Table 4

Referral to Revascularizationlegend

VariableUnivariate AnalysisMultivariate Analysis
Chi-Squarep ValueOR (95% CI)Chi-Squarep ValueOR (95% CI)
Extent of CAD385< 0.0012.8 (2.5–3.1)267< 0.0012.5 (2.3–2.8)
SRS(1-point increase)207< 0.0011.13 (1.11–1.15)57< 0.0011.07 (1.06–1.09)
SSS(5-point decrease)186< 0.0011.5 (1.4–1.6)
Chest pain class114< 0.0011.5 (1.4–1.7)62< 0.0011.4 (1.3–1.6)
Age (10-year increase)54< 0.0011.4 (1.3–1.5)
Symptom status49< 0.0012.3 (1.8–2.9)
Hyperlipidemia14< 0.0011.4 (1.2–1.7)
Cardiac enlargement90.0031.4 (1.1–2.6)
Gender (male)70.0101.3 (1.1–1.6)
Pharmacologic stress70.0100.8 (0.6–0.9)60.0140.7 (0.6–0.9)
Hypertension60.021.2 (1.0–1.5)
Smoking1NS1.1 (0.9–1.2)
Diabetes< 1NS0.9 (0.8–1.1)
  • Variables without numbers in the multivariate analysis did not meet the entry criteria into the final model;

  • a summed reversibility score (SRS) of 0 represents a nonischemic image;

  • a summed stress score (SSS) of 56 represents a normal image. Final model: chi-square = 652, p < 0.001; C-index = 0.82. Bootstrapping analysis the extent of CAD came into every model, SRS and chest pain class in 98% of the models and pharmacologic stress in 64%. In the initial multivariate stepwise analysis, symptom status just met the entry criteria into the model (chi-square = 4, p = 0.038). In the bootstrapping analysis, it came into only 13% of the models. This variable was therefore eliminated from the final multivariate model.

  • legend CAD = coronary artery disease; other abbreviations as in Table 3.