History and Physical (H&P) Template: What AI Fills In

A history and physical (H&P) note is the full workup you write on admission or a new patient. It runs in a fixed order: chief complaint, HPI, past medical and surgical history, medications, allergies, family and social history, review of systems, physical exam, then a problem-based assessment and plan. This page gives you that template section by section, then shows exactly what an AI scribe drafts for each part and what stays your job.

The template below follows the University of Central Florida College of Medicine’s comprehensive adult H&P write-up guide, which is about as close to a canonical teaching version as exists. Copy it. Then read the second half, because the interesting question in 2026 isn’t what an H&P contains. It’s which lines a machine can fill and which ones it can’t touch.

Key takeaways

• A comprehensive H&P has 12 sections, from chief complaint through assessment and plan (UCF College of Medicine H&P guide).
• The HPI is built with OLDCARTS: Onset, Location, Duration, Characterization, Aggravating/alleviating, Radiation, Temporal, Severity (StatPearls, 2023).
• An AI scribe can draft the history and the spoken exam. It cannot do your differential, order the workup, or sign.
• Primary-care physicians log a median of 36.2 minutes of EHR time per 30-minute visit (JAMA Network Open, 2023). The H&P is a big part of that.

What sections does a comprehensive H&P template have?

Here’s the full template, in the order you write it. Each section is one job. The UCF College of Medicine H&P guide is the reference for the wording and sequence.

Chief complaint (CC). The primary symptom that brought the patient in, ideally in their own words. One line.

Source and reliability. If the history came from someone other than the patient, say who. If the patient isn’t reliable, say why (“somnolent,” “intoxicated”). Skip when it’s the patient and they’re clear.

History of present illness (HPI). The story of the complaint, in order. Open with age, sex, and reason for the visit, plus pertinent past history. Then walk the dimensions of the complaint. The clean way to do this is OLDCARTS, from the StatPearls SOAP Notes reference:

• Onset: when did it start?
• Location: where is it?
• Duration: how long, and is it constant or intermittent?
• Characterization: how does the patient describe it?
• Aggravating and alleviating factors: what makes it worse or better?
• Radiation: does it move?
• Temporal: worse at a certain time of day?
• Severity: rate it 1 to 10.

Past medical and surgical history (PMH/PSH). Conditions with dates and complications, operations with indications, OB/GYN and psychiatric history where relevant, health maintenance, and significant childhood illness.

Medications. Every drug with dose, route, frequency, and generic name. Include over-the-counter meds and supplements. The guide is blunt here: do not use abbreviations.

Allergies. Not just the agent, the reaction. “Penicillin: hives” tells the next clinician more than “PCN allergy.”

Family history. Health or cause of death of parents, siblings, children, and the diseases that run in the family (HTN, CAD, CVA, DM, cancer).

Social history. Occupation, education, home situation, and quantified tobacco, alcohol, and drug use. Advance directives, functional status, and safety concerns live here too.

Review of systems (ROS). A system-by-system checklist of yes/no symptom questions that catches complaints the patient never volunteered. If something’s already in the HPI, note “refer to HPI” rather than repeating it.

Physical exam. Always start with vitals: temperature, pulse, blood pressure, respiratory rate, a pain rating, and pulse oximetry when you have it. Then general appearance, then a top-down system exam (HEENT, neck, heart, lungs, abdomen, extremities, neuro, MSK, skin).

Labs and imaging. The data relevant to the presentation, with your interpretation, not a copy-paste of the report.

Problem list, then assessment and plan. List problems, most important first. Then, per problem, a differential of at least three diagnoses ordered most-likely or most-serious first, followed by a diagnostic plan, treatment plan, and patient education.

What does an AI scribe fill in for each H&P section?

This is the table worth bookmarking. An ambient AI scribe listens to the visit and drafts a structured note. But an H&P is part transcription and part judgment, and the split is not even. Some sections a scribe drafts almost completely from the conversation. Others it can only stub, because the content lives in your head or your EHR, not in the room.

Here’s each section, what the AI can realistically draft, and what you still own.

