Half the antidepressant switches that happen shouldn't have happened, and the other half should have happened six months earlier.
Sections
Half the antidepressant switches that happen shouldn’t have happened, and the other half should have happened six months earlier. The when-to-switch and when-to-stay decision is one of the most common calls in outpatient psychiatry, and the rules of thumb aren’t as complicated as the size of the drug menu would suggest. Here’s how it actually plays out.
When to switch
There are three real reasons to switch: no response, intolerable side effects, or poop-out, and each one has its own threshold and its own logic.
No response at eight weeks at a real dose
The most common switch trigger. If you’ve been on a therapeutic dose (not the starting dose, the actual therapeutic dose, which is usually double or triple where you started) for eight weeks, and you have no improvement at all, the drug isn’t working for you.
The key qualifier is “at a real dose.” A lot of guys get told to switch because the starting dose didn’t do anything, when the right move was to push the dose to the therapeutic range first and give it another four to six weeks before concluding anything. Lexapro 10 mg for six weeks isn’t an adequate trial, Lexapro 20 mg for eight weeks is. Sertraline 50 mg for six weeks isn’t adequate, sertraline 150 to 200 mg for eight weeks is. Most failed first trials are dose failures, not drug failures, and switching too early just resets the clock without addressing what actually went wrong.
Intolerable side effects
Some side effects are tolerable and worth pushing through. Nausea in week two. Sleep disruption the first month. Mild headache. These mostly fade and aren’t reasons to switch.
Other side effects aren’t going to fade and aren’t worth tolerating long-term. Severe sexual dysfunction that doesn’t respond to dose adjustment or to adding Wellbutrin. Significant weight gain on a guy who can’t afford metabolic consequences. Sustained sleep disturbance after the first month. A persistent out-of-it or flat feeling that doesn’t lift after the first six weeks. These are real reasons to switch, especially after attempted fixes have failed.
Poop-out (also called tachyphylaxis, which is the textbook word for the same thing)
Some patients who responded well to an antidepressant for one to several years find the response fading. The depression creeps back even though the drug and dose haven’t changed. The mechanism isn’t fully worked out, but the phenomenon is clinically real and shows up in maybe ten to twenty percent of long-term responders.
The first move when poop-out is suspected is to verify it. Sometimes what looks like poop-out is actually something else (a new life stressor, sleep getting wrecked, drinking creeping up, the thyroid drifting, sleep apnea developing as the patient gained weight). If the depression has genuinely returned despite continued adherence to a previously effective drug, then either pushing the dose or switching is reasonable. Push the dose first if there’s room. Switch if there isn’t.
When to stay
There are three real reasons to stay on a drug that isn’t quite right: partial response, side effects that are still fading, or the life around the medication has gotten worse in ways that make switching the wrong call right now.
Partial response
This is the most common scenario, the patient is better but not all the way better, maybe sixty percent improvement… mood is up but not in remission, sleep is improved but not great, energy is up but not normal.
The right move for partial response, more often than not, isn’t switching, it’s augmentation. The drug is doing something. Switching means giving up that something for the chance a different drug might do more, which sounds fine until you realize you’re trading a known sixty percent for an unknown outcome that might be worse. Augmentation means keeping the something you have and adding to it, which is usually the better trade for partial responders, and the data on augmentation versus switching agrees with the clinical experience here.
The augmentations that earn their keep are Wellbutrin added to an SSRI (cleans up sexual side effects, adds motivation, augments the antidepressant effect, see the Wellbutrin post), lithium at 300 to 600 mg (best-evidenced augmentation with independent suicide-reduction data, the labs are a hassle but the result for partial responders is real), Abilify at 2 to 5 mg (works fast, watch the metabolic stuff and the risk of akathisia which is the inner-restlessness feeling some patients get), and T3 thyroid hormone at 25 to 50 micrograms (old strategy, still works for specific patients, mostly forgotten in current practice).
Side effects that are still fading
If you’re in week three or four and the side effects are bothersome but the mood is starting to lift, give it time. Most of the side effect profile peaks in weeks one through three and improves substantially by weeks six through eight. Switching at week four because of side effects is often switching just before the side effects would have eased off on their own. Patience here usually pays off.
If at week eight to twelve the side effects are still significant and the mood is partially or fully responding, then we’re in mitigation territory (add a counter-side-effect drug, like Wellbutrin for sexual issues) or switching territory. Before week eight, riding it out is usually right.
The life around the drug has gotten worse
This is the one nobody talks about. Sometimes the drug is working but the patient’s life has slid in ways that obscure the drug’s effect. New job stress, marriage problems, sleep wrecked by a new baby, an exercise routine that fell off, drinking that crept from two to five. The temptation in this situation is to switch the medication because the patient is suffering more than they were six months ago. The actual fix is usually to deal with whatever in the life situation has fallen apart, not to change the drug.
This conversation comes up a lot, and it’s an uncomfortable one because what the patient wants is for the prescriber to fix it with a different pill, and what’s actually needed is for the patient to have the hard conversation with his wife or quit drinking or admit his job is killing him. Switching the drug in this situation often makes things worse, because we lose the partial benefit we had and we add the side effects of starting a new drug, all to avoid the conversation that’s actually going to move the needle.
