Table 4.
Multivariate logistic regression for ICD prediction.
| Variables | Odds ratio | 95% CI | p-value | |
| Lower limit | Upper limit | |||
| Congestive heart failure | 3.99 | 1.46 | 10.87 | 0.01 |
| Ventricular arrhythmias | 33.94 | 9.20 | 125.18 | 0.001 |
| Pulmonary circulatory disorders | 4.40 | 1.34 | 14.44 | 0.02 |
| Diabetes mellitus | 0.29 | 0.08 | 1.02 | 0.05 |
| Other neurological disorders | 0.65 | 0.22 | 1.94 | 0.44 |
| Liver disease | 2.39 | 0.78 | 7.39 | 0.13 |
| Weight loss | 3.13 | 0.87 | 11.33 | 0.08 |
| Fluid and electrolyte disorders | 1.21 | 0.49 | 2.97 | 0.68 |
| Alcohol abuse | 4.41 | 1.22 | 15.94 | 0.02 |
CI, confidence interval; ICD, implantable cardioverter defibrillator.
The model was constructed based on univariate regression, with a threshold of 0.2.
The bolded p-value in Table 4 indicates that Ventricular arrhythmias like ventricular tachycardia or ventricular fibrillation, pulmonary circulation disorders like pulmonary embolism, and abusing alcohol (more than 3 drinks a day for men and 2 drinks per women) resulted in a greater likelihood of having an ICD placed.