Discontinuing Psychotropic Medication Before Psychotherapy for PTSD

Background and rationale A previous prospective cohort study of more than 6,000 patients receiving standardized intensive trauma-focused psychotherapy for PTSD at a high-volume psychotrauma expertise center during 2021-2024 found that concomitant use of several psychotropic medications was associated with smaller reductions in PTSD symptoms (https://doi.org/10.1159/000549259). Analyses adjusted for 27 baseline covariates. Psychotropic co-medication overall was associated with reduced symptom improvement compared with non-use, and this association persisted at 6-month follow-up. Antidepressants overall, amitriptyline, and mirtazapine showed consistent associations with poorer outcomes across sensitivity analyses. Anticonvulsants, antipsychotics, mood-stabilizing anticonvulsants, fluoxetine, zolpidem, and zopiclone showed similar negative associations, although estimates were less consistent across sensitivity analyses.

An important threat to causal inference in comparisons of psychotropic medication users with non-users is confounding by indication: patients receiving psychotropic medication had greater baseline PTSD symptom severity and psychiatric comorbidity. The present study will address this problem by comparing patients using a specific medication under two different clinical policies: patients who continued that medication during psychotherapy in 2021-2024 and patients for whom discontinuation of the same medication will be initiated before psychotherapy under the revised 2026 policy. Together with adjustment for baseline differences, this same-medication comparison will substantially improve clinical comparability. The principal remaining alternative explanation is a secular change in psychotherapy outcomes or patient characteristics between the historical and post-policy periods, which will be examined in dedicated sensitivity analyses. This design will significantly reduce residual confounding and strengthens the plausibility of a causal interpretation.

The purpose of this policy-driven natural experiment will therefore be to investigate whether tapering and discontinuing selected psychotropic medications before intensive trauma-focused psychotherapy, when clinically appropriate, is associated with greater improvement in PTSD symptoms than continuation of the same medications during psychotherapy.

Study design This will be an observational target trial emulation based on a policy-driven natural experiment. In September 2026, the center PSYTREC (Zeist, the Netherlands) will introduce a revised clinical medication-review protocol in which current and recent psychotropic medication use and treatment indication are systematically verified using available clinical and pharmacy information.

The target medications are amitriptyline, mirtazapine, quetiapine, topiramate, zolpidem, and zopiclone when prescribed off-label for PTSD-related symptoms. Under the revised policy, tapering and discontinuation before psychotherapy are initiated when clinically appropriate and safe. Medication management remains part of routine clinical care rather than a randomized research intervention. Tapering schedules are individualized according to medication, dose, duration of use, symptoms, withdrawal effects, and clinical circumstances. Psychotherapy can be postponed or the medication strategy modified when clinically necessary.

The 2026 post-policy cohort will consist of patients using one or more target medications for whom tapering and discontinuation are initiated under the revised protocol. The principal historical comparator will consist of patients treated in 2021-2024 who were using the same target medication or medications at treatment onset, during a period in which these medications were generally continued during psychotherapy. Classification of the 2026 policy cohort will therefore be based on initiation of the discontinuation strategy rather than successful completion of discontinuation.

All patients receive the center's standardized intensive trauma-focused psychotherapy program, including prolonged exposure and EMDR therapy. The primary clinical question is whether the strategy of discontinuing a target medication before psychotherapy is associated with greater improvement in clinician-rated PTSD symptoms than continuation of the same medication during psychotherapy. Longer-term PTSD symptoms and the feasibility and clinical consequences of medication discontinuation will also be evaluated.

Analytical approach Separate target-trial analyses will evaluate discontinuation versus continuation for the target medications. Propensity-score-based and outcome-modeling methods will be used to improve comparability between the 2026 and 2021-2024 cohorts and adjust for clinically relevant baseline factors associated with medication treatment or psychotherapy outcome. These include demographic characteristics, baseline PTSD symptom severity, trauma history, psychiatric comorbidity, previous psychotherapy, concomitant psychotropic medication use, and treatment duration.

Primary propensity-score matching will use nearest-neighbor matching without replacement, with up to two historical controls per 2026 patient and a caliper of 0.2 standard deviations of the logit of the propensity score. Covariate balance will be evaluated using standardized mean differences (<0.10) and variance ratios (0.5-2.0). If adequate balance cannot be achieved with nearest-neighbor matching, full matching will be used. Outcome models will additionally adjust for relevant baseline covariates.

Sensitivity analyses

Secular changes in psychotherapy outcomes will be evaluated in two policy-unaffected populations by comparing propensity-score-matched patients treated in 2026 with comparable patients treated in 2021-2024:

1. patients using no psychotropic medication, and
2. patients using psychotropic medication but none of the target medications. Specific psychotropic medication use will be included among the matching variables in the latter comparison.

Difference-in-differences analyses will then compare the temporal change for each target-medication group with the temporal change observed in the non-target medicated comparator cohort. The difference-in-differences estimate therefore represents the change in outcome among target-medication patients beyond the contemporaneous change observed among comparable patients whose medication management was unaffected by the discontinuation policy. Statistically, this effect will be estimated from the interaction between target-medication group and calendar period in an adjusted outcome model.

Pre-policy outcome trends during 2021-2024 will be examined for evidence that target-medication and comparator groups already showed differential temporal trends before implementation of the 2026 policy.

E-values will additionally quantify the strength of residual unmeasured confounding that would be required to explain the estimated associations.

Feasibility of discontinuation Clinical information collected during medication review will be used to characterize implementation of the discontinuation strategy, including completion or partial completion of tapering, modifications to the planned taper, clinically relevant discontinuation or rebound symptoms, temporary substitution or other medication changes, and postponement of psychotherapy. These analyses will describe the practical feasibility and clinical consequences of implementing the medication-review policy in routine PTSD care.