A scoping methodological review of simulation studies comparing statistical and machine learning approaches to risk prediction for time-to-event data

Table 3 The number of repetitions, number of data-generating mechanisms and factors varied in each article

	Repetitions	Factors varied in the data-generating mechanisms
	Repetitions	Number of DGMs	Sample size	Failure time distribution	Number of covariates	Covariate relationships	Covariate effects	Censoring
Geng et al. (2014) [26]	100	20	✓	✓	✓	✓	✓	✓
Golmakani et al. (2020) [31]	1	6	✓		✓	✓	✓
Gong et al. (2018)* [27]	500	57	✓		✓	✓	✓	✓
Hu and Steingrimsson (2018) [28]	1000	4	✓			✓	✓
Katzman et al. (2018) [29]	1	2					✓
Lowsky et al. (2012)** [25]	20	12	✓				✓
Omurlu et al. (2009) [24]	1000	4	✓
Steingrimsson and Morrison (2020) [32]	1000	16	✓	✓	✓		✓	✓
Wang and Li (2019) [30]	***	24		✓	✓
Xiang et al. (2000) [23]	50	9	✓		✓	✓	✓	✓

*Gong et al. (2018) [27] also included three data-generating mechanisms where data was based on clinical data
**All simulated datasets in Lowsky et al. (2012) [25] were based on a real kidney transplant dataset
***Numbers of repetitions were unclear in Wang and Li (2019) [30]

ISSN: 2397-7523