Antidepressants do not improve everything at once, and the order in which things improve confuses people. Sleep, appetite, and physical energy usually shift first, somewhere in the first two weeks. Concentration follows. Mood, the thing the patient actually came in about, is typically last, arriving somewhere between four and six weeks, with continued gains out to eight or twelve.
This creates a strange and very common experience. A patient tells me they are sleeping better and getting to the gym again but feel no less depressed, and they read that as failure. I read it as the medication doing precisely what it does, in the order it does it, and I want to hold the course.
A fair trial is six to eight weeks at an adequate dose. Anything less is not evidence.
This is where I find the most correctable mistakes. A patient arrives having failed three antidepressants, and when I look at the history, all three were stopped at the starting dose. The starting dose exists to let the body adjust, not to treat the illness. For most of these medications, there is meaningful room above it, and a sizable group of people who respond to nothing at 10mg respond well at 30 or 40.
So before I conclude a drug has failed, I want to know it was pushed to a genuinely therapeutic dose and held there long enough to judge. If it wasn’t, the honest answer is that we haven’t tested this medication yet.
This distinction drives the decision more than anything else.
If someone has improved somewhat but is not where they need to be, the medication is doing something, and I would rather build on it than discard it. That means raising the dose if there is room, or adding a second agent alongside it. Augmentation with an atypical antipsychotic, lithium, or thyroid hormone has a solid evidence base, and adding buspirone, bupropion, or mirtazapine to an existing SSRI is common and often well tolerated.
If someone has felt nothing at all after a proper trial at a proper dose, augmentation makes less sense. There is little to build on. That is when I switch.
The first question is whether to stay within the same class or leave it. If the first SSRI produced no response whatsoever, moving to another SSRI has a lower chance of success than changing mechanism, so I usually move to an SNRI, to bupropion, or to mirtazapine depending on the symptom picture. If the issue was side effects rather than effectiveness, staying in class often works fine, because tolerability varies a great deal between drugs that look similar on paper.
Symptoms shape the choice too. Significant fatigue and difficulty concentrating point me toward bupropion. Insomnia with weight loss points toward mirtazapine. Pain alongside the depression makes an SNRI more attractive. Anxiety that dominates the picture makes me cautious about anything activating and slower with the titration.
How we make the change matters as much as what we change to. Most switches are done as a cross taper, bringing one down while bringing the other up over a couple of weeks, which avoids leaving someone unmedicated in the middle. Certain combinations require a full washout for safety, and those are worth being rigid about.
One thing to watch for during any switch: discontinuation symptoms from the medication coming off can look like the depression returning. Dizziness, flu like feelings, electric shock sensations, irritability, vivid dreams. These start within days of a dose reduction and fade within a week or two, whereas a true relapse builds gradually. Patients who mistake one for the other sometimes conclude the old drug was working after all and want to go back, when what they are feeling is a taper that moved too fast.
There is a point where trying a fifth and sixth antidepressant stops being a plan. Broadly, if two well-conducted trials at adequate doses have failed, the odds that the seventh one is the answer are not good, and continuing to work through the list costs months.
That is the point to widen the approach rather than repeat it. TMS therapy and IV infusion treatment both exist for people in exactly this position, and they work through entirely different mechanisms than oral antidepressants do. It is also worth re-examining the diagnosis at that stage. Depression that consistently resists treatment is sometimes bipolar depression, or depression with an untreated anxiety disorder, trauma history, thyroid problem, or sleep disorder sitting underneath it.
The honest version. Not the version that avoids sounding difficult. When you started skipping doses and for how long, whether the side effect you mentioned briefly is actually the reason the bottle is still half full, whether you feel worse in a way that worries you. A medication change made on incomplete information is a guess, and guesses cost weeks.
If you are on something that does not feel like it is working and nobody has walked you through where you stand in this sequence, that conversation is overdue. You can schedule a visit here, by telemedicine if getting to the office is the obstacle, or read more about how I approach psychiatric care and the practice itself before you decide.
