Skip to main content

Table 3 Evidence profile table

From: Effect of brain-computer interface training based on non-invasive electroencephalography using motor imagery on functional recovery after stroke - a systematic review and meta-analysis

Certainty assessment

Nr. of patients

Effect

Quality of evidence

Importance

Nr. of studies

Study

design

Risk of bias

Inconsis-tency

Indirect-ness

Imprecision

Other

considerations

BCI

Usual therapy

Relative(95% CI)

Absolute(95% CI)

Motor function recovery upper extremity (follow-up: range 3 to 8 weeks; assessed with: Fugl-Meyer Assessment UE; scale from 0 to 54 or to 66 and MFT; scale from 0 to 32)

11

randomised trials

not serious

not serious

not serious

not serious

none

174

155

SMD 0.39 SD (0.17 to 0.62)

HIGH

IMPORTANT

: Motor recovery after stroke lower extremity (follow-up: range 3 days to 5 days; assessed with: Fugl-Meyer LE, BBS; scale from: 0 to (FM) 34 or (BBS) 56)

2

randomised trials

serious

not serious

serious a

serious b

none

18

14

SMD 0.41 SD

(−0.29 to 1.12)

VERY LOW

CRITICAL

: Brain function recovery (follow-up: range 5 days to 8 weeks; assessed with: Functional connectivity change indices, Attention index, Activation index and Lateralization index))

5

randomised trials

not serious

not serious

not serious

serious

none

51

34

SMD 1.11 SD (0.64 to 1.59)

MODERATE

IMPORTANT

: Short-term Follow-up (follow-up: mean 12 weeks; assessed with: Fugl-Meyer UE)

4

randomised trials

serious c

not serious

not serious

not serious

none

34

46

SMD 0.31 SD

(−0.11 to −0.54)

MODERATE

CRITICAL

: Long-term Follow-up (follow-up: range 24 weeks to 36 weeks; assessed with: Fugl-Meyer UE)

2

randomised trials

not serious

not serious

not serious

serious d

none

28

20

SMD 0.56 SD (0.01 to 1.11)

MODERATE

CRITICAL

  1. Legend: Nr. Number; BCI Brain-computer interface technology; usual therapy = without BCI; CI confidence interval; FM-LE Fugl-Meyer lower extremity; FM-UE Fugl-Meyer upper extremity; MFT Manual Function Test; BBS Berg Balance Scale; SMD standardised mean difference
  2. a = too short intervention, no generalizability, no transferability; b = small sample size; c = reduced intensity for BCI group, 136 repetitions per session vs 1040 sessions in control group in one study and less motor imagery trials in BCI group in second study; d = wide CI (uncertain of magnitude of the effect)