The default outpatient psychiatry intake template in Nextvisit is the one I use in my own practice at Psychvisit. It is also the template that ships first to every new psychiatrist, NP, and PA who joins the platform on Provider+. I want to write about how it ended up in its current shape, because the structure was not an accident, and the reasoning behind each section is the kind of thing that often gets lost between a clinician’s training and a software product’s release notes.
This is a founder’s note, so I will say upfront: I am the practicing psychiatrist co-founder. I still see patients. I still use this template every day. The template has gone through six revisions in three years. The version that ships today is the version I trust to read like an actual psychiatric intake, not a primary-care SOAP note retrofitted to fit one.
What an intake actually has to do
The job of an outpatient psychiatry intake is to do four things in 60 minutes: build rapport, gather the clinical material that informs the differential, assess and document risk, and arrive at a treatment plan the patient understands and can consent to. The chart that comes out of that visit has to support all four, plus survive a payer audit, plus inform the next clinician (or the next version of you) two years from now.
Most general-purpose AI scribe templates collapse this into a four-paragraph summary. The summary reads cleanly. It loses about 40 percent of the information that matters for psychiatry. The structure is the information. We rebuilt the template around that idea.
The sections, and why each is the way it is
The current default has nine top-level sections. Each one is its own field, not a paragraph in a larger blob. This matters because (1) the field structure is what payers and reviewers read, and (2) the field structure is what makes the chart queryable later.
History of presenting illness. The HPI in psychiatry needs to capture onset, course, severity, modifying factors, associated symptoms, and the patient’s own framing of the problem. Aria captures the first five into structured sub-fields and preserves the patient’s framing as a direct quote in the chief complaint. Reviewers can see the timeline at a glance. The patient’s voice is on the page.
Past psychiatric history. Hospitalizations, suicide attempts and self-harm history, prior medication trials with response and tolerability, prior therapy, prior diagnoses. Medication trials are a structured table; this is a conscious choice. Reading “patient previously tried sertraline, fluoxetine, and bupropion” is less useful than seeing each agent with the dose, duration, response, and reason for discontinuation. The structured table makes the pattern legible.
Family psychiatric history. Mood, anxiety, psychosis, substance use, completed suicide. These get captured per first-degree relative when the patient discusses them. The model does not invent family-history items the patient did not mention.
Social and developmental history. Trauma history is captured with care here. Aria treats trauma disclosure differently from other social-history elements: it captures the disclosure, but it does not paraphrase it. The patient’s words go in, with appropriate field labels. This was a deliberate choice and a request from clinicians during early testing. Trauma disclosure is too easily flattened.
Substance use. Current and past, by substance, with frequency and quantity. This is its own section because it is a distinct review of systems, not a subsection of social history. The template makes it visible.
Mental status exam. Appearance, behavior, speech, mood, affect, thought process, thought content, perception, cognition, insight, judgment. Aria captures observed elements during the visit. The clinician confirms or corrects. The MSE renders in the field that says MSE. It does not paraphrase into the HPI.
Risk assessment. This is the section that gets the most editorial care. Risk factors, protective factors, current ideation, plan, intent, means, and the clinical judgment that follows. Aria treats this as a structured field and does not flatten it. If a risk factor is mentioned anywhere in the visit (a recent fight with a partner, access to a stockpiled medication, a specific date as a goal), it surfaces in this section and prompts the clinician to address it. Risk assessment is the section where reconstruction at 9 PM does the most damage. It is also the section where structured capture during the visit pays off most.
Differential diagnosis. The differential is a list with reasoning, not a single line. Two or three plausible diagnoses with the evidence for each, and the clinician’s chosen working diagnosis with the evidence that distinguishes it. This is what training programs teach and what payers expect. Most general-purpose templates skip this and go straight to a diagnosis.
Treatment plan. Medication choice with rationale, dosing, monitoring plan, anticipated side effects discussed, and follow-up window. Therapy referrals if indicated. Lab work or screening if indicated. Consent and shared decision making documented. Each of these is its own structured element.
What we learned watching it run
Three observations from watching this template run across more than 300 practices.
Clinicians who came from a general-purpose scribe edit fewer characters per intake on the Nextvisit template than they did on their prior tool. The reduction is roughly 35 to 45 percent. The reason, when we look at the edits, is structural. They are not rewriting the structure of the note anymore. They are correcting clinical content where the model misheard or omitted, and the structural work is gone.
The risk assessment section is the most-edited section in the first week of use. It drops to roughly average by week three, as Aria adapts to the clinician’s preferred phrasing. The structure of the section does not change. The phrasing does.
Comorbidities surface more often when the template explicitly asks for them. Practices that switched to this template caught more documented comorbid conditions in the first 30 days than they had in the prior 30 days, not because the patients changed, but because the template prompts for them and Aria captures them when they are mentioned. This shows up in the coding sidebar as more accurate ICD-10 suggestions.
What clinicians can change
Everything. The template is a default, not a constraint. Every section can be reordered, renamed, removed, or added to. Custom prompts can be built from scratch. The default exists because most prescribers want to start with something that works on day one, not build a template from a blank page. After that, the template is yours.
If you want a copy of the source for the default outpatient psychiatry intake template to use as a starting point, it lives in the Templates app under “Outpatient Psychiatry Intake” once you sign in. We update it when the field updates. The version in your workspace becomes yours the moment you change it.
The right intake template is the one that supports the four jobs of the visit and survives the next two years of chart review. This is the version I trust to do that. It is also, not coincidentally, the version I use for my own patients.