Table 4.
Effect sizes (Hedges’ g) of ecological momentary intervention on mental health by study and intervention characteristics (within-subject analyses)a.
| Outcome | Random effect model | Heterogeneity | Test of difference | ||||||
| k b | nc | g (95% CI)d | Q e | I e | Q f | ||||
| Mental health |
|
|
27 | 1008 | 0.57 (0.45-0.70)g | 74.46g | 65.08 |
|
|
|
|
Design |
|
|
|
|
|
|
1.03 | |
|
|
|
RCTh | 11 | 481 | 0.65 (0.48-0.82)g | 24.10i | 58.50 |
|
|
|
|
|
Pre-post | 16 | 527 | 0.52 (0.33-0.71)g | 47.34g | 68.32 |
|
|
|
|
Sample |
|
|
|
|
|
|
1.79 | |
|
|
|
Clinical | 20 | 793 | 0.63 (0.50-0.76)g | 39.32i | 51.68 |
|
|
|
|
|
Healthy | 7 | 215 | 0.40 (0.10-0.71)j | 26.76g | 77.58 |
|
|
|
|
Agek, years |
|
|
|
2.19 | ||||
|
|
|
≤ 38.15 | 12 | 426 | 0.61 (0.36-0.86)g | 54.38g | 79.77 |
|
|
|
|
|
> 38.15 | 12 | 552 | 0.51 (0.37-0.64)g | 17.64l | 37.65 |
|
|
|
|
|
Unspecified | 3 | 30 | 0.80 (0.41-1.18)g | 0.40 | 0.00 |
|
|
|
|
Genderk |
|
|
|
|
|
1.96 | ||
|
|
|
≤ 60% female | 14 | 450 | 0.49 (0.28-0.70)g | 51.25g | 74.63 |
|
|
|
|
|
> 60% female | 11 | 550 | 0.67 (0.53-0.81)g | 15.94 | 37.26 |
|
|
|
|
|
Unspecified | 2 | 8 | 0.55 (−0.08 to 1.17)l | 1.12 | 10.43 |
|
|
|
|
Sample sizek |
|
|
|
1.18 | ||||
|
|
|
≤ 22 participants | 13 | 158 | 0.67 (0.46-0.87)g | 17.24 | 30.39 |
|
|
|
|
|
> 22 participants | 14 | 850 | 0.52 (0.36-0.69)g | 56.36g | 76.93 |
|
|
|
|
Training type |
|
|
|
0.32 | ||||
|
|
|
Active | 20 | 518 | 0.60 (0.42-0.78)g | 57.51g | 66.96 |
|
|
|
|
|
Passive | 7 | 490 | 0.53 (0.34-0.71)g | 16.65j | 63.97 |
|
|
|
|
Training trigger |
|
|
|
1.65 | ||||
|
|
|
Triggered | 9 | 535 | 0.52 (0.33-0.71)g | 26.96i | 70.45 |
|
|
|
|
|
On-demand | 11 | 256 | 0.49 (0.37-0.62)g | 9.41 | 0.00 |
|
|
|
|
|
Unspecified | 7 | 217 | 0.76 (0.38-1.14)g | 35.69g | 83.19 |
|
|
|
|
No. of daily training episodesk |
|
|
|
0.53 | ||||
|
|
|
≤ 2 | 7 | 370 | 0.55 (0.24-0.87)i | 32.65g | 81.62 |
|
|
|
|
|
> 2 | 6 | 259 | 0.51 (0.20-0.82)i | 22.81g | 78.08 |
|
|
|
|
|
Unspecified | 14 | 379 | 0.63 (0.49-0.77)g | 17.48 | 25.62 |
|
|
|
|
No. of total training episodesk |
|
|
|
0.92 | ||||
|
|
|
≤ 84 | 7 | 481 | 0.48 (0.21-0.75)i | 36.62g | 83.62 |
|
|
|
|
|
> 84 | 6 | 148 | 0.62 (0.27-0.97)i | 17.77i | 71.86 |
|
|
|
|
|
Unspecified | 14 | 379 | 0.63 (0.49-0.77)g | 17.48 | 25.62 |
|
|
|
|
Support MHPm |
|
|
|
6.77j | ||||
|
|
|
MHP-supported EMI | 14 | 474 | 0.73 (0.57-0.88)g | 20.67l | 37.10 |
|
|
|
|
|
Stand-alone EMI | 9 | 425 | 0.45 (0.22-0.69)g | 35.81j | 77.66 |
|
|
|
|
|
Stand-alone EMI with access to care as usual | 4 | 109 | 0.38 (0.11-0.64)i | 5.37 | 43.97 |
|
|
|
|
Quality assessmentk |
|
|
|
0.01 | ||||
|
|
|
≤ 3 | 17 | 781 | 0.57 (0.39-0.76)g | 57.68j | 72.26 |
|
|
|
|
|
> 3 | 10 | 227 | 0.59 (0.42-0.76)g | 16.78l | 46.38 |
|
|
aOutliers were excluded from the presented moderation analyses (ie, 6 studies).
bk=number of studies.
cn=number of participants.
dg=effect size Hedges’ g with 95% CI.
eQ and I2=heterogeneity statistics.
fQ=contrast between subgroups.
gP<.001.
hRCT=randomized controlled trial.
iP<.01.
jP<.05.
kData were categorized based on the median.
lP<.10.
mMHP=mental health professional.