Skip to main content
. 2023 Apr 24;12(7):953–962. doi: 10.1002/psp4.12965

TABLE 3.

RMSE and bias for AUC predictions from models for given number of concentration measurements over a 48‐h period on 1000 subsampled patients.

A priori One Two Three
Bias Adane et al.
53.4±0.5
29.9±0.4
23.8±0.3
21.1±0.2
Mangin et al.
63.4±0.5
23.2±0.3
20.7±0.3
9.7±0.2
Medellin‐G et al.
55.4±0.5
22.4±0.3
18.3±0.3
12.6±0.2
Revilla et al.
3.7±0.3
4.0±0.2
3.5±0.2
4.4±0.2
Roberts et al.
7.9±0.3
4.5±0.2
5.4±0.2
4.2±0.1
Thomson et al.
22.6±0.4
14.2±0.3
11.3±0.3
10.7±0.2
Naïve Ensemble
34.4±0.4
16.4±0.2
13.8±0.2
10.4±0.2
PBMA
34.4±0.4
4.7±0.2
4.8±0.2
3.4±0.2
SMC
18.4±0.5
8.2±0.3
7.1±0.2
5.2±0.2
SMC + PBMA
26.4±0.1
6.5±0.2
5.9±0.2
4.3±0.2
RMSE Adane et al.
80.0±0.7
47.3±0.4
40.9±0.6
32.5±0.3
Mangin et al.
83.3±0.6
36.2±0.3
33.0±0.3
21.4±0.2
Medellin‐G et al.
76.5±0.6
33.5±0.2
28.5±0.2
20.9±0.2
Revilla et al.
32.7±0.4
21.6±0.2
20.0±0.2
16.7±0.2
Roberts et al.
35.8±0.3
20.6±0.1
20.0±0.2
16.1±0.1
Thomson et al.
48.2±0.4
30.2±0.2
25.8±0.2
20.9±0.1
Naïve ensemble
55.7±0.4
28.5±0.2
24.9±0.2
19.3±0.2
PBMA
55.7±0.4
20.8±0.2
19.3±0.2
15.2±0.2
SMC
41.8±0.5
22.1±0.2
19.6±0.2
15.2±0.2
SMC + PBMA
47.5±0.5
20.5±0.1
18.7±0.2
14.6±0.2

Abbreviations: AUC, area under the curve; PBMA, Performance‐based Model Averaging; RMSE, root‐mean‐square error; SMC, Synthetic Model Combination.