What to do when an antidepressant isn’t working and the next step isn’t clear.
Work through the first decisions: was this a real trial, what kind of response you’re seeing, when to rethink the diagnosis—and what to do when the next step isn’t obvious.
Symptoms persist, improvement is partial or unclear, and you’re deciding whether to wait, increase, switch, augment, or rethink the frame entirely.
The question is not just “is the patient better?” It is whether you have enough structure to decide what should happen next.
Is this actually a failed treatment?
Before changing anything, the first question is whether this has actually been a meaningful trial.
- Was the dose still low or only partially titrated?
- Has enough time passed at a useful dose?
- Has adherence been consistent enough to judge response?
- Were expectations clear from the start?
You can recognize an incomplete trial — but when to wait, increase, or move on is less straightforward.
What kind of “not better” are you seeing?
Not all non-response is the same. The pattern matters before the plan does.
Are you sure depression is the main problem?
When treatment is not working, it is worth reconsidering what you are actually treating.
- Anxiety, ADHD, substance use, or sleep problems may be driving symptoms.
- Medical contributors or medications may be part of the picture.
- Bipolar-spectrum history can change the risk calculation.
- Chronic low mood may behave differently than an episodic major depressive episode.
Which next move are you leaning toward?
At this point, several reasonable options may be on the table — and they do not all lead to the same outcome.
Why this is harder than it looks
In practice, these decisions rarely line up cleanly.
- Partial improvement, but persistent functional impairment.
- Some benefit, but increasing side effects.
- No response, but an incomplete dose or adherence history.
- A patient who wants to stop medication despite some improvement.
- A history of “I’ve tried everything” without clear dose or duration details.
This is where most approaches break down
By now, you are trying to combine several decisions at once.
You are weighing trial adequacy, response type, side effects, diagnosis, patient expectations, and whether to optimize, switch, augment, or escalate.
There are general principles for each of these decisions. But it can be hard to find clear, step-by-step algorithms for how to put them together consistently across visits.
- You may know the options, but still be unsure which one fits this visit.
- You may see partial improvement, but not know when it is enough to keep going.
- You may move from one medication to the next without a clear threshold for changing course.
Continue with the full framework
You have seen the decision points. The harder part is knowing what to do when they do not all point in the same direction.
- Clarify the diagnosis and avoid common pitfalls like grief, burnout, medical contributors, or vague “depression/anxiety” labeling.
- Choose and start first-line treatment more deliberately.
- Set expectations with patients so early response, side effects, and delayed benefit do not derail care.
- Use a structured approach to deciding when to wait, increase, switch, augment, or escalate.
- Bring in behavioral activation, psychotherapy, exercise, and higher-level options like TMS, ECT, or ketamine when the frame needs to broaden.
Listen to the audio course through a private podcast feed while you’re on the go, or jump to specific sections when you want to hear the reasoning again.
The workspace is designed to be pulled up mid-visit—so you can move from uncertainty to a clear next step without leaving the encounter.
- Jump to the exact clinical question you’re facing
- Open the relevant flowchart or algorithm
- Review the reasoning behind the decision if needed
Includes: titration and augmentation flowcharts, medication selection tools, PHQ guide, differential checklist, side-effect guide, patient handouts and more.
If you want depression treatment decisions to feel less improvised — especially when the first medication does not produce a clean, obvious win — this is built for that.

