AI Pre-Visit Interviews in Primary Care: Benefits, Barriers, and Best Use Cases
Dr. Daniel Ngui
Primary care visits are often short, complex, and unpredictable. In a single appointment, a clinician may need to identify the patient’s main concern, review symptoms, check medication use, address preventive care gaps, and still leave time for assessment, decision-making, and documentation. That makes pre-visit preparation especially valuable.
One emerging idea is the use of AI-enabled patient interviews before the visit. In this model, an AI conversational agent uses voice or chat to collect structured clinical information before the appointment and generate a summary for the clinician. The goal is not to replace clinician history taking. The goal is to improve pre-visit preparation, support better workflow, and help the clinician start the visit with clearer context.
Whether this works in primary care depends less on the technology itself and more on how well it fits real clinical workflow.
What AI pre-visit interviews are
AI pre-visit interviews use a conversational agent to interact with the patient before the appointment. This interaction may take place through a smartphone app, web portal, phone call, or kiosk in the waiting room.
The AI asks structured clinical questions and produces a pre-visit summary for the clinician. Depending on the design, this summary may include:
Chief concern
Timeline of symptoms
Medication adherence
Review of systems
Screening questionnaire results
Risk factors
Social determinants
Preventive care gaps
In the best-case version of this workflow, the output becomes a concise pre-visit summary that can be reviewed before the encounter and integrated into the EMR.
Why this is appealing in primary care
Primary care visits are often limited to 10 to 15 minutes. In that setting, even a small amount of preparation can change the quality of the visit.
A structured pre-visit interview could help by collecting history in advance, organizing the visit agenda, highlighting urgent issues, and reducing some of the documentation burden tied to routine intake. Instead of starting the visit from zero, the clinician could enter the room already oriented to the patient’s main concern and the most relevant context.
A simple pre-visit summary might look like this:
Chief concern: persistent cough for 3 weeks
Associated symptoms: fatigue, mild shortness of breath
No fever
Smoking history: 20 pack years
Medications: unchanged
Preventive gaps: due for pneumococcal vaccine
A summary like this does not replace the clinical encounter, but it can help the clinician start with a clearer frame.
Potential benefits for clinicians
Time efficiency
AI pre-visit interviews may help collect part of the history before the appointment. In primary care, that is valuable because time pressure affects almost every visit. If key information is collected in advance and presented clearly, the clinician may spend less time on routine intake and more time on decision-making and patient discussion.
Improved pre-visit planning
Pre-visit interviews may also support better planning before the patient arrives. For example, the system could identify medication issues, overdue labs, chronic disease monitoring needs, or preventive screening gaps.
In a diabetes visit, a pre-visit summary might flag that the last A1C was eight months ago, the patient is not on a statin, and the foot exam is overdue. This gives the care team useful information before the encounter starts.
Better data capture
Patients sometimes forget details during the visit or remember them only after the conversation has moved on. A structured interview can ask follow-up questions, clarify timelines, and capture symptom progression in a more systematic way.
More structured data for the EMR
Instead of producing only free text, AI pre-visit interviews could generate structured fields such as symptom duration, severity scale, risk factors, or screening scores. Structured input may be more useful for downstream documentation and clinical decision support than an unstructured patient narrative alone.
Major barriers to adoption
The benefits are easy to imagine. Adoption is harder.
Patient trust
Patients may worry about who sees the information, whether AI is replacing the doctor, and whether the conversation is being recorded. This means trust cannot be treated as an afterthought.
Clear messaging is essential. Patients need to understand that their responses are confidential, that the information is being collected to help prepare for the visit, and that the summary is shared with their healthcare team.
Technology barriers
Primary care serves diverse patient populations, including elderly patients, people with limited digital literacy, patients with language barriers, and patients who may not have smartphones or stable internet access.
That means any successful model must account for voice interaction, simple interfaces, multilingual support, and optional rather than mandatory use.
Privacy and confidentiality
Questions about privacy are central. Clinics will need to consider where the data is stored, whether the system is cloud-based, whether the workflow is compliant with Canadian privacy requirements, and whether data is encrypted.
A secure design may involve patient authentication through a clinic portal and a pre-visit conversation that takes place within a secure system without using unnecessary identifiers.
Accuracy and reliability
AI may misinterpret symptoms, miss red flags, or generate incomplete or incorrect summaries. For that reason, AI pre-visit interviews cannot replace clinician history taking. They need to remain assistive, not diagnostic.
Will this work in primary care?
