Effectiveness of Action Observation Training on Upper Limb Motor Function in Children with Hemiplegic Cerebral Palsy: A Systematic Review of Randomized Controlled Trials

Figures & data

Table 1 Methodological Quality Assessment of Included RCTs

PEDro Items	Buccino et al, 2012^Citation18	Sgandurra et al, 2013^Citation20	Young Kim et al, 2014^Citation32	Kirkpatrick et al, 2016^Citation35	Kim et al, 2018^Citation36	Buccino et al, 2018^Citation34	Simon-Martinez et al, 2019^Citation33	Simon-Martinez, et al, 2020^Citation31	Kim, 2020^Citation30
Eligibility criteria	Yes	No	Yes	Yes	Yes	Yes	No	Yes	Yes
Random allocation	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Concealed allocation	No	Yes	No	Yes	No	No	No	No	No
Baseline comparability	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Blind participants	Yes	No	No	No	No	Yes	No	No	No
Blind therapists	No	No	No	No	No	No	No	No	No
Adequate follow-up	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes
Blind assessors	Yes	Yes	No	Yes	Yes	Yes	No	No	No
Intention to treat analysis	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Between group comparisons	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Point estimates and variability	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Total score out of 10	8/10	8/10	6/10	8/10	7/10	7/10	6/10	6/10	6/10
Graded approach scale	High	High	Moderate	High	Moderate	Moderate	Moderate	Moderate	Moderate

