Table 1.
Characteristics of included studies.
| Study | n (E/C) | Age (yr) | Time since stroke | Primary outcome measures | Experiment | Control | Therapy dose | Results |
|---|---|---|---|---|---|---|---|---|
| Adie et al. (34) | 117/118 | E: 66.8 ± 14.6 C: 68 ± 11.9 |
E: 57.3 ± 48.3 (d) C: 56.3 ± 50.1 (d) |
ARAT | Home-based Wii grp | Home exercise handout | 45 min, daily for 6 wks | No between grp difference (MD: −1.7, 95% CI −3.9–0.5, p = 0.12) on ARAT score to improve UL function |
| Ballester et al. (35) | 17/18 | E: 65.1 ± 10.3 C: 61.8 ± 12.9 |
E: 1,073.4 ± 767.7 (d) C: 798.1 ± 421.8 (d) |
FM, CAHAI | Home-Based VR | Home-based OT | E: 26 min 40 s, 1–3 times/d, 5 d/wk, 3 wks C: 20 min, 1–3 times/d, 5 d/wk, 3 wks |
VR was more effective to improve UL function measured by CAHAI scale [1.53 (2.4), p = 0.01] than home-based OT |
| Barzel et al. (36) | 85/71 | E: 62.6 ± 13.7 C: 65.3 ± 13.7 |
E: 56.6 ± 47.4 (mo) C: 45.7 ± 57.7 (mo) |
MAL WFMT |
Home-Based CIMT | NDT clinic-based | E: 50–60 min, 5 times/5 wks + 40 h in 20 d of self-practice C: 25–30 min, 10 times/5 wks or 50–60 min, 5 times/5 wks |
Home-based CIMT grp improved more in MAL scores (MD: 0.26, 95% CI 0.05–0.46, p = 0.016) than NDT grp |
| Choudhury et al. (37) | 32/32 | E: 51 ± 12.1 C1: 53 ± 9.9 C2: 53.0 ± 10.6 |
E: 55 ± 142 (mo) C1: 43 ± 94 (mo) C2: 30 ± 29 (mo) |
ARAT, MA S, power and pinch strength, maximum force at wrist joint | Paired stim | C1: Random stim C2: Usual care |
4 h/d over 4 wks | Paired stim grp improved more ARAT (median baseline: 7.5, week 8: 11.5, p = 0.019) than the other two trainings |
| Cramer et al. (38) | 62/62 | E: 62 ± 14 C: 60 ± 13 |
E: 132 ± 65 (d) C: 129 ± 59 (d) |
FM | Home-Based telerehab | Clinic | 18 supervised and 18 unsupervised 70 min sessions, over 4 wks.; 5 min/d ×3 times of stroke education | No between grp difference on FM score (0.06, 95% CI −2.14–2.26, p = 0.96) was found |
| dos Santos-Fontes et al. (39) | 10/10 | E: 52.2 ± 11.1 C: 59.1 ± 11.1 |
E: 3.8 ± 4.5 (yr) C: 3.3 ± 2.1 (yr) |
JTT Compliance rate |
Home-Based RPSS stim | Sham | 2 h of stim daily before motor training, over 4 wks Motor training for 15 min, 2 times/d in 4 wks at home |
Electrical stim grp improved more in JTT performance than sham grp (14.3%, CI = 1.06–25.6%) |
| Duncan et al. (17) | 50/50 | E: 68.5 ± 9 C: 70.2 ± 11.4 |
E: 77.5 ± 28.7 (d) C: 73.5 ± 27.1 (d) |
OPS, FM, Grip strength, WMFT | Home therapeutic exercise | Usual care | E: 36 sessions, 90 min over 12–14 wks C: not specific |
The overall effect of therapeutic exercise had greater gain than usual care (Wilk's λ = 0.64, p = 0.0056) |
| Emmerson et al. (40) | 30/32 | E: 68 ± 15 C: 63 ± 18 |
E: 122 (77–193; d, median) C: 133 (58–228; d, median) |
Adherence rate WMFT |
Home-Based iPad grp | Home exercise handout | 1–2 times/d with no of exercises varied per d, for 4 wks | No between grp difference (MD: 0.02s, 95% CI −0.1–0.1) on WMFT log-transformed time to improve UL function |
| Hara et al. (41) | 10/10 | E: 56 C: 60.5 |
E: 13 (mo) C: 13 (mo) |
SIAS, ROM, MAS, 10-CMT, & 9-HPT | Home-Based FES grp | Clinic | E: 30 min, 5 d/wk for first 10 days, then 1 h/session, 5 d/wk for 5 mo C:40 min, once/wk for 5 mo |
Home-based FES was more effective to improve UL function than outpatient rehab (10-CMT: F = 18.72, p < 0.01) |
| Hsieh et al. (18) | 12/12 | E: 53.2 ± 19.2 C: 56.4 ± 18 |
E: 15.9 ± 13 (mo) C: 13.7 ± 11 (mo) |
