Why Your Psychiatrist Should Also Do Therapy
By John Nicholas Jung Shumate, MD, JD | The Jeong Center for Psychiatry and Psychotherapy
Last updated: May 13, 2026
Seeing one clinician for both medication and therapy means your prescriber and your therapist share the same understanding of who you are (because they're the same person). This integrated model leads to more precise medication decisions, fewer missed diagnoses, and a deeper therapeutic relationship, but it has become rare because insurance incentives push most psychiatrists toward brief medication-only visits.
There is a word in Korean (jeong, 정) that describes a bond formed not through a single grand gesture but through accumulated presence: the slow, patient work of showing up. It has no clean English equivalent.
The closest approximation might be deep connection built over time, but that flattens something the original holds in tension (the way such bonds require both vulnerability and constancy, the way they cannot be rushed or contracted out). I think about this word often, because the model of care that dominates American psychiatry has, in many ways, engineered jeong out of the equation.
A patient I'll call Priya (a junior physician in her early thirties who had spent a decade learning to carry other people's emergencies while quietly suffocating under her own) once described her previous psychiatric care this way:
"I see my therapist on Tuesdays. I see my prescriber once every three months for fifteen minutes. They've never spoken to each other. I'm the telephone between them."
She said this with the flat, diagnostic precision of someone who had been charting her own treatment and found the notes incomplete. She was not wrong.
Why do most psychiatrists no longer do therapy?
The short answer is economics, not evidence. Over the past two decades, insurance reimbursement structures have pushed psychiatrists toward brief medication-management visits (fifteen minutes, sometimes less) and away from the longer sessions that therapy requires. The data is stark: according to a nationally representative study published in the American Journal of Psychiatry, the percentage of psychiatrist visits involving psychotherapy dropped from 44.4% in 1996 to just 21.6% by 2016. By the mid-2010s, 53% of U.S. psychiatrists provided no psychotherapy at all. An even smaller number specialize in therapy.
This did not happen because the evidence turned against integrated care.
It happened because the math favored volume. A psychiatrist seeing four patients an hour for medication checks generates more revenue (for the system, if not for the patient) than one seeing a single patient for fifty minutes of therapy and medication management. In my training at Harvard's psychiatry residency at Beth Israel Deaconess Medical Center, I watched this tension daily: we were trained to do both, then released into a system that incentivized doing only one.
Is it better to see one doctor for medication and therapy?
The evidence says yes, and the reasons go beyond convenience. A landmark network meta-analysis by Cuijpers et al. in World Psychiatry found that combined psychotherapy and pharmacotherapy was significantly more effective than either treatment alone for depression.
But the case for integration is even stronger than the case for mere combination, because it addresses something the split-care research consistently identifies as a problem: the gap between what two providers know about the same patient.
Consider what a psychiatrist learns in a therapy session that a fifteen-minute medication check will never reveal. The patient mentions, almost in passing, that she has been sleeping on the couch because her marriage is unraveling. The patient describes irritability that sounds like a medication side effect but turns out to be the leading edge of something unprocessed from childhood. The patient is, for the first time, able to cry in session, and that timing matters for whether this is the right moment to adjust a dose or the wrong one.
When I am both the therapist and the prescriber, these details are not relayed secondhand (if they are relayed at all). They live in the same room, in the same relationship, observed by the same clinician. I don't need a phone call to learn what happened in therapy last week.
I was there.
What can go wrong in the split-care model?
Split care (the arrangement where a therapist provides psychotherapy and a separate psychiatrist manages medication) is not inherently bad, and I work hand-in-hand with some wonderful therapists where my role is more as a prescriber. There are situations where it makes sense (when a patient needs a type of therapy their psychiatrist doesn't offer, or when access to an integrated provider isn't available). I want to be honest about that.
But the split model introduces vulnerabilities that are structural, not personal. A study cited in The Lancet Psychiatry found that in roughly 30% of split-care arrangements, the two providers had never communicated at all. Priya's image of herself as "the telephone" is not an outlier.
