Skip to main content

Table 2 Prognostic test characteristics for in-hospital mortality

From: External validation of prognostic rules for early post-pulmonary embolism mortality: assessment of a claims-based and three clinical-based approaches

 

IMPACT

PESI

sPESI

Hestia

Low-Risk Mortality

1/230

2/309

0/250

0/211

n/N (%)

(0.4 %)

(0.6 %)

(0 %)

(0 %)

High-Risk Mortality

20/577

19/498

21/557

21/596

n/N (%)

(3.5 %)

(3.8 %)

(3.8 %)

(3.5 %)

Sensitivity

95.2 %

90.5 %

100 %

100 %

(95 % CI)

(74.1–99.8 %)

(68.2–98.3 %)

(80.8–100 %)

(80.8–100 %)

Specificity

29.1 %

39.1 %

31.8 %

26.8 %

(95 % CI)

(26.0–32.5 %)

(35.6–42.6 %)

(28.6–35.2 %)

(23.8–30.1 %)

PPV

3.5 %

3.8 %

3.8 %

3.5 %

(95 % CI)

(2.2–5.4 %)

(2.4–6.0 %)

(2.4–5.8 %)

(2.2–5.4 %)

NPV

99.6 %

99.4 %

100 %

100 %

(95 % CI)

(97.2–100 %)

(97.4–99.9 %)

(98.1–100 %)

(97.8–100 %)

C-statistic

0.766

0.792

0.762

0.857

(95 % CI)

(0.685–0.848)

(0.696–0.889)

(0.682–0.842)

(0.796–0.918)

  1. CI confidence interval, IMPACT In-hospital Mortality for Pulmonary embolism using Claims data, NPV negative predictive value, PESI Pulmonary Embolism Severity Index, PPV positive predictive value, sPESI simplified Pulmonary Embolism Severity Index