Table 4 Transdiagnostic-focused apps vs control at follow-up

From: Transdiagnostic-focused apps for depression and anxiety: a meta-analysis

Analysis

Ncomparisons

g (95% CI)

I2

NNT

PI

Combined outcomes

 Total effect

11

0.25 (0.10, 0.41)

44%

12.9

−0.13, 0.63

  One effect per study (smallest)

10

0.25 (0.09, 0.41)

48%

12.9

−0.15, 0.65

  One effect per study (largest)

10

0.25 (0.09, 0.41)

48%

12.9

−0.16, 0.66

  High risk of bias excluded

8

0.31 (0.14, 0.47)

28%

10.6

−0.04, 0.64

  Trim-and-fill estimate

 

0.14 (0.01, 0.28)

 

24.1

 

Depression

 Total effect

10

0.21 (0.04, 0.38)

51%

15.6

−0.22, 0.64

  One effect per study (smallest)

9

0.21 (0.04, 0.39)

57%

15.6

−0.26, 0.68

  One effect per study (largest)

9

0.22 (0.04, 0.39)

56%

14.9

−0.25, 0.59

  High risk of bias excluded

7

0.27 (0.07, 0.48)

49%

11.9

−0.24, 0.78

  Trim-and-fill estimate

 

0.21 (0.04, 0.37)

 

20.9

 

Anxiety

 Total effect

10

0.25 (0.10, 0.40)

40%

12.9

−0.11, 0.61

  One effect per study (smallest)

9

0.25 (0.09, 0.40)

44%

12.9

−0.12, 0.62

  One effect per study (largest)

8

0.25 (0.09, 0.41)

46%

12.9

−0.13, 0.63

  High risk of bias excluded

7

0.30 (0.14, 0.47)

19%

10.6

0.01, 0.58

  Trim-and-fill estimate

 

0.08 (−0.07, 0.24)

 

43.2

 
  1. NNT Number-Needed-To-Treat, PI Prediction Interval.