There is no symptom score that tells you it is time to start a psychiatric medication. The more useful test is functional: how much is this costing you, how long has it been going on, and has anything you have already tried moved it? Medication becomes worth considering when symptoms are persistent, are interfering with work, sleep or relationships, and have not responded to the changes available to you.
Just as importantly, considering medication is not the same as committing to it. A good evaluation ends with a decision you helped make — including, sometimes, the decision not to prescribe anything yet.
What actually makes medication worth considering?
A few patterns come up repeatedly:
- Duration. Symptoms present most days for weeks, rather than a hard fortnight that resolved.
- Interference. Work, sleep, parenting or relationships are measurably affected.
- Diminishing returns. Therapy, exercise, sleep changes or reduced drinking have helped some, but not enough.
- Physical symptoms. Sleep that will not come, appetite changes, or a body that stays braced all day.
- Access. Symptoms are severe enough that engaging with therapy is itself hard — a case where medication sometimes lowers the volume enough for other work to be possible.
What is not a good test: whether other people think you are struggling enough to deserve treatment.
Does starting medication mean staying on it forever?
No, and that assumption stops people from asking about it at all. Duration depends on the diagnosis, the history, and what you want — a first episode of depression is a different conversation from a fourth. It is a question that gets revisited at follow-up visits rather than settled once at the start.
Tapering, when it is clinically appropriate, is a normal part of the work. It is done gradually and with monitoring — not by stopping on your own between appointments, which is the most common way a reasonable plan goes wrong.
How long before you know whether it is working?
Longer than most people are told, which is why so many first attempts get abandoned early.
The STAR*D study — the largest real-world trial of depression treatment, following 2,876 outpatients treated with citalopram in primary and psychiatric care settings — found that a substantial portion of participants who responded or reached remission did so at or after eight weeks of treatment (Trivedi et al., 2006). Remission rates on that first medication were 28% by clinician-rated scale and 33% by self-report, with a response rate of 47%.
Two things follow from that. First, patience during a properly monitored trial is not passivity — it is how the trial produces an answer. Second, a first medication not working is a common, expected outcome rather than evidence that nothing will. Roughly half of people did not respond to that first agent, and the study existed precisely to work out what to do next.
How do you tell whether it is actually working?
“Do you feel better?” is a weak instrument. Memory for how bad things were six weeks ago is unreliable, and improvement is often gradual enough to be invisible from the inside.
The alternative is measurement-based care: brief, structured symptom rating scales completed regularly, with the results used during the appointment to make decisions. A review of 51 articles found that virtually all randomised controlled trials with frequent, timely feedback of patient-reported symptoms to the provider during medication management significantly improved outcomes — while one-time screening, infrequent assessment, and feeding results back outside the clinical encounter did not (Fortney et al., 2017).
Practically, that means a good follow-up visit should be able to answer “is this better than last month, and by how much?” with something more durable than a shared impression.
What if you have tried medication before and it did not help?
That history is one of the most useful things you can bring to an evaluation, provided it is specific. What was tried, at what dose, for how long, and what exactly happened — no effect, partial effect, or side effects that ended it before the trial was complete?
Those distinctions change the next step entirely. A medication stopped after ten days because of nausea has not actually been tested. One taken at a low dose for three months without change is a different situation. “Nothing works for me” frequently turns out to mean “nothing has yet been given a fair trial with monitoring.”
What if the answer is no medication?
Sometimes it is, and a prescriber should be willing to say so. Where sleep, alcohol, an untreated medical issue or the absence of therapy is doing most of the work, starting a medication on top of that treats the symptom and leaves the driver running. At NCNE, medication is used judiciously and strategically rather than as a first or only solution — which necessarily means there are visits that end without one.
Getting started
NCNE provides adult psychiatric medication management by secure video to people located in Connecticut, Massachusetts, New Hampshire, Maine, Rhode Island and Virginia. If you are weighing this decision, the initial evaluation is where it gets worked through properly — there is no obligation to leave with a prescription. You can send a message and we will respond within two business days, or check insurance and fees first.
If you are in immediate danger, call 911. For urgent support at any hour, call or text 988 (Suicide & Crisis Lifeline). This article is general information, not medical advice. Never start, stop or change a prescribed medication without speaking to your prescriber.
References
Fortney, J. C., Unützer, J., Wrenn, G., Pyne, J. M., Smith, G. R., Schoenbaum, M., & Harbin, H. T. (2017). A tipping point for measurement-based care. Psychiatric Services, 68(2), 179–188. https://pubmed.ncbi.nlm.nih.gov/27582237/
Trivedi, M. H., Rush, A. J., Wisniewski, S. R., Nierenberg, A. A., Warden, D., Ritz, L., … Fava, M. (2006). Evaluation of outcomes with citalopram for depression using measurement-based care in STAR*D: Implications for clinical practice. The American Journal of Psychiatry, 163(1), 28–40. https://pubmed.ncbi.nlm.nih.gov/16390886/