, the elderly are the fastest growing segment of the US population, and this group often develops new-onset heart failure as it ages. As Thomas and Rich have previously documented, heart failure is the leading cause of hospitalization in the subgroup over 65 years.¹ In their

incisive article in this issue

, Rich et al point out that determining effective outpatient management strategies for older patients depends on first determining predictors of long-term survival in a cohort of elderly patients hospitalized with heart failure.

Accordingly, Rich and colleagues enrolled 282 patients ≥70 years in a prospective, randomized, multidisciplinary intervention trial. Data from that trial was used to develop a mortality risk score based on what they termed "readily available clinical parameters." Statistical analysis revealed 7 clinical variables that were independent predictors of reduced survival time. These were (beside increasing age) reduced serum sodium level, presence of coronary artery disease, dementia, peripheral vascular disease, increasing systolic blood pressure, and increasing blood urea nitrogen levels. With the risk score that was developed—and the mortality rates correlated with this score—it became apparent that the 1-year mortality rates in these patients varied markedly depending on the score. For example, when the score was low (0-1), the 1-year mortality was 9%; when it was moderate (2-3), 22%; and when it was high (4 or above), 73%. The authors look at the risk score approach as one way to reliably estimate prognosis in a syndrome with considerable heterogeneity in survival, ie, 25% of patients die within 1 year, and only 25% survive for more than 5 years. At a bare minimum this score can help decide who gets costly new devices installed (such as implantable defibrillators² and biventricular pacemakers) and who doesn't. The length of the study (14-year follow-up) makes up for the relatively small number of patients, and this means clinicians should pay attention to its results.

Even though others have studied heart failure survival,³ the few exclusion criteria used in this study make it especially relevant to the type of elderly patient seen in routine clinical practice. I also want to compliment the authors for the detailed case study they submitted with their article since it illustrates their conclusions in a very practical manner.