Table 3.
Original and imputed patient characteristics and associations with treatment delay in univariate analyses (n = 1017)
| Variable | Missing (%) | Original value | Imputed value | p-value* |
|---|---|---|---|---|
| Age (mean (SD) years) | 1 (0.1 %) | 62.5 (12.5) | 62.5 (12.5) | 0.16 |
| Female | 0 (0 %) | 271 (26.6 %) | 271 (26.6 %) | 0.63 |
| Diabetes Mellitus | 78 (7.7 %) | 124 (13.2 %) | 136 (13.4 %) | 0.91 |
| Hypertension | 84 (8.3 %) | 375 (40.2 %) | 406 (39.9 %) | 0.52 |
| Hyperlipidemiaa | 89 (8.8 %) | 303 (32.7 %) | 335 (32.9 %) | 0.35 |
| Obesity (BMI > 30 kg/m2) | 87 (8.6 %) | 80 (8.6 %) | 90 (8.8 %) | 0.54 |
| (Ex-) smoker | 88 (8.7 %) | 526 (56.6 %) | 563 (55.4 %) | 0.38 |
| Prior myocardial infarction | 102 (10.0 %) | 123 (13.4 %) | 144 (14.2 %) | 0.046 |
| Prior PCI | 102 (10.0 %) | 132 (14.4 %) | 158 (15.5 %) | 0.06 |
| Prior CABG | 103 (10.1 %) | 21 (2.3 %) | 30 (2.9 %) | 0.55 |
| Prior use of anticoagulants (≤7 days) | 86 (8.5 %) | 183 (19.7 %) | 205 (20.2 %) | 0.007 |
Significant results are highlighted in bold
SD standard deviation, BMI body mass index, kg/m2 kg per square meter, PCI percutaneous coronary intervention, CABG coronary artery bypass graft
*P-values are calculated using the Wald statistics, comparing a generalized linear model with and without the imputed variable, corrected for clustering of patients in PCI centers
aHyperlipidemia is defined as described in patient's history or statin use before admission