How to Intermediate · 4 min read

Clinical note summarization with AI

Q: Clinical note summarization with AI

Use a large language model like gpt-4o-mini to summarize clinical notes by sending the note text as a prompt and requesting a concise summary. This can be done via the OpenAI Python SDK with prompt engineering to ensure clinical relevance and privacy compliance.

Quick answer

Use a large language model like gpt-4o-mini to summarize clinical notes by sending the note text as a prompt and requesting a concise summary. This can be done via the OpenAI Python SDK with prompt engineering to ensure clinical relevance and privacy compliance.

PREREQUISITES

Python 3.8+
OpenAI API key (free tier works)
pip install openai>=1.0

Setup

Install the openai Python package and set your API key as an environment variable for secure access.

bash

pip install openai>=1.0

output

Collecting openai
  Downloading openai-1.x.x-py3-none-any.whl (xx kB)
Installing collected packages: openai
Successfully installed openai-1.x.x

Step by step

This example shows how to send a clinical note to the gpt-4o-mini model and receive a concise summary. The prompt instructs the model to focus on key clinical information.

python

import os
from openai import OpenAI

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

clinical_note = """
Patient is a 65-year-old male with a history of hypertension and diabetes mellitus type 2. He presents with chest pain and shortness of breath. ECG shows ST elevation in leads II, III, and aVF. Troponin levels are elevated. Treatment started with aspirin and nitroglycerin.
"""

prompt = f"Summarize the following clinical note focusing on diagnosis, treatment, and key findings:\n\n{clinical_note}\n\nSummary:" 

response = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": prompt}],
    max_tokens=150
)

summary = response.choices[0].message.content
print("Clinical note summary:\n", summary)

output

Clinical note summary:
Patient, a 65-year-old male with hypertension and type 2 diabetes, presents with chest pain and shortness of breath. ECG indicates ST elevation in leads II, III, and aVF, and elevated troponin confirms myocardial infarction. Treatment initiated with aspirin and nitroglycerin.

Common variations

You can use asynchronous calls for better performance in web apps, switch to streaming responses for real-time display, or try other models like claude-3-5-haiku-20241022 for different summarization styles.

python

import os
import asyncio
from openai import OpenAI

async def async_summarize(note: str):
    client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
    prompt = f"Summarize the clinical note:\n\n{note}\n\nSummary:"
    response = await client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=150
    )
    return response.choices[0].message.content

clinical_note = "Patient with fever and cough, diagnosed with pneumonia. Started antibiotics."

summary = asyncio.run(async_summarize(clinical_note))
print("Async summary:\n", summary)

output

Async summary:
Patient diagnosed with pneumonia presenting with fever and cough; treatment initiated with antibiotics.

Troubleshooting

If you receive incomplete summaries, increase max_tokens or refine your prompt for clarity.
Ensure your clinical data is de-identified to comply with HIPAA and privacy regulations.
If API calls fail, verify your OPENAI_API_KEY environment variable is set correctly and your network allows outbound HTTPS requests.

✅

Key Takeaways

Use gpt-4o-mini with prompt engineering to extract concise clinical summaries.
Always de-identify patient data before sending to AI APIs to maintain privacy compliance.
Async and streaming API calls improve responsiveness in clinical applications.
Adjust max_tokens and prompt specificity to optimize summary quality.

Verified 2026-04 · gpt-4o-mini, claude-3-5-haiku-20241022

Verify ↗