Table 2.
Multivariate Cox proportional hazard models for the association between concomitant PTA at the time of PCI and clinical outcomes.
| Clinical outcomes | Before PSM | After PSM | ||||||
|---|---|---|---|---|---|---|---|---|
| Events | Models* | Events | Models* | |||||
| N = 279 | N = 6284 | HR (95% CI) | P value | N = 279 | N = 1385 | HR (95% CI) | P value | |
| Death | 73 (26.2) | 919 (14.6) | 1.69 (1.32–2.15) | < 0.001 | 73 (26.2) | 273 (19.7) | 1.52 (1.17–1.98) | 0.002 |
| CV death | 21 (7.5) | 283 (4.5) | 1.54 (0.99–2.42) | 0.058 | 21 (7.5) | 69 (5.0) | 1.59 (0.97–2.59) | 0.065 |
| MACEs | 79 (28.3) | 1611 (25.6) | 1.11 (0.89–1.40) | 0.358 | 79 (28.3) | 393 (28.4) | 1.00 (0.78–1.27) | 0.969 |
| Myocardial infarction | 13 (4.7) | 414 (6.6) | 0.87 (0.50–1.52) | 0.625 | 13 (4.7) | 80 (5.8) | 0.85 (0.47–1.53) | 0.591 |
| Revascularization | 49 (17.6) | 962 (15.3) | 1.25 (0.93–1.66) | 0.136 | 49 (17.6) | 237 (17.1) | 1.09 (0.80–1.48) | 0.594 |
| Stroke | 16 (5.7) | 344 (5.5) | 1.09 (0.66–1.81) | 0.727 | 16 (5.7) | 101 (7.3) | 0.88 (0.52–1.50) | 0.639 |
| Any bleeding | 9 (3.2) | 209 (3.3) | 0.97 (0.50–1.89) | 0.920 | 9 (3.2) | 49 (3.5) | 0.91 (0.44–1.85) | 0.787 |
| ESRD | 28 (10.0) | 206 (3.3) | 2.05 (1.37–3.06) | < 0.001 | 28 (10.0) | 93 (6.7) | 1.60 (1.05–2.45) | 0.029 |
| Amputation | 20 (7.2) | 108 (1.7) | 3.81 (2.36–6.16) | < 0.001 | 20 (7.2) | 34 (2.5) | 3.41(1.96–5.92) | < 0.001 |
| PTA after discharge | 68 (24.4) | 168 (2.7) | 9.27 (6.93–12.4) | < 0.001 | 68 (24.4) | 70 (5.1) | 5.94 (4.25–8.30) | < 0.001 |
PSM, propensity score matching; PCI, percutaneous coronary intervention; PTA, Percutaneous transluminal angioplasty; CV cardiovascular; MACE, major adverse cardiovascular events; ESRD, endstage renal disease; HR, hazard ratio; CI, confidence interval.
*The multivariate model includes age, sex, BMI, waist circumference, MI at index admission, number of coronary stents used, income ≥ median, prior bleeding events, MI, PAD, stroke, diabetes, hypertension, peptic ulcer, DAPT duration and statin use as covariates. The model was reduced using a backward selection method (cutoff criterion, p > 0.1).