Objective: The hypothesis that traditionally defined preoperative risk
factors predict prolonged mechanical ventilation after coronary arter
y bypass graft surgery (CABG) was tested in our cohort. The predictive
power of these factors was quantified, and specific patient subsets d
estined for prolonged mechanical ventilation after CABG surgery were d
efined. Design: Five hundred thirteen consecutive patients undergoing
CABG were prospectively evaluated. Preoperative pulmonary evaluation i
ncluded clinical historic data, standard spirometry, and arterial bloo
d gas. Preoperative cardiac parameters included clinical parameters an
d left ventricular function assessment. Nonthoracic organ (renal, endo
crine, pancreas, liver) function was assessed. Setting: University-bas
ed, tertiary referral center. Interventions: None (observational only)
. Outcomes measured: Duration of mechanical ventilation, duration of s
urgical ICU stay, and mortality. Results: Multivariate regression anal
yses revealed that for the patient undergoing routine elective surgery
and the patient undergoing urgent surgery, prolonged mechanical venti
lation and death were rare events (8.3% and 2.0%, respectively). The c
ombination of reduced left ventricular ejection fraction and the prese
nce of selected preexisting comorbid conditions (clinical congestive h
eart failure, angina, current smoking, diabetes) served as modest risk
factors for prolonged mechanical ventilation; their absence strongly
predicted an uncomplicated postoperative respiratory course. No pulmon
ary diagnosis, mechanical lung function, or blood gas parameter substa
ntially contributed to predicting adverse outcome. Classification and
regression tree subgroup analysis refined specific factors important i
n specific subgroups. Conclusion: With the exception of left ventricul
ar ejection fraction, no preoperative factors emerge as good predictor
s across all subgroups. This series suggests that pulmonary diagnosis,
lung mechanics, and blood gas parameters do not offer the clinician g
lobal rules in predicting postoperative respiratory outcome, nor shoul
d they be used as exclusion criteria for CABG surgery.