Table 2 Summary of Included Randomized Controlled Trials

Authors, Publication Years	Participant Characteristics	Interventions	Outcome Measures	Results/Conclusion
Buccino et al, 2012^Citation18	−15 total participants: EG; (n=8), CG; (n=7) -Sex: 9 males and 6 females, age rage: 7 to 11 years -Median age/years: EG; 7 and 6 Months, CG; 8 years. GMFCS: not stated	-EG: asked to observe videos with motor content (every day arm⁄ hand actions), afterwards to imitate them for 3 weeks. -CG: asked to observe videos with no motor specific content and afterwards to execute the same actions as the case group for 3 weeks. -Both groups received routine conventional rehabilitation program.	-Melbourne Assessment Scale	- After treatment, the functional score gain (D) was significantly different in the case and control groups (p=0.02). A Bonferroni post-hoc test indicated that motor scores as measured by the Melbourne Assessment Scale at T3 (mean 88.3) improved compared with T2 (mean 85.1; p=0.04) and T1 (mean 83.9; p=0.004). A Bonferroni post-hoc test showed that an increased score on the Melbourne Assessment Scale was achieved only in the case group (T1=86.8, T2=88.0, T3=94.2; T1 vs T3, p=0.002; T2 vs T3, p=0.01), but not in the control group (T1=80.5, T2=81.8, T3=81.5; T1 vs T3, p=1.0; T2 vs T3, p=1.0). -Action observation treatment can be an effective part of the rehabilitation program in children with CP.
Sgandurra et al, 2013^Citation20	−24 total participants: EG; (n=12), CG; (n=12) -Sex (M/F): 16/8 -Age: mean age (years) (SD): EG; 9.48 (2.12), CG; 9.94 (2.77) Range: EG; 6.4–12.6, CG; 6.1–14.3. GMFCS: not stated	-EG: observed goal directed Videos each lasted 3 minutes and child was required to perform the same observed action for 3 minutes in the exact order and with the same objects as shown in the video for 60 minutes per day for 3 consecutive weeks. -CG: watched computer games and then were verbally instructed to perform the same for 60 minutes per day for 3 consecutive weeks.	-AHA - MUUL ABILHAND-Kids	-The experimental group improved more (p=0.008) in score changes for the AHA at the primary endpoints T1 (p=0.008), T2 (p=0.019), and T3 (p=0.049). No between-group significant changes were found for ABILHAND-Kids or Melbourne assessment. -The Upper Limb Children Action Observation Training (UP-CAT) improved daily upper limb (UL) activities with UCP children, suggesting a new rehabilitation approach based on a neurophysiological model of motor learning.
Young Kim et al, 2014^Citation32	−16 total participants: EG; (n=8), CG; (n=8) -Sex (M/F) 8/8 -Mean age/year (SD): EG; 9.13 (2.36) CG; 9.25 (3.15). GMFCS: EG; 2.38 ± 0.74, CG; 2.25 ± 1.16	EG: received AOPT & practiced repeating the actions they observed on video clips for twelve 30-min sessions, 3 days a week, for 4 weeks. For each video five minutes and and 10 minutes per tasks, two videos were shown during each training session. CG: PT group performed the same actions as the AOPT group did after observing landscape photographs	-MAS -BBT -WeeFIM -Grasp power	-The results of this study showed that, in comparison with the physiotherapy group, the functioning of the upper extremities in the AOT group was significantly improved in body functions and activity and participation according to the ICF framework. Comparison between the groups before and after training indicated significantly more improvement for the action observation physical training group than for the physical training group (p<0.05); MAS in AOT (−0.6 ± 0.3) vs in PT (−0.2 ± 0.3), BBT in AOT (7.8 ± 2.9) vs in PT (2.3 ± 1.9), ABILHAND-Kids in AOT (3.7 ± 3.0) vs PT (1.6 ± 2.6) and WeeFIM in AOT (1.1 ± 0.8) vs (0.7 ± 0.7). -Indicates that AOT has a positive influence on the functioning of the upper extremities in children with CP.
Kirkpatrick et al, 2016^Citation35	−59 total participants: EG; (n=28), CG (n= 31) -Sex (M/F): 28/31 -Median age (IQR): EG; 5y 2mo (4y), CG; 5y 4mo (3y 4mo), age range, 3 to 10 years. GMFCS: not stated	-EG: received AO plus RP 15 minutes/day, 5 days/week for 3 months CG: received 15 minutes/day, 5day/week for 3 months.	- AHA -MA2 ABILHAND-Kids	-There were no between-group differences in AHA, MA2, or ABILHAND-Kids at 3 or 6 months versus baseline (all p>0.05). -Combined-group improvements (p<0.001), observed in AHA and MA2 at 3 months, were maintained at 6 months. ABILHAND-Kids also showed improvement at 3 months (p=0.003), maintained at 6 months. -This study did not demonstrate a significant improvement in hand function with AO plus repeated practice (RP) compared to RP alone with children age 3 to 10 years with UCP.
Kim et al, 2018^Citation36	−12 total participants: EG; (n=6), CG (n=6) -Sex (M/F): 7/5 - Age ranged from 8 to 11 years. GMFCS: not stated	-EG: received live AOT for 20 treatment sessions, each 30 minutes in duration, 5 days/week for a month. -CG: received video AOT for 20 treatment sessions, each 30 minutes in duration, 5 days/week for 4 weeks.	-JTHF -BBT	- JTHF scores were significantly lower in the live group (p<0.05). The BBT score was significantly higher in the live than in the video group (p< 0.05). The findings suggest that live AOT is more effective than video AOT for improving upper limb movement acceleration and function. The functional measures, JTHF scores were significantly better for the live group (167.6 ± 10.6 s) than for the video group (181.2 ± 9.5 s) at the follow-up test (p=0.03). Scores on the BBT were significantly higher in the live group (59.6 ± 8.8 EA) than in the video group (47.8 ± 2.9 EA) at the follow-up test (p=0.03).
