Guideline recommendations for long-term treatment of depression with antidepressants in primary care-a critical review

Background: Long-term treatment with antidepressants is considered effective in preventing recurrence of major depressive disorder (MDD). It is unclear whether this is true for primary care. Objectives: We investigated whether current guideline recommendations for long-term treatment with antidepressants in primary care are supported by evidence from primary care. Methods: Data sources for studies on antidepressants: PubMed, Cochrane Library, Embase, PsycInfo, Cinahl, articles from reference lists, cited reference search. Selection criteria: adults in primary care, continuation or maintenance treatment with antidepressants, with outcome relapse or recurrence, (randomized controlled) trial/naturalistic study/review. Limits: published before October 2009 in English. Results: Thirteen depression guidelines were collected. These guidelines recommend continuation treatment with antidepressants after remission for all patients including patients from primary care, and maintenance treatment for those at high risk of recurrence. Recommendations vary for duration of treatment and definitions of high risk. We screened 804 literature records (title, abstract), and considered 27 full-text articles. Only two studies performed in primary care addressed the efficacy of antidepressants in the long-term treatment of recurrent MDD. A double-blind RCT comparing mirtazapine (n = 99) and paroxetine (n = 98) prescribed for 24 weeks reported that in both groups 2 patients relapsed. An open study of 1031 patients receiving sertraline for 24 weeks, who were naturalistically followed-up for up to two years, revealed that adherent patients had a longer mean time to relapse.

Conclusions : No RCTs addressing the efficacy of maintenance treatment with antidepressants as compared to placebo were performed in primary care. Recommendations on maintenance treatment with antidepressants in primary care cannot be considered evidence-based.