This is the key question.
Some specialist settings may adopt this kind of workflow sooner because their intake processes are longer and more predictable. Gastroenterology intake, sleep clinic questionnaires, mental health screening, and surgical pre-assessment are all easier to standardize.
Primary care is more difficult because of high patient volume, unpredictable visit reasons, short appointments, and variable patient preparation.
That said, some primary care use cases are more promising than others.
Best use cases in primary care
Chronic disease monitoring
This may be one of the strongest use cases. Before a diabetes visit, an AI pre-visit interview could ask about home glucose trends, medication adherence, symptoms of hypoglycemia, and lifestyle factors. That gives the clinician a more prepared starting point for the encounter.
Preventive care visits
Preventive visits often benefit from structured pre-visit information. AI could help collect family history, cancer screening eligibility, vaccination history, and lifestyle risks before the appointment begins.
Medication reviews
Medication review is another practical use case. AI can ask about side effects, adherence, and dosing confusion before the visit, making it easier to focus the discussion when the clinician and patient meet.
Mental health screening
AI pre-visit interviews may also be useful for administering validated tools such as PHQ-9, GAD-7, or sleep assessments before the visit. In this use case, the value comes from structured input and preparation rather than free-form conversation alone.
A realistic workflow example
A practical pre-visit workflow might look like this:
The patient books an appointment online
The AI sends a link asking the patient to complete a short health interview before the visit
The patient speaks with the AI for five to seven minutes
The AI produces a structured summary
The clinician reviews the summary before the appointment
The clinician confirms key points during the visit
In this model, the AI is not replacing the encounter. It is functioning as a preparation tool.
The real issue is workflow integration
The central question is not whether AI can ask patients questions. It can.
The real question is whether this process fits primary care workflow well enough to be adopted.
Primary care clinicians are unlikely to adopt AI pre-visit interviews unless the tool saves time, reduces documentation burden, and improves visit quality. If it adds complexity, creates long summaries, or introduces another disconnected system, adoption will be limited.
Critical design principles for success
For AI pre-visit interviews to work in primary care, the design likely needs to follow a few clear principles:
Optional for patients
Very short interaction, ideally five minutes or less
Clear privacy protections
Integrated into the EMR or clinic workflow
Summary must be concise
Clinician verification required
Multilingual voice interface
These are not minor details. They determine whether the tool supports care or adds friction.
A more realistic strategic model: AI-assisted team-based care
One of the most practical opportunities may be AI-assisted team-based care.
In this model, the AI interview produces a summary, a nurse or MOA reviews it, and the clinician sees the patient with that preparation already in place. That makes the AI less of a direct substitute for intake and more of a preparation tool for the care team.
This model is more realistic in primary care because it aligns with team-based workflows rather than trying to replace clinician judgment or patient interaction.
One important risk
A major risk is that patients may over-disclose information, leading to long, unfiltered summaries that clinicians do not want to read.
That means the output cannot be a transcript dump. It needs to be a structured summary with prioritized concerns and, where appropriate, red flag alerts. If the output is too long or too noisy, clinicians are likely to ignore it.
Final Thoughts
AI pre-visit interviews can be useful in primary care if they function as a preparation tool rather than a replacement for the clinical encounter.
The strongest use cases are likely to be short, structured, optional, privacy-safe, and integrated into real workflow. Chronic disease monitoring, preventive care, medication review, and mental health screening are more promising than an open-ended attempt to automate the full primary care history.
The real opportunity is not simply AI asking questions before the visit. It is whether that process helps the clinician start the encounter better prepared without adding complexity to the day.
FAQ
What is an AI pre-visit interview?
An AI pre-visit interview is a voice or chat based interaction that collects structured patient information before the appointment and produces a summary for the clinician.
Can AI pre-visit interviews replace clinician history taking?
No. They should function as an assistive preparation tool, not a replacement for clinician history taking or clinical judgment.
What are the best use cases in primary care?
The strongest use cases are chronic disease monitoring, preventive care visits, medication reviews, and structured mental health screening.
What are the biggest barriers to adoption?
The main barriers are patient trust, technology access, digital literacy, privacy concerns, reliability, and poor workflow integration.
What makes this more likely to work in primary care?
Short interactions, concise summaries, optional use, multilingual support, privacy protections, EMR integration, and clinician verification are all important for adoption.
See how Empathia supports review, documentation alignment, and billing-aware workflow in real-world primary care. Book a demo to explore the workflow in practice.
More resources:
