Background Quality of life (QOL) is an increasingly important outcome measu
re after hospital admission for acute myocardial infarction (AMI). However,
the ability to adjust these outcomes for differences between compared grou
ps of patients is limited because the predictors of QOL after AMI are unkno
wn.
Methods To identify any clinical, demographic, and psychosocial characteris
tics of patients at admission that were independent predictors of QOL 6 mon
ths and I year after AMI, we measured physical and mental QOL (Short Form-3
6 Physical and Mental Component summary scores) and overall QOL (EuroQol he
alth perception scale) in a prospective cohort of 587 patients admitted at
10 hospitals in Quebec. A set of plausible multivariate linear regression m
odels was created for each outcome measure with use of the Bayesian Informa
tion Criterion.
Results Mean physical, mental, and overall QOL scores corresponding to the
time immediately before admission (baseline) were 45 (95% confidence interv
al [CI] 44-46), 47 (95% CI 46-48), and 70 (95% CI 68-72), respectively. By
1 year, mean physical, mental, and overall QOL scores were close to baselin
e (45 [95% CI 44-46], 48 [95% CI 47-49], and 73 [95% CI 71-74], respectivel
y). The predictors of physical, mental, and overall QOL were similar at 6 m
onths and 1 year. Important predictors of physical QOL were the correspondi
ng score at baseline, age, and previous bypass surgery (beta coefficients a
t 1 year: 5 [per 10-point difference in baseline score], -1 [per 10-year ag
e difference], 5.3; 95% Cls 4 to 5, -2 to -1, -9.2 to - 1.3, respectively).
Predictors of mental QOL were the corresponding score at baseline and depr
ession (beta coefficients at 1 year: 3 [per 10-point difference in baseline
score], - 3 [per 10-point difference in depression score]; 95% Cls 2 to 4,
-5 to -2, respectively). Predictors of overall QOL included the correspond
ing score at baseline and age (beta at 1 year: 2 [per 10-point score differ
ence], -3 [per 10-year age difference]; 95% Cis 1 to 3, -4 to - 1, respecti
vely). Depression was also a predictor of impaired physical and overall QOL
at 6 months (beta at 6 months: -1.6 [per 10-point score difference], -5.4
[per 10-point score difference]; 95% Cls -2.9 to -0.4, -7.7 to -3.2, respec
tively). No variables related to treatments received in-hospital were found
in the most clinically relevant models.
Conclusions These results suggest that age and psychosocial characteristics
at baseline are the most important predictors of QOL after AMI. Other clin
ical characteristics and treatments received in-hospital do not appear to s
trongly affect patients' long-term perceptions of QOL.