Aims: We sought to describe the frequency, type and timing of clinical events, and delineate patterns in their transitions, after a first episode of heart failure (HF).
Methods and results: In 1036 Framingham participants with new-onset HF (mean age 79 years; 53% women), we used mixture models to estimate probabilities of, and time to cardiac death, other cardiovascular disease (CVD) death, recurrent HF, cardiac events and other CVD events, accounting for age, sex, HF type (preserved vs. reduced ejection fraction), and prevalent cardiac/CVD events. The most common first events after new-onset HF were cardiac (36%), recurrent HF (28%) and death (29%). Compared with recurrent HF (referent transition state), prevalent cardiac events were associated with higher odds of fatal [odds ratio (OR) 1.90, 95% confidence interval (CI) 1.11-3.23] and non-fatal (OR 2.13, 95% CI 1.52-3.00) cardiac events; prevalent CVD increased odds of other CVD death (OR 1.90, 95% CI 1.04-3.47). Among 715 participants without a fatal initial event, there were 3337 distinct epochs (inter-event time periods), with median 3.0 epochs/participant [49% cardiac (n = 1639); 27% recurrent HF (n = 912)]. Median inter-event times varied between 12 to 285 days (recurrent HF to other CVD death and non-fatal other CVD, respectively). Prior HF, cardiac and other CVD events significantly increased odds of developing the same event-type (OR ∼ 5-7-fold), with shortened time to recurrence, indicating 'rapid cycling loops' of the same event type. HF type did not impact the nature or timing of future events.
Conclusions: Comorbidities but not HF type impact clinical course of HF by influencing the type and timing of subsequent events, denoting 'natural history loops' within the overall HF population.
Keywords: Clinical course; Competing causes; Heart failure; Prognosis; Risk.
© 2020 European Society of Cardiology.