Introduction:
Hospitalization is a complex health exposure and the period immediately following acute-care hospitalization is a high-risk state. Self-rated health is a subjective indicator of health and the long-term trends in self-rated health after hospitalization are not well characterized.
Hypotheses:
1. Self-rated health decreases after hospitalization, with only partial recovery in the following years. 2. Poor self-rated health after hospitalization is associated with increased mortality.
Methods:
We analyzed 13,758 participants in the Atherosclerosis Risk in Communities (ARIC) Study with at least 1 hospitalization. Self-rated health was assessed annually and rated on a 4-point scale as follows: “Over the past year, compared to other people your age, would you say that your health has been excellent(=4), good(=3), fair(=2) or poor(=1)?" Using Cox regression and Kaplan-Meier methods, we evaluated mortality after hospitalization for myocardial infarction, congestive heart failure, cerebrovascular disease, pneumonia or diabetes mellitus with complications.
Results:
The mean self-rated health the year prior to hospitalization was 2.82 and the nadir of self-rated health was 2.62, occurring 1 year after hospitalization (Fig 1a). As compared to “excellent” self-rated health, “poor” self-rated health after any hospitalization was strongly associated with mortality (HR 4.65, 95% CI 4.27-5.07). Corresponding HRs (95% CI) for mortality post-hospitalization were 3.12 (2.30-4.22) for acute myocardial infarction, 3.08 (2.39-3.96) for congestive heart failure, 2.15 (1.43-3.23) for acute cerebrovascular disease, 4.54 (3.39-6.09) for pneumonia, and 3.32 (2.35-4.69) for diabetes mellitus with complications (Fig 1b).
Conclusion:
Mean self-rated health decreases significantly after hospitalization and worse self-rated health is associated with higher mortality. Self-rated health is an easily obtained patient centered outcome with valuable prognostic information.
Trends in mortality, co-morbidity and treatment after acute myocardial infarction in patients with rheumatoid arthritis 1998–2013
