Table 2 Summary statistics of remission outcomes

From: Technology-assisted cognitive-behavioral therapy for perinatal depression delivered by lived-experience peers: a cluster-randomized noninferiority trial

Analysis population

Analysis methods

Remission outcome

n/N (%) of participants with events

P value

THP-TAP (N = 487)

WHO-THP (N = 493)

Risk difference (two-sided 95% CI)a

ITT

Crude analysis

Remission from MDE at 3 months postnatal (primary outcome)

395/428 (92.3)

349/418 (83.5)

8.91 (4.25, 13.56)

<0.0001c

Remission from MDE at 6 months postnatal

389/421 (92.0)

375/410 (89.7)

2.91 (−1.30, 7.11)

0.176

Adjusted analysisb

Remission from MDE at 3 months postnatal (primary outcome)

395/428 (92.3)

349/418 (83.5)

9.22 (4.46, 13.97)

<0.0001c

Remission from MDE at 6 months postnatal

389/421 (92.0)

375/410 (89.7)

2.62 (−1.98, 7.21)

0.2643

Crude analysis with missing outcomes imputed

Remission from MDE at 3 months postnatal (primary outcome)

449/487 (92.2)

411/493 (83.4)

8.88 (4.41, 13.34)

<0.0001c

Remission from MDE at 6 months postnatal

448/487 (92.1)

442/493 (89.7)

2.14 (−2.21, 6.48)

0.3356

Per protocol

Crude analysis

Remission from MDE at 3 months postnatal (primary outcome)

389/421 (92.4)

345/410 (84.1)

8.32 (3.72, 12.91)

<0.0001c

Remission from MDE at 6 months postnatal

387/421 (91.9)

368/410 (89.8)

2.83 (−1.44, 7.09)

0.1939

Adjusted analysisb

Remission from MDE at 3 months postnatal (primary outcome)

389/421 (92.4)

345/410 (84.1)

8.68 (3.99, 13.38)

<0.0001c

Remission from MDE at 6 months postnatal

387/421 (91.9)

368/410 (89.8)

2.51 (−2.12, 7.13)

0.2884

Crude analysis with missing outcomes imputed

Remission from MDE at 3 months postnatal (primary outcome)

391/423 (92.4)

352/418 (84.2)

8.25 (3.64, 12.86)

<0.0001c

Remission from MDE at 6 months postnatal

389/423 (92.0)

375/418 (89.7)

2.91 (−1.30, 7.11)

0.176

  1. MDE, current major depressive episode diagnosed with SCID MDE module.
  2. aRisk difference was calculated using the GEE model.
  3. bAdjusted risk difference was calculated by the GEE model weighted by inverse probability of treatment calculated from logistic regression based on the prespecified covariates (age, parity, household income and PHQ-9 score at baseline).
  4. cP value for one-sided noninferiority test.
Back to article page