Buccino et al, 2018^Citation34	−18 total participants: EG; (n=11), CG; (n=7) -Age (years): 5–11 -Sex (M/F); 9/9 GMFCS: ranged from 1 to 4.	- EG: observed videos with motor content (every day arm/hand actions) and perform each action plus CPT for each action 9–12 minutes, 5 times/week for 3 weeks -CG: observed videos with no specific motor content plus CPT for each action 9–12 minutes, 5 times/week for 3 weeks	- MUUL - AHA	- Significant interaction between time and treatment, both for MUUL (b1 interaction=2.7, t36=3.9, and p=0.00) and for AHA (b1 interaction=2.3, t18=3.6, p=0.002). Score improvements, in both scales, were higher in the treated participants than in the controls; furthermore, in the treatment group, those improvements were not only maintained but became even stronger at T3. Post hoc analysis showed that for MUUL, results at T2 were significantly different from the results at T1 only in cases (p<0.001), but not in controls. As for AHA, results at T2 were significantly different from the results at T1 (p<0.001). Even more interestingly, results at T3 were different from the results at T2 (p<0.001) for both scales, but again only in cases, but not in controls.
Simon-Martinez et al, 2019^Citation33	−36 total participants: EG; (n=20), CG; (n=16) -Mean age (SD): 9 years (1 year) in both groups -Sex (males): MACS (n (%)); EG, I, II, III, 5 (25), 7 (35), 8 (40) and CG, 3 (19), 6 (37), 7 (44) respectively	-EG: received mCIMT plus AOT, 15 hours of AOT, ie video-observation and execution of unimanual tasks at 6 months follow-up. -CG: received mCIMT plus placebo, watched biological-motion free videos and executed the same tasks at 6 months follow-up.	-Motor control parameters (RU, RGV)	-The combination of mCIMT plus AOT resulted in a shorter movement time during a reaching task compared to those who did not receive additional AOT. Both groups showed improvement in RU and RGV (p=0.07) with large effect sizes (partial η2=0.15 for both tasks. Post-hoc analyses for RU showed significant improvements between T1-T3 (p=0.02) and T2-T3 (p=0.03). Improvements during RGV were found immediately after the intervention (T1-T2, p=0.04) and maintained at follow-up (T1-T3, p=0.02). This indicated the improvement in lateral flexion during RU (p=0.03, 67–100% of the movement cycle), although post-hoc analyses did not survive Bonferroni correction. -The mCIMT plus placebo group showed no significant improvements in movement duration (p>0.05). The added AOT training may have a long-lasting effect on movement duration during reaching
Simon- Martinez et al, 2020^Citation31	−44 total participants: EG; (n=22), CG; (n=22). Sex(males): 27. Mean age 9 years 6 months (1 year 10 months) MACS I, 15 MACS II, and 20 MACS III], and were allocated to the CIMT + AOT group (n=22) and CIMT + placebo group (n= 22)	EG: Received the CIMT plus AOT group received 15 hours of AOT (ie, video-observation) and executed the observed tasks. CG: the CIMT plus AOT group watched videos free of biological motion and executed the same tasks.	- AHA -MAS -MA2 ABILHAND-Kids	- Although no differences between-groups were observed (p>0.05; η2=0–16), the addition of AOT resulted in higher improvements in children with initially lower bimanual performance (p=0.02; η2=0.14). Both groups improved in all outcome measures after the intervention and retained the gains at follow-up (p<0.01; η2=0.02–0.71). Poor sensory function resulted in larger improvements in the total group (p=0.03; η2=0.25) and high amounts of mirror movements tended to result in a better response to the additional AOT training (p=0.06; η2=0.18). The total group improved over time in the AHA (p<0.001). - Adding AOT to CIMT resulted in a better outcome for children with poor motor function and high amounts of mirror movements.
Kim, 2020^Citation30	−10 total participants: EG; (n=5), CG; (n=5) Sex (male: female): EG; (2: 3), CG; (2: 3) Mean age (years; EG; (10.2 (1.3)), CG; (9.4 (1.3)). EG; 3(14) MACS I, 7 (32) MACS II, and 12(55) MACS III, CG; 6 (27) MACS I, 8 (36.5) MACS II, and 8 (36.5) MACS III.	EG: observed video tasks of goal-directed action before performing tasks, for 30 minutes per day, three times per week, for 4 weeks. CG: observed and performed the same actions, 60 minutes per day, three times per week, for 4 weeks.	-Grip power ABILHAND Kids - QUEST	-Grip power, and ABILHAND-Kids test results improved significantly in both groups (p<0.05). The ABILHAND-Kids test within-group score improved significantly, from 1.6 ± 0.4 logits to 2.3 ± 0.7 logits in the 30 minutes training group (p=0.04) and from 1.4 ± 0.3 logits to 2.1 ± 0.5 logits in the 60 minute group (p=0.04). However, there was no between-group difference in the 30 minutes (0.7 ± 0.3 logits) or the 60 minute group (0.6 ± 0.1 logits; p=1.0). Grip strength at the follow-up evaluation was significantly higher, from 3.8 ± 1.7 to 4.8 ± 1.7 in the 30 minutes training group (p=0.03) and from 4.1 ± 1.7 to 5.1 ± 2.0 in the 60 minute group (p=0.04). However, there was no significant difference between the 30 minutes (1.0 ± 0.3) and 60 minutes groups (1.0 ± 0.5) at the end of each intervention (p=0.8). The 30 minutes AOT was as effective as 60 minutes AOT in improving grip power and upper limb function in children with CP. QUEST scores improved significantly, from 68.8 ± 7.5% (pre-test) to 77.6 ± 6.4% (post-test) in the 30 minutes (p=0.04) and from 64.9 ± 6.1% (pretest) to 72.3 ± 3.5% (post-test) in the 60 minutes groups (p=0.04).