The augmentation question, more directly
Augmentation usually beats switching for partial responders. The data leans this direction, the clinical experience leans this direction, and the practical advantage is that augmentation lets you build on existing benefit rather than restart the clock. Plus you don’t have to white-knuckle through week two again.
Augmentation also beats switching for side-effect issues if the side effect can be addressed by the add-on. Wellbutrin for sexual side effects. Low-dose mirtazapine for sleep or appetite. Buspirone for residual anxiety. Each of these turns a partial-response-with-side-effects situation into a partial-response-without-side-effects situation, which is usually a better outcome than starting over with a different primary drug.
For non-responders (zero response, not partial), switching usually beats augmentation. If the drug isn’t doing anything at eight weeks at a real dose, augmenting it with a second drug usually just adds the side effects of the second drug to the failure of the first. Switch to a different mechanism instead.
The cross-taper, which patients underestimate
When we do switch, the cross-taper is where the felt-suffering of a switch happens. Most SSRI-to-SSRI switches can be done with a fairly fast cross-taper over one to two weeks (drop the old one as the new one comes up). SSRI to Wellbutrin similar.
The harder switches are anything involving Effexor, Cymbalta, or Paxil, all of which have a brutal discontinuation profile and need to come down slowly. Going from Effexor to anything else usually takes four to eight weeks of overlap, sometimes longer if the patient’s been on it for years. Going to or from an MAOI (an older class of antidepressants that come with a strict food list to follow, because eating aged cheese or fermented stuff on them can spike your blood pressure to dangerous levels, see the depression post for the MAOI digression) requires a washout period to avoid serotonin syndrome, which is the rare but real complication where serotonin levels stack and the patient ends up in the ER.
The cross-taper is the underappreciated piece. The transition between drugs is where most of the felt-suffering of a switch happens, and a careful cross-taper can be the difference between a switch that lands smoothly and a switch that puts the patient in two months of feeling worse than they did on the old drug.
What’s nice to hear in all this
Most of these decisions work out fine if the prescriber is patient and the patient is honest, which is the part that gets buried when everyone’s focused on what can go wrong. People who augment partial responses to sixty percent end up at eighty-five percent. People who switch carefully off a drug that wasn’t working find one that does. The drugs do work for most of the people who go through the full process, and that tends to get lost in the risk-focused version of this conversation. The story doesn’t have to end with treatment-resistant depression. It mostly doesn’t.
The pattern that ends up augmenting instead of switching
Say you’ve got a guy who comes in wanting to switch, he’s been on Lexapro 20 mg for two years with about sixty percent improvement of his depression. Side effects are minimal. Sexual function is somewhat decreased but not absent. Marriage is fine, work is fine, he’s just not all the way back, and he wants to swap the drug because something different has to be better than this. We talk about it. Switching means giving up the sixty percent on a gamble the new drug does more. Augmenting means keeping the sixty percent and adding something on top to push for the missing piece. We add 150 mg Wellbutrin XL in the morning. Six weeks later he’s at eighty-five percent improvement, his sexual function is back, he has more morning energy than he had in a while, and he stays on the combination. The reflex to switch when something isn’t perfect is often the wrong reflex. Building on what’s working is usually the move.
If your life has slid and you want to switch the drug because of it, that’s usually just avoiding the conversation you actually need to have.
What to actually do
If you’re not responding at all at eight weeks at a real dose, switch. Within class or across class based on judgment.
If you’re partially responding, augment before you switch. Wellbutrin for sexual side effects or low motivation, lithium for the broader augmentation effect, Abilify if you need a fast response.
If your side effects are bad but mood is improving and you’re inside the first eight weeks, ride it out. Most side effects fade.
If your life has gotten worse and you want to switch because of it, address the life first. The drug usually isn’t the problem.
If poop-out is real, push the dose first if you can, then switch.
Bottom line
Switching too fast and staying too long are roughly equally common, which is awkward but true. The threshold for either decision is whether you’ve actually given the existing drug a fair shake at a real dose for a real length of time (usually no), and whether what’s wrong is the drug or something else in your life that’s slid since you started (often the second). Both questions get skipped more than they should, which is why this post exists.
Sources
- Papakostas GI. Managing partial response or nonresponse: switching, augmentation, and combination strategies for major depressive disorder. J Clin Psychiatry. 2009;70(Suppl 6):16-25. PMID 19922740.
- Rush AJ, Trivedi MH, Wisniewski SR, et al. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry. 2006;163(11):1905-1917. PMID 17074942.
- Fornaro M, Anastasia A, Novello S, et al. The emergence of loss of efficacy during antidepressant drug treatment for major depressive disorder: An integrative review of evidence, mechanisms, and clinical implications. Pharmacol Res. 2019;139:494-502. PMID 30385364.
- Clayton AH, Warnock JK, Kornstein SG, Pinkerton R, Sheldon-Keller A, McGarvey EL. A placebo-controlled trial of bupropion SR as an antidote for selective serotonin reuptake inhibitor-induced sexual dysfunction. J Clin Psychiatry. 2004;65(1):62-7. PMID 14744170.
