Table 3.
12 h Cut-Off Model: Cox model with hours in perpetuity as a time-varying covariate.
| Estimated Coefficient (95% CI) | Relative Risk (95% CI) | χ2 | df | p Value | |
|---|---|---|---|---|---|
| Hours in Perpetuity (>8 mL/kg) | 0.1042 (0.0254, 0.183) | 1.1099 (1.0257, 1.2009) | 6.7183 | 1 | 0.0095 |
| Age | 0.0279 (0.0217, 0.0341) | 1.0283 (1.022, 1.0347) | 78.1274 | 1 | 0.0000 |
| O2 | −0.014 (−0.0205, −0.0076) | 0.9861 (0.9797, 0.9924) | 18.2653 | 1 | 0.0000 |
| Hours in Acuity (Linear) | 0.0145 (−0.0423, 0.0712) | 1.0146 (0.9586, 1.0738) | 0.2491 | 1 | 0.6177 |
| Hours in Acuity (Nonlinear) | 9.4794 | 3 | 0.0236 | ||
| Arrival by EMS Ground | 0.0743 (−0.3989, 0.5475) | 1.0771 (0.671, 1.729) | 0.0947 | 1 | 0.7583 |
| Other Mode of Arrival | 0.0663 (−0.6111, 0.7437) | 1.0685 (0.5427, 2.1037) | 0.0368 | 1 | 0.8479 |
| POV Mode of Arrival | −0.2338 (−0.8186, 0.351) | 0.7915 (0.4411, 1.4205) | 0.6140 | 1 | 0.4333 |
| BUMCS | 0.4415 (0.242, 0.6409) | 1.555 (1.2738, 1.8983) | 18.8148 | 1 | 0.0000 |
| NIPPV prior to intubation | −0.94 (−1.2658, −0.6142) | 0.3906 (0.282, 0.5411) | 31.9737 | 1 | 0.0000 |
| Ideal Body Weight | −6 × 10−4 (−0.01, 0.0089) | 0.9994 (0.99, 1.0089) | 0.0141 | 1 | 0.9054 |
| BMI | 0.0011 (−0.009, 0.0112) | 1.0011 (0.991, 1.0113) | 0.0442 | 1 | 0.8334 |
| Heart Rate (Linear) | 7 × 10−4 (−0.0021, 0.0036) | 1.0007 (0.9979, 1.0036) | 0.2398 | 1 | 0.6244 |
| Heart Rate (Nonlinear) | 11.4993 | 2 | 0.0035 | ||
| Systolic Blood Pressure | −0.0025 (−0.0091, 0.0041) | 0.9975 (0.991, 1.0041) | 0.5474 | 1 | 0.4594 |
| Mean Arterial Pressure | −0.0049 (−0.0137, 0.004) | 0.9951 (0.9863, 1.004) | 1.1622 | 1 | 0.2810 |
| Acuity Score | −0.4323 (−0.6313, −0.2332) | 0.649 (0.5319, 0.792) | 18.1107 | 1 | 0.0000 |
| Hours in Perpetuity × Age | −4 × 10−4 (−7 × 10−4, 0) | 0.9996 (0.9993, 1) | 4.4328 | 1 | 0.0353 |
| Hours in Perpetuity × O2 | −9 × 10−4 (−0.0017, −1 × 10−4) | 0.9991 (0.9983, 0.9999) | 5.1694 | 1 | 0.0230 |
The estimated coefficient is on the log(hazard rate) scale. A relative risk over one indicates that having the associated covariate (or each unit of the associated covariate) increases the hazard of dying in the hospital. The nonlinear portion of the spline terms for hours in acuity and heart rate do not have single estimated coefficients. Thus, only χ2 tests and the associated p values are given for these rows. 2025 patients are included in this model.