Predictor characteristics necessary for building a clinically useful risk prediction model: a simulation study

Table 3 Model performance measures according to odds ratio, number, and prevalence of simulated predictors

Simulated predictor characteristics			Model performance measures
OR of simulated predictors	Number of simulated predictors added to original model	Prevalence of simulated predictors	Proportion of population (%) with informative likelihood ratio ^a	Proportion of population (%) assigned to clinically distinct risk group ^b	AUC	Nagelkerke’s r² (%)
2	3	10 %	0.0	27.2	0.71	10.0
2	3	20 %	0.0	29.9	0.73	11.7
2	3	40 %	0.0	34.8	0.74	12.9
2	5	10 %	0.0	28.9	0.73	11.8
2	5	20 %	0.0	33.3	0.75	14.6
2	5	40 %	0.0	39.2	0.77	16.6
6	3	10 %	0.0	54.4	0.83	29.4
6	3	20 %	63.6	63.6	0.87	37.1
6	3	40 %	70.2	70.0	0.88	37.1
6	5	10 %	66.9	66.9	0.88	40.6
6	5	20 %	72.0	72.0	0.92	50.7
6	5	40 %	73.8	73.8	0.93	51.2

^aDefined as the proportion of the population classified into a stratum with a likelihood ratio <0.10 or >10.0
^bDefined as the proportion of the population classified into a stratum with predicted risk meaningfully different than the baseline rate of pre-eclampsia in the population (<0.03 or >0.15)

ISSN: 1471-2288