TABLE 4

Univariable and multivariable logistic regression of risk factors with optimal cut-points for in-hospital adverse outcomes in normotensive acute pulmonary embolism (PE) patients who are high-risk simplified Pulmonary Embolism Severity Index (sPESI)

Univariable models OR (95% CI)p-valueMultivariable models OR (95% CI)
1) hs-TnT ≥50 pg·mL−1, CT RV/LV ≥1.5, heart rate ≥100 beats·min−1, central PA embolism, SBP 90–100 mmHg2) CT RV/LV ≥1.5, heart rate ≥100 beats·min−1, central PA embolism, SBP 90–100 mmHg3) CT RV/LV ≥1.5, heart rate ≥100 beats·min−1, central PA embolism4) CT RV/LV ≥1.5, heart rate ≥100 beats·min−1
Patients n1179149814981498
Age per year increase0.98 (0.96–0.99)0.049
Lower extremity DVT present#9.61 (2.24–41.196)0.002
Elevated lactate >2.2 mmol·L−15.06 (2.17–11.81)<0.001
Oxygen saturation <90%2.86 (1.09–7.46)0.032
Syncope1.30 (0.39–4.32)0.673
hs-TnT ≥50 pg·mL−18.37 (3.58–19.57)<0.0011.90 (0.67–5.40)
p=0.223
CT RV/LV ratio ≥1.5¶,+22.92 (8.68–60.52)<0.0015.55 (1.77–17.04)
p=0.003
9.02 (3.06–26.58)
p<0.001
9.11 (3.09–26.8)
p<0.001
15.35 (5.76–40.88)
p<0.001
Central PA embolism§9.85 (4.32–22.46)<0.0012.93 (1.10–7.80)
p=0.031
2.86 (1.13–7.23)
p=0.027
2.91 (1.15–7.36)
p=0.24
Heart rate ≥100 beats·min−14.90 (2.19–10.96)<0.0012.61 (1.01–6.72)
p=0.047
3.02 (1.18–7.70)
p=0.021
2.96 (1.17–7.51)
p=0.022
3.36 (1.33–8.43)
p=0.010
SBP 90–100 mmHg3.26 (1.11–9.56)0.0313.29 (0.99–10.88)
p=0.051
3.51 (1.07–11.50)
p=0.038
Model performance measures
 Akaike information criterion217.0216.6/228.9ƒ230.4234
 C-statistic0.880.88/0.89ƒ0.890.87

Hs-TnT: high-sensitivity troponin; CT RV/LV: computed tomography right/left ventricle ratio; PA: pulmonary artery; SBP: systolic blood pressure; DVT: deep vein thrombosis. #: documented positive if reported on duplex ultrasound of the lower extremities; : cut-points determined by Youden's index; +: measured by dividing the right and left ventricle diameter at the valvular level of the CT angiogram axial cuts; §: defined as thrombus present within the central pulmonary arteries proximal to a lobar artery; ƒ: the first value is calculated using a model limited to the 1179 patients in model 1; the second value is calculated using the 1498 patients in models 2–4. There were 29 adverse in-hospital outcomes in models 2–4.