Table 2.
Cardiac testing, radiation exposure, disposition, length of hospital stay, time to diagnosis, and healthcare costs as predicted using decision rules based on hs-cTnI and cardiovascular risk factors vs as observed in the ROMICAT II trial.
ROMICAT II as observed (n = 255) | ROMICAT II as predicted (n = 255) | P | |
---|---|---|---|
Cardiac testing (%)a | <0.001 | ||
No testing | 0 (0.0) | 86 (33.7) | |
1 test | 203 (79.6) | 127 (49.8) | |
≥2 tests | 52 (20.4) | 42 (16.5) | |
Invasive coronary angiography (%) | 23 (9.0) | 22 (8.6) | 1.000 |
Intervention (%) | |||
PCIb | 14 (5.5) | 14 (5.5) | 1.000 |
CABG | 1 (0.4) | 1 (0.4) | 1.000 |
Cumulative radiation exposure, mSv/patientc | 10.2 ± 11.9 | 7.7 ± 12.4 | <0.001 |
Disposition | <0.001 | ||
ED discharge | 62 (24.3%, 95% CI, 19.1–30.1%) | 128 (50.2%, 95% CI, 43.9–56.5%) | |
Observational unit admission | 150 (58.8%, 95% CI, 52.5–64.9%) | 92 (36.1, 95% CI, 30.2–42.3%) | |
Hospital admission | 41 (16.1%, 95% CI, 11.8–21.2%) | 34 (13.3%, 95% CI, 9.4–18.1%) | |
Left against medical advice | 2 (0.8%, 95% CI, 0.1–2.8%) | 1 (0.4%, 95% CI, 0.0–2.1%) | |
Length of hospital stay, hours | <0.001 | ||
Median (25th-75th percentile) | 21.4 (8.0–28.6) | 8.2 (2.0–27.1) | |
Time to diagnosis, hours | <0.001 | ||
Median (25th-75th percentile) | 8.2 (5.8–22.7) | 5.7 (1.0–20.4) | |
Healthcare cost per patient in US$ | |||
Median (25th-75th percentile) | 2698 (2698–2698)d | 2698 (655–2698)d | <0.001 |
Mean ± SD | 4066 ± 4858 | 3342 ± 5089 | |
Non-ACS patients only | |||
Median (25th-75th percentile) | 2698 (1836–2698) | 1837 (655–2698) | <0.001 |
Mean ± SD | 2922 ±1425 | 2136 ±1634 |
Cardiac testing included coronary computed tomography angiography, exercise treadmill test, stress echocardiography, nuclear myocardial perfusion imaging, and invasive coronary angiography.
CABG, coronary artery bypass grafting; PCI, percutaneous coronary intervention.
Radiation exposure included exposure from coronary computed tomography, nuclear myocardial perfusion imaging, and invasive coronary angiography.
Because all changes in costs occurred in patients below the median of costs, observed and predicted median costs values are identical.