Table 3.

Summary of the data preprocessing step.

Study, year	Outcomes	Prediction time horizon	Class imbalance ratio (%)	Data imbalance handling methods	Feature selection methods	Number of features
Inselman et al [27], 2022	Asthma exacerbation	180 d	22.60	Unknown	Unknown	21
Hurst et al [25], 2022	Asthma exacerbation	30, 90, and 180 d	37	Unknown	Unknown	41
Hogan et al [28], 2022	Asthma readmission	180 d	5.70	Unknown	Backward stepwise variable selection	21
Zein et al [29], 2021	Asthma exacerbation	28 d	Nonsevere=32.80 Severe=2.90	Unknown	Unknown	82
Sills et al [30], 2021	Asthma-related hospitalization	Admission after A&Ea department visit	22.50	Oversampling	Automated feature selection	13
Hozawa et al [31], 2021	Asthma exacerbation	365 d	13.70	Unknown	Unknown	25
Lisspers et al [32], 2021	Asthma exacerbation	15 d	0.04	Undersampling and weighting method	Correlation and LGBMb model	>500
Ananth et al [23], 2021	Asthma exacerbation	365 d	50	Unknown	Unknown	17
Tong et al [33], 2021	Asthma-related hospitalization or A&E department visit	365 d	1.66	WEKAc	Automated feature selection	234
Mehrish et al [24], 2021	Asthma prevalence, asthma-related hospitalization, or hospital readmission	Unknown	Unknown	Unknown	Unknown	7
Xiang et al [4], 2020	Asthma exacerbation	365 d	7.20	SMOTEd	Automated feature selection	Unknown
Cobian et al [34], 2020	Asthma exacerbation	90 d	Unknown	Unknown	Unknown	>25
Luo et al [35], 2020	Asthma-related hospitalization	365 d	3.59	Unknown	Automated feature selection	235
Roe et al [22], 2020	Asthma-related mortality	Unknown	49	Unknown	Unknown	42
Luo et al [26], 2020	Asthma-related hospitalization	365 d	2.30	Unknown	Automated feature selection	337
Wu et al [21], 2018	Asthma relapse	Unknown	32.89	Random undersampling	Unknown	60
Patel et al [11], 2018	Asthma-related hospitalization	Admission after EDe visit	17	Unknown	Unknown	100

^aA&E: accident and emergency.

^bLGBM: light gradient boosting method.

^cWEKA: Waikato Environment for Knowledge Analysis.

^dSMOTE: synthetic minority oversampling technique.

^eED: emergency department.