Inpatient stroke rehabilitation: prediction of clinical outcomes using a machine-learning approach

Table 1 Demographic information, stroke characteristics, and clinical tests of study participants (N = 50)

Numeric variables
Characteristic	Mean	SD	Range
Age (y)	57.5	14.15	22–86
Height (cm)	172.9	11.3	149.9–195.6
Weight (kg)	81.3	19.9	42.7–118
BMI (kg/m²)	27.1	5.76	16.7–44.7
Education (y)	14.9	3.3	6–20
Rehab duration (days)	18.7	10.4	4–57
Time from stroke to admission (days)	18.8	29.6	3–181
Categorical variables
Characteristic	#	%
Sex
Male	31	62
Female	19	38
Race
White	24	48
African American	21	42
Hispanic	4	8
Asian	1	2
Lifestyle
Sedentary	21	42
Moderately active	11	22
Highly active	18	36
Stroke type
Ischemic	39	78
Hemorrhage	11	22
Cognitive/Communication diagnosis
Yes	37	74
No	13	26
Speech diagnosis
Yes	37	74
No	13	26
Language diagnosis
Yes	10	20
No	40	80
Dysphagia
Yes	35	70
No	15	30
Clinical tests at admission
Test	Mean	SD	Range
FIM	55.9	19.1	17–98
TMWT (m/s)	0.47	0.49	0–1.55
SMWT (m)	103.34	122.77	0–461
BBS	20.6	16.26	1–55
Clinical tests at discharge
Test	Mean	SD	Range
FIM	82.5	19.5	42–120
TMWT (m/s)	0.76	0.61	0–2.06
SMWT (m)	188.38	163.44	0–562
BBS	29.6	17.41	0–56

ISSN: 1743-0003