Table 10

Performance of various values of PetCO2 and duration of CPR for prediction of survival in patients with nonshockable initial rhythm in cardiac arrest

PetCO2

Cut-off (kPa [mmHg])

n

Min-max (kPa [mmHg])

Mean ± SD (kPa [mmHg])

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

AUROC (95% CI)


initial

≤1.3 (10)

97

0.0–1.3 (0.0–10)

0.66 ± 0.32 (5.1 ± 1.9)

100

27

23

100

0.58 (0.52–0.63)

>1.3 (10)

336

1.4–8.4 (10.1–63.2)

4.17 ± 2.12 (30.9 ± 15.8)

0–10 min

≤1.7 (12.8)

229

0.4–1.7 (3.5–12.8)

0.99 ± 0.36 (7.1 ± 3.3)

100

64

37

100

0.88 (0.84–0.91)

>1.7 (12.8)

204

1.8–5.6 (12.9–42.1)

2.79 ± 0.86 (21.5 ± 6.2)

10 min

≤1.6 (12.1)

199

0.3–1.6 (2.2–12.1)

1.03 ± 0.38 (7.6 ± 3.1)

100

56

32

100

0.87 (0.83–0.91)

>1.6 (12.1)

234

1.7–7.2 (12.2–54.2)

3.02 ± 1.08 (22.6 ± 6.5)

11–15 min

≤1.7 (12.8)

239

0.4–1.7 (3.4–12.8)

0.99 ± 0.31 (7.3 ± 2.8)

100

67

39

100

0.86 (0.83–0.90)

>1.7 (12.8)

194

1.4–5.1 (12.9–38.4)

3.14 ± 0.81 (23.3 ± 6.7)

15 min

≤1.9 (14.3)

241

0.3–1.9 (2.4–14.3)

1.12 ± 0.41 (7.9 ± 3.8)

100

67

39

100

0.87 (0.84–0.91)

>1.9 (14.3)

192

2.1–6.1 (14.4–45.9)

3.62 ± 0.94 (28.6 ± 7.5)

20 min

≤2.1 (15.8)

242

0.3–2.1 (2.3–15.8)

0.91 ± 0.31 (7.2 ± 2.2)

100

68

40

100

0.87 (0.84.0.91)

>2.1 (15.8)

190

2.3–7.8 (15.9–58.7)

4.39–1.11 (33.1 ± 7.7)

max

≤2.8 (21.1)

231

0.8–2.8 (6.1–21.1)

1.68 ± 0.44 (13.4 ± 3.5)

100

65

38

100

0.89 (0.86–0.92)

>2.8 (21.1)

202

2.9–10.7 (21–80.5)

5.26 ± 1.48 (39.1 ± 12.2)

Final

≤1.6 (12.1)

240

0.2–1.6 (1.9–12.1)

0.97 ± 0.33 (6.9 ± 2.7)

100

67

39

100

0.87 (0.84–0.91)

>1.6 (12.1)

193

1.9–6.6 (12.2–49.6)

3.06 ± 0.97 (27.1 ± 7.2)


'Nonshockable' was defined as asystole or pulseless electrical activity. AUROC, area under the receiver operating characteristic curve; CI, confidence interval; CPR, cardiopulmonary resuscitation; NPV, negative predictive value; PetCO2, partial pressure of end-tidal carbon dioxide; PPV, positive predictive value; ROSC, restoration of spontaneous circulation; SD, standard deviation.

Kolar et al. Critical Care 2008 12:R115   doi:10.1186/cc7009

Open Data