FM, BBT, Revised NSA, MAL, 10 m walk, sit-to-stand test, COPM, EuroQoL-5D | Home-Based MT | MT in clinic | 75–105 min, for 12 sessions over 4 wks | Home-based MT grp improved more than clinic MT on MAL (p = 0.01) |
| Kimberly et al. (42) | 8/8 | E: 58.4 C: 62.8 |
E: 24.6 (mo) C: 38.5 (mo) |
Grip strength, BBT, MAL, JTT, Isometric finger extension strength | Home-Based NMES | Sham | 3–6 h, for 10 d over 3 wks | Home-based NMES improved arm function more than sham [BBT: t(7) = 2.06, p = 0.039; JTT: t(7) = 3.82, p = 0.003; MAL-AOU: t(7) = 7.6, p < 0.001; MAL-QOM: t(7) = 3.82, p = 0.003] |
| Mortenson et al. (43) | 8/8 | E: 65.5 C: 60.8 |
E: 32 (mo) C: 28.8 (mo) |
JTT, grip strength | Home-Based transcranial stim | Home-Based CT | 30 min per session, 5 times | Both groups improved in JTT over time (p < 0.01). Anodal grp improved more in grip strength than sham (p = 0.025) |
| Michielsen et al. (44) | 20/20 | E: 55.3 ± 12 C: 58.7 ± 13.5 |
E: 4.7 ± 3.6 (yr) C: 4.5 ± 2.6 (yr) |
FM, Grip strength, Tardieu scale, VAS, ARAT, ABILHAND, Stroke-ULAM, EQ-5D | Home-Based MT | Home-Based bilateral UL training | 1 h per session, 5 times/wk at home, 1 time/wk at center over 6 wks | MT grp improved more in FM than bilateral training grp after Rx (3.6 ± 1.5, p < 0.05) |
| Nijenhuis et al. (45) | 9/10 | E: 58 (48–65) C: 62 (54–70) |
E: 11 (10–26; mo) C: 12 (10–30; mo) |
IMI, FM, grip strength, MAL, ARAT, BBT, SIS | Home-Based robotic | Home-Based CT | 30 min per session, 5 times/wk over 6 wks at home | CT grp reported higher training duration (189 vs. 118 min per wk, p = 0.025). No between groups difference in UL outcomes (p ≥ 0.165) |
| Piron et al. (46) | 18/18 | E: 66.0 ± 7.9 C: 64.4 ± 7.9 |
E: 14.7 ± 6.6 C: 11.9 ± 3.7 |
FM, ABILHAND scale, Ashworth scale | Home-based telerehab | Clinic | 1 h per session, 5 times/wk over 4 wks at home | Telerehab grp improved more in FM (53.6 ± 7.7) than clinic (49.5 ± 4.8), p < 0.05 |
| Saadatnia et al. (9) | 20/20 | E: 62 ± 12.4 C: 66 ± 10.3 |
Nil data | BI, FM, MRS | Home-Based video exercise | Usual care (in clinic) | E: 1 h per session, 2 times/d, daily over 12 wks at home + usual care C: usual care |
Video exercise grp improved more in BI, FM, and MRS score than usual care grp (p < 0.001) |
| Standen et al. (47) | 17/10 | E: 59 ± 12 C: 63 ± 12 |
E: 22 (16, 59.5; mo) C: 12 (7.75, 20.25; mo) |
WMFT, 9-HPT, MAL, Nottingham extended activities of daily living | Home-Based Nintendo VR | No Rx | E: 20 min per session, 3 times/wk over 8 wks C: nil |
VR grp improved more than control grp in WMFT (r = 0.51, p < 0.05) at midpoint and MAL-AOU (r = 2.26, p < 0.05) at final point |
| Street et al. (48) | 6/6 | E: 53.2 ± 21.9 C: 67.6 ± 18.3 |
E: 19 (mo) C: 13.8 (mo) |
ARAT, 9-HPT | Home-Based (TIMP) | No treatment | E: 20–30 min per session, 2 times/wk over 6 wks C: nil |
No between grp difference in overall ARAT score 1.313 (SE:0.674, 95%CI: −0.073–2.698) and 9-HPT 0.169 (SE:0.823, 95%CI: −1.53–1.87) |
| Stinear et al. (49) | 16/16 | E: 57.9 (38–78) C: 52.6 (25–73) |
E: 28.8 (6–144; mo) C: 20.3 (6–73; mo) |
FM, NIHSS, grip strength | Home-Based (APBT) | Self-Directed task training | 10–15 min per session, 3 times/wk over 4 wks | APBT grp improved more UL function (p < 0.025) than control grp |
| Sullivan et al. (50) | 20/18 | E: 61.6 ± SD (37–88) C: 59.5 ± SD (41–85) |
E: 7.7 ± SD (1–29; yr) C: 6.6 ± SD (3–14; yr) |
