Table 3.

Previous research regarding COVID-19 prediction models

Class	Studies	Prediction type	Outcome variable	Data type	Sample size	Easy-to-measure input features
	Our model	Prognosis	Mortality	Nationwide	149,471	Yes
1	Zoabi et al. [25]	Diagnosis	RT-PCR	Nationwide	99,232	Yes
	Yanamala et al. [27]	Diagnosis	RT-PCR	Local	3883	No
	Gozes et al. [28]	Diagnosis	RT-PCR	Local	157	No
	Song et al. [29]	Diagnosis	RT-PCR	Local	275	No
	Feng et al. [32]	Diagnosis	RT-PCR	Local	164	No
	Jin et al. [30]	Diagnosis	RT-PCR	Local	11,356	No
	Punn et al. [31]	Diagnosis	RT-PCR	Local	1214	No
	Menni et al. [26]	Diagnosis	RT-PCR	Nationwide	2,618,862	Yes
2	Cifuentes et al. [33]	Prognosis	Mortality	Nationwide	1,033,218	Yes
	Cho et al. [35]	Prognosis	Mortality	Nationwide	7590	No
	Ikemura et al. [36]	Prognosis	Mortality	Local	4313	No
	Her et al. [34]	Prognosis	Mortality	Nationwide	5628	No
3	Subudhi et al. [38]	Prognosis	Complication or mortality	Local	10,826	No
	Shamout et al. [37]	Prognosis	Complication or mortality	Local	3661	No
	Marcos et al. [39]	Prognosis	Complication or mortality	Local	1270	No
	Kim et al. [40]	Prognosis	Complication or mortality	Nationwide	4787	Yes
	Su et al. [41]	Prognosis	Complication or mortality	Local	14,418	No
4	Rinderknecht et al. [42]	Prognosis	Complication	Nationwide	15,753	Yes
	Wang et al. [43]	Prognosis	Complication	Local	3008	No

RT-PCR reverse transcription polymerase chain reaction