Skip to main content

Advertisement

Table 4 Univariate and multivariate logistic regression analyses of risk factors for early mortality in the study subjects†.

From: Preoperative ejection fraction as a predictor of survival after coronary artery bypass grafting: comparison with a matched general population

Risk factors OR early mortality Univariate analysis P value OR early mortality Multivariate analysis P value
EF 35% - 50% 2.623 (1.923 -3.578) < .0001 1.9 (1.335 -12.693) < .0001
EF < 35% 7.592 (5.143 -11.207) < .0001 4.206 (2.6 -6.805) < .0001
Age (y)* 1.08 (1.062 -1.098) < .0001 1.031 (1.004 -1.059) .026
Male sex 0.802 (0.601 -1.070) .071   
NYHA class 1.338 (1.122 -1.595) .001 1.168 (0.908 -1.503) .227
Angina class 1.002 (0.951 -1.055) .277   
Hypertension 0.929 (0.716 -1.207) .662   
COPD 1.966 (1.438 -2.687) < .0001 1.479 (0.943 -2.319) .089
Preoperative Hb level 0.692 (0.634 -0.755) < .0001 0.883 (0.779 -1.001 .051
Diabetes 1.524 (1.148 -2.023) .004 1.743 (1.195 -2.543) .004
Preoperative CrCl 0.965 (0.962 -0.968) < .0001 0.978 (0.967 -0.988) < .0001
PVD 1.633 (1.164 -2.290) .005 1.441 (991 -2.277) .118
Prior cardiac surgery 4.542 (3.304 -6.244) < .0001 3.064 (1.847 -5.083) < .0001
No. of anastomoses 0.903 (0.801-1.011) .077   
Off-pump 0.631 (0.366 -1.087) .097   
Use of IMA 0.231 (0.176 -0.303) < .001   
Cardioplegia 1.312 (0.995 -1.731) .054   
Emergency 6.550 (4.722 -9.087) < .0001 3.307 (1.597 -6.846) .001
Perioperative MI 5.938 (4.053 -8.699) < .0001   
Re-exploration 5.810 (4.261-7.922) < .0001   
IABP 13.974 (9.916 -19.691) < .0001   
  1. OR = Odds ratio, EF = ejection fraction, NYHA = New York Heart Association, COPD = chronic obstructive pulmonary disease, Hb = hemoglobin, CrCl = creatinine clearance, PVD = peripheral vascular disease, IMA = internal mammary artery, MI = myocardial infarction, IABP = intra-aortic balloon pump support.
  2. † Only preoperative variables which are significant in the univariate analysis were entered into the multivariate analysis
  3. *Entered as a continuous variable.