Relapse of depression more likely after discontinuation of medication (number needed to treat to harm = 6).

Reference

Lewis G, Marston L, Duffy L, et al. Maintenance or discontinuation of antidepressants in primary care. N Engl J Med 2021;385(14):1257-1267.

Synopsis

These researchers identified primary care patients who had at least 2 episodes of depression or who had been taking antidepressants for at least 2 years. All patients felt well enough to consider discontinuing their medication; those with current depressive symptoms were excluded. The patients who had been taking been taking citalopram 20 mg, fluoxetine 20 mg, sertraline 100 mg, or mirtazapine 30 mg for at least 9 months were randomized to either continue their medication or to discontinue their medication over a 2-month period by substituting placebo. At baseline, the mean age of participants was 54 years, 73% were women, 95% were white, approximately half were taking citalopram, and approximately three-quarters had been taking an antidepressant for more than 3 years. The primary outcome was relapse of depression, defined as a 2-week spell of feeling sad, miserable, or depressed, or being unable to enjoy or take an interest in things as much as usual, accompanied by at least one other depressive symptom. After one year, this outcome occurred significantly more often in the placebo group (56% vs 39%; hazard ratio [HR] 2.06; 95% CI 1.56 - 2.70; number needed to treat to harm [NNTH] = 6). Depressive and anxiety symptoms were, on average, more severe in the discontinuation group. In addition, more patients in the discontinuation group stopped taking their placebo (48% vs 30%; HR 2.28; 1.68 - 3.08) and more resumed antidepressant medications (39% vs 20%). The mean scores on the 7-item Generalized Anxiety Disorder Questionnaire and the 9-question Patient Health Questionnaire were also significantly higher in the discontinuation group.