H&P section	What an AI scribe can draft	What the clinician still owns
Chief complaint	The patient’s stated reason, captured in their words	Deciding what the real complaint is when the patient buries it
HPI	The narrative, structured by OLDCARTS, from the spoken history	Pulling the thread that matters, and the pertinent negatives you chose to ask
PMH / PSH	Whatever the patient recites in the room	Reconciling it against the chart; nothing pre-fills from your EHR
Medications	The med list as the patient states it	Reconciliation against pharmacy and chart, doses, and what they actually take
Allergies	Agent and reaction, as reported	Confirming and clarifying vague reports (“upset stomach” vs true allergy)
Family history	Named conditions and relatives from the conversation	Judging relevance to the current presentation
Social history	Occupation, habits, and quantities the patient shares	Screening you did that the patient didn’t narrate
Review of systems	The positives and negatives you voice during the visit	The completeness of the screen; the AI drafts only what was said
Physical exam	Findings you speak aloud as you examine	Every finding you didn’t verbalize, and the exam itself
Labs / imaging	Results you dictate or read out	Your interpretation and what to order next
Assessment / plan	A first-pass structure if you talk through your reasoning	The differential, the diagnosis, the workup, the treatment. All of it.

Read that last row twice. The assessment is where an H&P earns its keep, and it’s the section a scribe can least stand in for. A transcript can hold your spoken reasoning, but the reasoning is yours. Everything above the line is capture; the plan is medicine.

One more thing the table makes concrete: an AI scribe drafts from what happened in the room, not from your chart. It doesn’t pull a med list from the EHR or a problem list from prior notes, so medication reconciliation and history verification stay hands-on. Anyone implying otherwise is describing an EHR integration, not an ambient scribe.

How does an AI scribe hand back the H&P sections?

You start a session, get consent, and see the patient. The tool captures the conversation ambiently while you work. When the visit ends, it drafts the note, sorts the history into the right sections, and hands you something to review instead of a blank template at 9pm.

The AI Medical Scribe by Patient Square is the ambient scribe module inside Practice Copilot: it listens during the visit and hands back a structured SOAP note, ICD-10 suggestions, and a prescription draft, ready to review and sign minutes after the visit. For an H&P, that structured note is the scaffold: the history sections populated from the conversation, the exam findings you spoke, and a starting assessment structure. The ICD-10 codes are suggestions you confirm, never a coding engine, and never an E&M level. The prescription is a draft, too. The scribe doesn’t screen it for interactions or dosing; you review it and sign it, the same clinical check you make on every script now.

What it does not do: pre-fill from your EHR, write back into it, generate a per-specialty H&P template, or produce a patient after-visit summary. The notes always come out in clean clinical English, and the visit audio is processed in memory and discarded once the note is drafted, so there’s no recording sitting anywhere. The full data posture is on our security page. If you’re a hospitalist weighing this for admission workups, the hospitalist walkthrough covers the ward-round workflow in more detail.

Where does the AI-drafted H&P still need you most?

Three places, every time.

Medication reconciliation. The scribe writes down what the patient says. It has no line into the pharmacy or the chart, so the drafted med list is a claim to verify, not a reconciled list. This is a patient-safety step, not a formatting one, and it stays manual.

The physical exam. A scribe can only record findings you verbalize. If you examine in silence, the exam section comes back thin. The fix is habit, not software: speak your findings as you examine, the way you would to a student. What you don’t say aloud, the AI can’t write.

The assessment and plan. Drafts contain errors. Models mishear a drug name, compress two problems into one, or write something plausible that didn’t happen. The assessment is exactly where a plausible-but-wrong line does the most damage, because it’s the reasoning the next clinician trusts. Read it hardest. For the shorter daily notes that follow the admission H&P, our progress note examples show the same review discipline on the SOAP format, and what accuracy really means for an AI scribe is worth reading before you trust any vendor’s demo.

The chart is yours, legally, no matter who typed the draft. A tool that fills the history and the spoken exam at the pace of the visit is the difference between signing the H&P before you leave the floor and finishing it in bed. That’s the honest pitch: it drafts the parts that are transcription, and hands you back the time to do the parts that are judgment.

If you want to see it run on a real admission, book a demo and bring your own workflow. Run it on a live clinic day during the trial and read every section closely for the first week. That’s the evaluation that tells you the truth.

Frequently asked questions about H&P templates and AI

The FAQ block below is the short version. The template above is the one to copy.