While drug therapy and medical management improved markedly over the last decade, the basic clinical characteristics of the heart failure patient population treated at the study hospital changed little. This offers an excellent opportunity to study potential heart failure survival improvements for a general patient population.

Vital status follow-up through 2001 was obtained from the Social Security Death Index for all 2323 patients aged ≥65 years at the time of an initial, medically managed heart failure hospitalization between October 1989 and March 2000. Kaplan Meier survival probabilities were compared across 4 time periods in the 1990s. A Cox proportional hazards model was used to estimate age, sex, race and comorbidity-adjusted differences in survival among patients admitted in 1989–1991 and 3 subsequent multi-year periods.

There was an increase in the proportion of older female patients with more chronic conditions. Compared with patients admitted in 1989–1991, survival probabilities for patients admitted in 1999–2000 had improved about 5% at 30 days (to 95%) and 10% at 1 year in 1999–2001 (to 73.5%). For those admitted between 1989–1998, there was a 9% improvement over 1989–1991 at 5 years (to 36%). Hazards model results indicated that patients admitted in 1999–2000 had a relative risk of death only 66% that of patients admitted in 1989–1991 (P < .0001).

These findings provide evidence of modest but significant short-term survival improvements, particularly after 1998, when drug therapy had became optimal in the inpatient setting, patient education and discharge planning became better documented, and inpatient mortality rates had declined substantially.