How do you choose an antidepressant?

You see a 24 yo woman in a severe major depressive episode and offer her sertraline and a referral to therapy. After the visit, the medical student rotating with you asks why you chose sertraline for this patient. He follows with, “I guess I’m asking…how do you choose an antidepressant?”

You answer his question in three parts.

1) Which patients with depressive symptoms are likely to benefit from treatment with antidepressants?
A patient with major depression that is moderate or severe, or a patient with persistent depressive disorder (formerly dysthymia) may benefit from treatment with an antidepressant. Patients with subthreshold depression or major depression that is mild in severity improve as much (on average) from placebo pills as from antidepressants (Vöhringer, 2011). Review this PsychSnap to learn more about subthreshold depression and persistent depressive disorder. 

2) Is this patient willing to consider a medication for their symptoms?
The best antidepressant is the one that the patient is willing to take. And some patients will not take any. So ask first. I find these questions helpful.

“What do you think is going on with you?”  
“What do you think might help?”
And if it makes sense,
“What are your thoughts on taking a medication for [whatever they have called the problem]?”

3) Which antidepressant? A three step approach.

Use what worked before:
If they have felt like this before, what medication helped them and at what dose? Restart the antidepressant that previously worked and titrate as quickly as possible to the same dose that worked before. If I know that a patient needed sertraline 150mg daily to treat a previous depressive episode, I try to get them to that dose within 2 weeks.The titration schedule could be: sertraline 50mg pills, ½ tab (25mg) x4 days, then 1 tab (50mg) daily x5 days, then 2 tabs (100mg) daily x5 days, then 3 tabs (150mg) daily, slowing down the titration if side effects like GI upset are problematic.

Don’t use what didn’t work before:
If they have felt like this before, what did they try that wasn’t helpful or that they couldn’t tolerate? Don’t do that again. This may require waiting a week to request and review outside records, or naming a bunch of antidepressants to see if any of them sound familiar to the patient. A full antidepressant trial lasts about 3 months, so an extra week or two to clarify what didn’t work before may reduce patient suffering overall.

Start with an SSRI (escitalopram, sertraline) or bupropion:*
If you know how to use escitalopram, sertraline, and bupropion, you can provide first-line antidepressant treatment to most people with depression who present in primary care.

The SSRIs are effective, reasonably well-tolerated, and robustly studied. SSRIs are first-line treatments for major depression (moderate-severe) AND generalized anxiety disorder, social anxiety disorder, panic disorder, PTSD, and OCD.  For more information on why escitalopram & sertraline are “first line” SSRIs, including a helpful reference table for choosing between SSRIs, check out this PsychSnap.

Bupropion is another first-line treatment for major depression. Unlike the SSRIs, bupropion does not treat anxiety disorders, PTSD, or OCD.  Bupropion is activating and for some people, can increase nervousness, cause agitation or create sleep problems. But people taking bupropion don’t gain weight or develop sexual side effects – the two side effects that most commonly cause people to stop SSRIs. For people who have significant anxiety as part of a major depressive episode but do not have a separate comorbid anxiety disorder, bupropion can be a reasonable choice for the treatment of depression, particularly if they have not tolerated SSRIs. Bupropion can be used as monotherapy for depression, or as an augmentation agent in combination with SSRIs or SNRIs without worries of causing serotonin syndrome. For practical tips on prescribing bupropion, check out the second half of this PsychSnap. 

To summarize: An antidepressant may be an appropriate treatment for a patient who has moderate to severe major depression or persistent depressive disorder and who is open to treatment with medications.

When choosing an antidepressant:

  1. Use what worked before
  2. Don’t use what didn’t work before
  3. Start with either an SSRI (if there is a comorbid anxiety disorder, PTSD, or OCD) or bupropion (if sexual side effects or weight gain from SSRIs are concerning)*

*There are times when a psychiatrist might start with an SNRI like venlafaxine or an antidepressant like mirtazapine instead of an SSRI or bupropion. For someone with insomnia and low appetite, mirtazapine might be a good first line antidepressant. For someone with neuropathic pain and depression, an SNRI would be a reasonable first line treatment. While there are more tolerability problems with SNRIs, mirtazapine, and tricyclic antidepressants, they may be marginally more effective for depression than SSRIs or bupropion, which can have clinical relevance in patients with very severe depression requiring psychiatric care (Figure 3, Cipriani, 2018). With these important caveats, the vast majority of cases of new depression that present in primary care could be managed initially with the above algorithm. 


Cipriani, Andrea, et al. “Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis.” Focus 16.4 (2018): 420-429.

Vöhringer, Paul A., and S. Nassir Ghaemi. “Solving the antidepressant efficacy question: effect sizes in major depressive disorder.” Clinical therapeutics 33.12 (2011): B49-B61.