Initial treatment choices for long term remission of insomnia disorder in adults: a systematic review and network meta-analysis

ABSTRACTBackgroundCognitive behavioral therapy for insomnia (CBT-I), pharmacotherapy and their combination are effective for insomnia. However, it remains unclear which treatment is more likely to lead to favorable long-term outcomes when used as the initial treatment. We aimed to evaluate the comparative efficacy and acceptability of CBT-I, pharmacotherapy, and their combination in the long- and short-terms among adults with insomnia disorder.MethodsWe searched PubMed, CENTRAL, PsycINFO and WHO ICTRP from database inception to Dec 27, 2023, to identify published and unpublished randomized controlled trials. We included trials in hypnotic-free adults with insomnia disorder comparing at least two of the following: CBT-I with at least one effective component (sleep restriction, stimulus control, cognitive restructuring, and third wave components), pharmacotherapy, or their combination. We assessed the confidence in evidence using CINeMA. The primary outcome was long-term remission (longest follow-up between 3 to 12 months). Secondary outcomes included all-cause dropout and self-reported sleep continuity measures at long-term follow-up, and the same outcomes at the end of the acute treatment phase. We performed frequentist random-effects network meta-analyses. We used odds ratio (OR) for dichotomous outcomes and mean difference for continuous outcomes, expressed in minutes and percent. This study is registered in PROSPERO (CRD42024505519).FindingsWe identified 13 trials, including 823 randomized participants (mean age, 47.8 years, 60% women). Results suggested that CBT-I was more beneficial than pharmacotherapy in the long-term (remission OR 1.82 [95% Confidence Interval (CI), 1.15 to 2.87; certainty of evidence: high]), while there was weaker evidence of benefit of combination against pharmacotherapy (OR 1.71 [95%CI, 0.88 to 3.30: moderate]) and no clear evidence of difference of CBT-I against combination (OR 1.07 [95%CI, 0.63 to 1.80: moderate]). CBT-I was associated with less dropouts than pharmacotherapy in the long-term. Short-term outcomes also favored CBT-I over pharmacotherapy except total sleep time. Given the average long-term remission rate in the pharmacotherapy-initiating arms of 28%, CBT-I resulted in a long-term remission rate of 41% (95% CI: 31% to 53%) and combination 40% (95% CI: 25% to 56%).InterpretationThis study found that starting with CBT-I for the treatment of adults with chronic insomnia leads to better outcomes than starting with pharmacotherapy. Combination therapy may be better than pharmacotherapy alone, but unlikely to be worth the additional burden over CBT-I alone.FundingNone.