Table 1.

Reviewed studied of preoperative patient selection and planning in hip and knee arthroplasty.

Author, year	Pathology/Surgery	ML algorithms	Prediction outputs	Patients in testing set (n)	Avg. age	%Female	Data source
Alam et al., 2019 [14]	THA	ANN, regression	Costs	10,000	—	—	Multicenter
Aram et al., 2018 [15]	TKA	ANN, decision tree	Readmissions/reoperation	6137	70.2	57.1	National Joint Registry (UK)
Bonakdari et al., 2020 [16]	TKA, THA	ANN	Readmissions/reoperation	5251	60.3	66.3	UK National Institute for Health and Care Excellence (NICE)
Borjali et al., 2020 [30]	THA	ANN	Preop patient selection/planning	25	61.3	47	Single institution
Cafri et al., 2019 [31]	TKA	Decision tree, regression	Preop patient selection/Planning	74,520	65.5	57.4	Kaiser Permanente Total Joint Replacement Registry (KPTJRR)
Fontana et al., 2019 [32]	TJA	Regression, SVM, decision tree	Preop patient selection/planning	3430	63	—	Patient database
Gabriel et al., 2019 [17]	THA	Regression, decision tree	Discharge/LOS	240	—	50.5	Single institution
Hirvasniemi et al., 2019 [33]	THA	Regression	Preop patient selection/planning	197	55.7	83.8	CHECK cohort
Hyer et al., 2019 [18]	TJA	Regression, decision tree	Preop patient selection/planning	262,290	73	55.8	Medicare
Hyer et al., 2020 [19]	All	Decision tree	Costs, readmissions/reoperation	262,290	73	55.8	Medicare
Hyer et al., 2020 [20]	THA, TKA	Cluster analysis	Costs	19,522	—	—	Medicare
Jafarzadeh et al., 2020 [44]	TKA	ANN, regression	Preop patient selection/planning	2357	61.6	62	Multicenter Osteoarthritis (MOST) Study
Jodeiri et al., 2020 [34]	THA	ANN	Preop patient selection/planning	95	—	—	Single institution
Jones et al., 2019 [21]	THA, TKA	Regression, boosting	Readmissions/reoperation	—	—	—	Medicare
Kang et al., 2020 [35]	THA	ANN	Preop patient selection/planning	1202	—	—	Multicenter
Karnuta et al., 2019 [22]	Hip fracture	Bayesian	Costs, discharge/LOS	98,562	—	73.5	New York Statewide Planning and Research Cooperative System database
Karnuta et al., 2019 [46]	TJA	ANN	Costs	73,901	—	—	New York State inpatient administrative database
Lee et al., 2017 [26]	TJA	Regression	Readmissions/reoperation	26	—	—	Analysis of patient records to provide risk prediction for readmissions.
Lee et al., 2019 [27]	TJA	Boosting, regression	Costs	131	—	—	Single institution
Navarro et al., 2018 [28]	TKA	Bayesian	Costs, discharge/LOS	35,362	—	—	Administrative database
Pareek et al., 2019 [37]	Knee fracture	ANN, decision tree, regression, boosting, Bayesian	Preop patient selection/planning	62	64.6	68	Single institution
Ramkumar et al., 2019 [23]	TKA	ANN	Costs, discharge/LOS	175,042	73.5	64	National inpatient sample
Ramkumar et al., 2019 [24]	THA	Bayesian	Costs, discharge/LOS	30,584	—	—	Patient database
Ramkumar et al., 2019 [25]	THA	ANN	Costs, discharge/LOS	78,335	75.3	63.6	National inpatient sample
Sherafati et al., 2020 [38]	THA	ANN	Preop patient selection/planning	78	63.1	47	Single institution
Tiulpin et al., 2019 [39]	TKA	ANN, regression, boosting	Preop patient selection/planning	3918	61.16/62.50	57.2/61.2	Osteoarthritis Initiative (OAI) and MOST data sets
Tolpadi et al., 2020 [40]	TKA, THA	ANN	Preop patient selection/planning	719	61	58	OAI database
Twiggs et al., 2019 [41]	TKA	Bayesian	Preop patient selection/planning	150	65.7	53	Single institution
Van et al., 2019 [29]	THA	ANN	Costs, preop patient selection/planning	100	—	—	Single institution
Yi et al., 2019 [42]	TKA	ANN	Preop patient selection/planning	154	—	—	Single institution
Yoo et al., 2013 [43]	TKA	SVM	Preop patient selection/planning	—	0	0	Single institution

ANN, artificial neural network; LOS, length of stay; ML, machine learning; SVM, support vector machine; THA, total hip arthroplasty; TJA, total joint arthroplasty; TKA, total knee arthroplasty.