Background: Lung disease accounts for most of the mortality in patient
s with cystic fibrosis (CF). Lung transplantation is an option for pat
ients severely impaired, being recommended when life expectancy is est
imated to be <2 years. Our objectives were to evaluate in our patient
population the validity of currently accepted criteria for low life ex
pectancy and to identify other potentially useful criteria. Methods: D
ata were retrieved from CF patients follow ed up at our center who rea
ched and kept an FEV1 <30% predicted. A life table was created and str
atified according to characteristics believed to be of importance. In
addition, the rate of decline in percent predicted FEV1 was analyzed.
These characteristics were evaluated as predictors of risk of death. R
esults: The median survival was 3.9 years (95% confidence interval, 2.
88 to 4.12 years), with no significant differences according to gender
, nutritional status, presence of diabetes, or decade in which the pat
ient was cared for. Only by age was there a significant difference in
the median survival (p<0.05). By proportional hazards regression, only
the rate of decline in percent predicted FEV1 was a significant predi
ctor of the risk of death, with a borderline effect from younger age (
p=0.06). Conclusion: In our patient population, a cutoff value of FEV1
of <30% predicted is not a reliable predictor of high risk of death w
ithin 2 years. The yearly rate of decline of percent predicted FEV1 is
a better parameter to identify those patients at high risk for death.