A practical solution to estimate the sample size required for clinical prediction models generated from observational research on data

European Radiology Experimental

Table 4 Variability of sample size with R²_Nagelkerke

Model	R²_Nagelkerke = 0.15		R²_Nagelkerke = 0.5		R²_Nagelkerke = 0.8
Model	Sample size	EPP	Sample size	EPP	Sample size	EPP
NB 5-year mortality risk	2383	24.39	666	6.82	550	5.63
NB relapse/progression risk	2495	21.42	704	6.04	590	5.06
NB time to death	2951	14.79	858	4.30	749	3.75
HR-NB 5-year mortality risk	2247	37.45	620	10.33	501	8.35
HR-NB 5-year relapse/progression risk	2280	31.01	631	8.58	513	6.98
HR-NB time to relapse/progression	2280	24.15	631	6.68	513	5.43
DIPG 1-year mortality risk	2247	33.70	620	9.30	501	7.52
DIPG 2-year mortality risk	2848	16.04	824	4.64	713	4.02
DIPG 1-year progression risk	2575	20.17	731	5.73	618	4.84
DIPG time to death	2705	16.72	775	4.79	663	4.10
DIPG time to progression	2419	20.69	679	5.81	563	4.82

Sample size calculations with different values of R²_Nagelkerke (0.15, 0.5, and 0.8). The number of events per predictor variable was obtained with the pmsampsize package