Maintenance treatment aims at preventing new episodes (recurrences) and thus is appropriate for recurrent or chronic depression but not for single-episode major depressive disorder. Maintenance medication is effective in virtually all studies reported to date. Strong evidence suggests that patients with three or more major depressive episodes are candidates for maintenance-phase treatment (Frank et al. 1990; Reynolds et al. 1999). Indeed, even at 5 years in highly recurrent major depressive disorder, maintenance medication has prophylactic efficacy (Kupfer et al. 1992). How strongly to recommend maintenance treatment for those with only two major depressive episodes is less clear. Factors that might indicate the need for maintenance treatment in such situations are a previous incomplete interepisode recovery (Depression Guideline Panel 1993; Judd et al. 1997, 1998a, 1998b), the presence of two episodes within the last 3 years, and a positive family history of depression or bipolar disorder, any of which, when present, point toward a higher likelihood of an earlier new episode (recurrence) than would be the case for patients without such histories. In any case, these factors need to be weighed by clinicians and patients together to decide on the utility of maintenance-phase treatment. An alternative, if maintenance treatment is declined, is to diligently monitor symptoms with symptom self-reports so that the development of a new episode can be detected early. Should a new episode develop, early intervention can shorten the length of the episode.
Symptom breakthrough may be encountered in continuation- and maintenance-phase treatment. Symptom breakthrough may be only modest and time limited, requiring little shift in the plan for continuation- or maintenance-phase treatment (e.g., dose adjustment, reassurance) (Rush 1999). On the other hand, if symptom breakthrough is profound, prolonged, disabling, or unresponsive to dose adjustment and reassurance, practitioners must decide how it should be treated in the absence of randomized controlled trials addressing this problem. Perhaps the simplest approach is to augment the current medication with an additional medication (e.g., lithium, thyroid hormone, or another antidepressant). Should this strategy prove effective, then the augmenting medication may be discontinued after a time to empirically evaluate whether it is necessary over the longer term. If the augmenting medication fails, then switching to a different treatment may be needed.
If symptom breakthrough occurs, it could as well be remediated by psychotherapy, but this option has not been formally studied. Perhaps psychotherapy would be indicated if the symptoms were caused by disturbed interpersonal relationships or life events (e.g., divorce or unemployment).
Another tactical problem encountered in continuation- and maintenance-phase treatment is the management of the depression in the context of 1) intercurrent general medical illnesses requiring medication, 2) the need for surgery, or 3) pregnancy. Because pregnancy may last for a prolonged period, and given the evidence for the efficacy of IPT alone in maintenance treatment (Frank et al. 1990; Reynolds et al. 1999), psychotherapy without medication may provide a clinically useful drug-free period. Alternatively, fluoxetine (Chambers et al. 1996; Pastuszak et al. 1993) (or perhaps other SSRIs) may be considered. The development of other general medical illnesses and the need for nonpsychotropic medications during continuation- and maintenance-phase treatment are not uncommon. These circumstances need to be managed, taking into account pharmacokinetics and drug interactions with continuation/maintenance medication.
When to discontinue maintenance medication treatment is unclear. As noted at the beginning of this section, findings from a 5-year study of patients with highly recurrent depression (i.e., more than three episodes) indicate that the efficacy of maintenance-phase treatment continues for at least 2-5 years. Some patients may require lifetime maintenance medication or very prolonged periods (e.g., a decade) of treatment. When discontinuation occurs, careful monitoring after discontinuation is needed, because the first 6 months after discontinuation appear to constitute a period of particular risk for recurrences.
Revision date: July 8, 2011
Last revised: by Janet A. Staessen, MD, PhD