FM, AMAT | Home-Based sensory electrical stimulation (SES) | Sham | 30 min, 2 times/d, 5 d/wk over 4 wks | No between grp differences but SES grp improved more on AMAT median time (p = 0.003, 95% CI:−1.4304, −6.365, effect size: 0.84) after Rx |
| Tariah et al. (51) | 10/8 | E: 54.8 ± 10.9 C: 60.6 ± 4.9 |
E: 9.2 ± 5.8 (mo) C: 9.6 ± 4 (mo) |
WMFT | Home-Based CIMT | Outpatient NDT | 2 h/d, 7 d/wk over 8 wks | CIMT grp improved more in WMFT-FAS [F(1, 15) = 12.68, p = 0.003] as compared to NDT grp |
| Turton et al. (52) | 24/23 | E: 66 (54.3, 75.1; median; IQR) C: 66.1 (57.6, 76.5; median; IQR) |
E: 111.5 (82, 241) (d) C: 135 (103, 171) (d) |
ARAT, WMFT | Home-Based reach-to-grasp (RTG) | Usual care | E:14 visits, 1 h/visit over 6 weeks + 56 h of self-practice C: not specific |
RTG grp improved 6 points for median score of ARAT after Rx but not the usual care grp |
| Wei et al. (8) | 32/25/27 | E: 59.2 ± 11.3 C1: 60.4 ± 10.4 C2: 63.1 ± 10.3 |
E: 47.8 ± 21.9 (d) C1: 61.1 ± 41.3 (d) C2: 53.7 ± 41.2 (d) |
FM, ARAT, BBT | Home-Based wearable device | C1: sham C2: usual care |
E & C1: 3 h/d,7 d/wk over 4 wks C2: not specific |
Wearable grp improved more in ARAT score than sham (MD = 6.283, 95% CI 0.812–11.752, p = 0.019) and control (MD = 5.767, 95% CI 0.299–11.235, p = 0.035) |
| Wolf et al. (53) | 51/48 | E: 59.1 ± 14.1 C: 54.7 ± 12.2 |
E: 115.5 ± 53.1 (d) C: 127.1 ± 46.2 (d) |
ARAT | Home-Based robotic | Home exercise handout | 3 h/d, 5 d/wk over 8 wks | Control group improved more in WMFT than robotic grp (p =0.012) |
| Zondervan et al. (54) | 8/8 | E: 61 ± 17 C: 54 ± 14 |
E: 39 ± 46 (mo) C: 24 ± 8 (mo) |
FM | Home-Based Resonating arm exercise (RAE) | Conventional therapy | 3 h/3 sessions/wk over 3 wks | Both groups improved in FM (p < 0.05) after Rx. RAE grp improved more in distal FM than CT (p = 0.02) |
| Zondervan et al. (55) | 9/8 | E: 60 (bib45–74) C: 59 (35–74) |
E: 5.33 ± 4.14 (y) C: 3.17 ± 1.66 (y) |
BBT, ARAT, MAL, & 9-hole peg test | Home-Based music glove (VR) | Home-Based task-specific training | 3 h/wk over at least 3 sessions/wk for 3 wks | No between grp difference in ARAT. VR grp improved more in both subscales of MAL (p = 0.007, p = 0.04) |
AMAT, Arm Motor Ability Test; ARAT, Action Research Arm Test; APBT, active passive bilateral training; BI, Barthel Index; BBT, Box and Block Test; C, control; COPM, Canadian Occupational Performance Measure; CAHAI, Chedoke Arm and Hand Inventory; CI, confidence index; CT, conventional therapy; E, experiment; EQ-5D-3L; d, days; FIM, functional independence measure; FM, Fugl-Meyer; h, hours; HEP, home exercise programme; IMI, Intrinsic Motivation Inventory; JTT, Jebsen–Taylor Test; MAL, Motor Activity Log; mo, months; MMT, manual muscle testing; MRS, Modified Rankin Scale; NIHSS, National Institute of Health Stroke Scale; 9-HPT, Nine hole Peg Test; OPS, Orpington Prognostic Scale; MT, mirror therapy; RNSA, Revised Nottingham Sensory Assessment; RCT, randomized controlled trial; RPSS, repetitive peripheral sensory; Rx, Treatment; SIS, Stroke Impact Scale; SIAS, Stroke Impairment Assessment Scale; SE, standard error; TEMPA, the upper extremity performance test; TIMP, therapeutic instrumental music performance;10-CMT, 10-cup-moving test; VAS, visual analog scale for pain; WMFT, Wolf Motor Function Test; WMFT-FAS, Wolf Motor Function Test -functional ability score; wk, weeks; yr, years; Rx, treatment.