The Psychopharmacology Algorithm Project at the Harvard South Shore Program presents evidence-based recommendations considering efficacy, tolerability, safety, and cost. Two previous algorithms for unipolar nonpsychotic depression were published in 1993 and 1998. New studies over the last 20 years suggest that another update is needed.
The references reviewed for the previous algorithms were reevaluated, and a new literature search was conducted to identify studies that would either support or alter the previous recommendations. Other guidelines and algorithms were consulted. We considered exceptions to the main algorithm, as for pregnant women and patients with anxious distress, mixed features, or common medical and psychiatric comorbidities.
For inpatients with severe melancholic depression and acute safety concerns, electroconvulsive therapy (or ketamine if ECT refused or ineffective) may be the first-line treatment. In the absence of an urgent indication, we recommend trialing venlafaxine, mirtazapine, or a tricyclic antidepressant. These may be augmented if necessary with lithium or T3 (triiodothyronine). For inpatients with non-melancholic depression and most depressed outpatients, sertraline, escitalopram, and bupropion are reasonable first choices. If no response, the prescriber (in collaboration with the patient) has many choices for the second trial in this algorithm because there is no clear preference based on evidence, and there are many individual patient considerations to take into account. If no response to the second medication trial, the patient is considered to have a medication treatment–resistant depression. If the patient meets criteria for the atypical features specifier, a monoamine oxidase inhibitor could be considered. If not, reconsider (for the third trial) some of the same options suggested for the second trial. Some other choices can also considered at this stage. If the patient has comorbidities such as chronic pain, obsessive-compulsive disorder, attention-deficit/hyperactivity disorder, or posttraumatic stress disorder, the depression could be secondary; evidence-based treatments for those disorders would then be recommended.