Abbreviations: AHA, Assisting Hand Assessment scale; UCP, unilateral cerebral palsy; MUUL, Melbourne Assessment of Unilateral Upper Limb Function; MA2, Melbourne Assessment 2; MAS, Modified Ashworth Scale; BBT, Box and Block Test; WeeFIM, Wee Functional Independent Measure; AOPT, action observation physical training; JTHF, Jebsen–Taylor Hand Function; BBT, Box and Block Test; mCIMT, modified constraint-induced movement therapy; RU, reach upwards; RGV, reach-to-grasp a vertically oriented cylinder; MAS, Modified Ashworth Scale; QUEST, Quality of Upper Extremity Skills Test; SD, standard deviation; EG, experimental group; CG, control group; GMFCS, Gross Motor Function Classification System; IQR, interquartile range; MACS, Manual Ability Classification System.

Figure 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) diagram.

Note: Adapted from Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLoS Med. 6(7): e1000097.³⁷

Buccino G, Arisi D, Gough P, et al. Improving upper limb motor functions through action observation treatment: a pilot study in children with cerebral palsy. Dev Med Child Neurol. 2012;54(9):822–828. doi:10.1111/j.1469-8749.2012.04334.x22765352

 PubMed Web of Science ®Google Scholar

Sgandurra G, Ferrari A, Cossu G, et al. Randomized trial of observation and execution of upper extremity actions versus action alone in children with unilateral cerebral palsy. Neurorehabil Neural Repair. 2013;27(9):808–815. doi:10.1177/154596831349710123886886

 PubMed Web of Science ®Google Scholar

Kim JY, Kim J-M, Ko EY. The effect of the action observation physical training on the upper extremity function in children with cerebral palsy. J Exerc Rehabil. 2014;10(3):176. doi:10.12965/jer.14011425061598

 PubMedGoogle Scholar

Kirkpatrick E, Pearse J, James P, et al. Effect of parent‐delivered action observation therapy on upper limb function in unilateral cerebral palsy: a randomized controlled trial. Dev Med Child Neurol. 2016;58(10):1049–1056. doi:10.1111/dmcn.1310927038153

 PubMed Web of Science ®Google Scholar

Kim D-H, An D-H, Yoo W-G. Effects of live and video form action observation training on upper limb function in children with hemiparetic cerebral palsy. Technol Health Care. 2018;26(3):437–443. doi:10.3233/THC-181220

 Web of Science ®Google Scholar

Buccino G, Molinaro A, Ambrosi C, et al. Action observation treatment improves upper limb motor functions in children with cerebral palsy: a combined clinical and brain imaging study. Neural Plast. 2018;2018:1–11. doi:10.1155/2018/4843985

 Web of Science ®Google Scholar

Simon-Martinez C, Mailleux L, Jaspers E, et al. Effects of combining constraint-induced movement therapy and action-observation training on upper limb kinematics in children with unilateral cerebral palsy: a randomized controlled trial. Sci Rep. 2020;10:1–15.31913322

 PubMed Web of Science ®Google Scholar

Simon-Martinez C, Mailleux L, Hoskens J, et al. Randomized controlled trial combining constraint-induced movement therapy and action-observation training in unilateral cerebral palsy: clinical effects and influencing factors of treatment response. Ther Adv Neurol Disord. 2020;13:1756286419898065. doi:10.1177/175628641989806532031542

 PubMed Web of Science ®Google Scholar

Kim DH. Comparison of short-and long-time action observation training (AOT) on upper limb function in children with cerebral palsy. Physiother Pract Res. (Preprint):1–6.

 Google Scholar

Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. PLoS Med. 6(7):e1000097.19621072

 PubMed Web of Science ®Google Scholar