Traumatic childbirth can lead to persistent posttraumatic stress and depressive symptoms. Evidence on the durability of Eye Movement Desensitization and Reprocessing (EMDR) and on the optimal timing of intervention in the perinatal context remains limited.

ACTES is a single-centre exploratory naturalistic study conducted at Strasbourg University Hospitals (France). Women with childbirth-related traumatic stress were treated within one of three a priori care pathways defined by the timing of EMDR initiation: immediate postpartum (in-hospital Recent Traumatic Episode Protocol, n = 8), early postpartum (1–12 months, standard 8-phase EMDR, n = 8), or late postpartum ( > 12 months, standard 8-phase EMDR, n = 8). A baseline PCL-5 score ≥ 30 was used to define probable PTSD. The PCL-5, Edinburgh Postnatal Depression Scale (EPDS), and Subjective Units of Disturbance (SUD) were collected pre-treatment (T0) and at 1, 6, 12, and 18 months post-treatment (T1–T4). Reliable change between T0 and T1 was estimated using the Reliable Change Index. Post-treatment trajectories were modelled with non-parametric mixed-effects models based on the Aligned Rank Transform.

Of 36 women included, 24 (67%) completed treatment and all four post-treatment assessments. Baseline symptom burden was high (median PCL-5 47.0; SUD 9.0; EPDS 13.0). 20 of 24 women (83%) showed a reliable PCL-5 improvement at T1. 15 of the 20 women initially above the PCL-5 cut-off (75%) crossed below the threshold at T1. Whole-sample median scores remained below baseline at 18 months (PCL-5 12.0; SUD 2.0; EPDS 4.0). In T1–T4 models, only the time effect on SUD reached significance (F(3, 66) = 3.03; p = 0.036); no group effect or interaction was detected.

These exploratory observations are consistent with the clinical feasibility of an integrated EMDR care pathway across three timing windows in a tertiary maternity setting. The small sample size and the absence of a comparison condition preclude any conclusion regarding the comparative efficacy of intervention timing or protocol; confirmatory trials with adequate statistical power are required.