Pharmacotherapy for sleep disturbances in post-traumatic stress disorder (PTSD): A network meta-analysis

Background: Sleep disturbances are an important symptom dimension of post-traumatic-stress-disorder (PTSD). There is no meta-analytic evidence examining the effects of all types of pharmacotherapy on sleep outcomes among patients with PTSD.

Methods: Medline/Embase/PsychInfo/CENTRAL/clinicaltrials.gov/ICTRP, reference lists of published reviews and all included studies were searched for Randomised Controlled Trials (RCTs) examining any pharmacotherapy vs. placebo or any other drug among patients with PTSD.

Primary outcomes: total sleep time, nightmares, sleep quality.

Secondary outcomes: sleep onset latency, number of nocturnal awakenings, time spent awake following sleep onset, dropouts due to sleep-related adverse-effects, insomnia/somnolence/vivid-dreams as adverse-effects. Pairwise and network meta-analyses were performed.

Results: 99 RCTs with 10,481 participants were included. Prazosin may be the most effective treatment for insomnia (SMD = -0.88, 95%CI = [-1.22;-0.54], nightmares (SMD = -0.44, 95%CI = [-0.84;-0.04]) and poor sleep quality (SMD = -0.55, 95%CI = [-1.01;-0.10]). Evidence is scarce and indicates lack of efficacy for SSRIs, Mirtazapine, z-drugs and benzodiazepines, which are widely used in daily practice. Risperidone and Quetiapine carry a high risk of causing somnolence without having a clear therapeutic benefit. Hydroxyzine, Trazodone, Nabilone, Paroxetine and MDMA-assisted psychotherapy may be promising options, but more research is needed.

Conclusions: Underpowered individual comparisons and very-low to moderate confidence in effect estimates hinder the generalisability of the results. More RCTs, specifically reporting on sleep-related outcomes, are urgently needed.