It is, for many patients, the norm.
The risks are subtle but accrue on both sides of the aisle. A therapist notices increasing anxiety but attributes it to the therapeutic process. The prescriber, seeing the patient briefly, increases the SSRI dose. Neither realizes that the anxiety is a response to a workplace situation the patient hasn't fully disclosed to either of them (because neither appointment feels long enough or relationally deep enough to hold the full picture).
The patient gets more medication when what they needed was more understanding.
I have seen this pattern more times than I can count, and early in my career, I was complicit in it. Before I built my practice around integrated care, I spent time on the prescriber side of the split model. I kept my appointments brief. I told myself the therapist had the rest. And I missed things (important things, things I would have caught if I had been sitting with the patient for more than a quarter of an hour).
What does integrated care actually look like?
In my practice in Boston, integrated sessions are typically fifty to sixty minutes. We do therapy. We also discuss medication when it's relevant (not every session, but available in every session). The medication conversation is embedded in the therapeutic relationship rather than bolted onto it from the outside.
This matters because research on therapeutic alliance consistently shows it is one of the strongest predictors of treatment outcomes across all forms of therapy.
When the person prescribing your medication is also the person who knows your story (your marriage, your childhood, the way your jaw tightens when you talk about your father), the prescribing becomes more precise. The pharmacology changes because the understanding changes as well.
The word jeong surfaces again here. The kind of bond that makes psychiatric care work is built through accumulated presence: the session where nothing dramatic happens but something shifts, or the moment a patient trusts you enough to say the thing they've been circling for months.
You cannot build that in fifteen minutes four times a year. You cannot split it across two strangers and expect it to arrive intact.
If this resonates with you, I'd welcome a conversation about what integrated care might look like for your situation. You can learn more about what to expect from a first appointment or read about resources to assist your search for a psychiatrist who does integrated care.
Priya still comes on Tuesdays.
But now there is no second appointment, no quarterly check-in with a stranger, no telephone game. Some sessions we talk about what happens in her body during overnight call. Some sessions we talk about medication. Most sessions, the two are not as separate as the healthcare system pretends. Last week she mentioned, almost offhandedly, that she had cooked dinner for the first time in months (something simple, just dal and rice, the kind her mother used to make).
The office held the sentence for a moment, the way a room holds warmth after someone has opened and closed a door.
Frequently Asked Questions
Can a psychiatrist be your therapist?
Yes. Psychiatrists are trained in both psychotherapy and medication management during residency. However, many psychiatrists have moved away from providing therapy due to insurance and economic pressures. As of the mid-2010s, only about half of U.S. psychiatrists still provided any psychotherapy at all (and far fewer specialize in it), so you may need to look specifically for a psychiatrist who offers integrated care.
Is it better to see one doctor for medication and therapy?
For many patients, yes. Combined treatment (psychotherapy plus medication) is more effective than either alone for conditions like depression and anxiety. Seeing one clinician for both means your prescriber understands the full context of your life and therapy, which leads to more precise medication decisions and fewer things falling through the cracks. However, split care with strong communication between providers can also work well.
How long are appointments with a psychiatrist who does therapy?
Integrated sessions are typically fifty to sixty minutes, compared to the fifteen-minute medication checks common in split-care models. The longer format allows the psychiatrist to provide genuine psychotherapy while also addressing medication as needed within the same relationship.
How do I find a psychiatrist who also does therapy near Boston?
Look for psychiatrists who specifically describe their practice as offering psychotherapy or integrated care (not just "medication management"). In the greater Boston area (including Brookline, Cambridge, Jamaica Plain, Newton, and beyond), solo practitioners and small group practices are more likely to offer this model than large health system clinics. Ask during a consultation call whether the psychiatrist provides ongoing psychotherapy or only prescribes.
Patient details have been changed to protect privacy. Composites drawn from clinical experience are used throughout. The emotional and clinical truths are preserved; the identifying particulars are not. Prior results do not guarantee a particular outcome.
