Factors related to survival analyzed by the Cox
proportional hazards model included age, dyspnea,
symptoms of chronic bronchitis, airway obstruction,
and Dlco as previously reported correlates of survival.
11, 17 However, the correlation of FEV1 with the
survival rate was weak in the present study. Degree
of airway obstruction was not included among the
entry criteria, although this factor has been used to
screen participants in previous studies such as the
Intermittent Positive Pressure Breathing trial.11
Therefore, the present study might have enrolled
more patients with COPD with mild airway obstruction
than other studies. This, in turn, might cause
underestimation or overestimation of the effect of
FEV1 on the survival rate. Alternatively, the strong
correlation between dyspnea and survival may have
weakened the degree of correlation between airway
obstruction and survival.