Skip to main content

Table 3 AUC and C-statistics for DRSF model with the most recent window

From: Dynamic prediction of hospital admission with medical claim data

Training /Testing Window Index Date

Evaluation window Index Date

# of subjects in evaluation

# of events in evaluation

Mean time to event

AUC at the 60th Day

AUC at the 180th Day

Harrell’s C statistics

# of covariates1

Feb 1,14

Aug 1,14

5175

263

175.17

0.67a (0.61,0.73)b

0.64 (0.60,0.67)

0.66 (0.63,0.70)

14

Mar 1,14

Sep 1,14

5143

250

175.14

0.67 (0.61,0.72)

0.67 (0.64,0.71)

0.68 (0.65,0.72)

17

Apr 1,14

Oct 1,14

5069

247

175.07

0.70 (0.65,0.75)

0.68 (0.65,0.71)

0.68 (0.65,0.71)

14

May 1,14

Nov 1,14

4988

230

175.31

0.65 (0.60,0.71)

0.66 (0.63,0.70)

0.69 (0.66,0.72)

20

Jun 1,14

Dec 1,14

4958

225

175.30

0.66 (0.61,0.71)

0.65 (0.62,0.70)

0.69 (0.65,0.72)

24

Jul 1,14

Jan 1,15

4261

201

175.20

0.74 (0.68,0.79)

0.69 (0.66,0.73)

0.71 (0.67,0.74)

16

Aug 1,14

Feb 1,15

4233

176

175.70

0.71 (0.65,0.77)

0.69 (0.65,0.72)

0.71 (0.67,0.74)

24

Sep 1,14

Mar 1,15

4222

172

175.85

0.61 (0.55,0.67)

0.65 (0.61,0.69)

0.68 (0.64,0.71)

14

  1. A larger AUC or C statistics represents a better model prediction performance
  2. 1Number of covariates selected by RSF model
  3. aScore obtained with the original dataset
  4. b95% confidence interval of the score obtained with 500 bootstrapped datasets