Write better prompts to get the most out of Hathr AI
Explore how to build the best possible prompts to use Hathr AI’s HIPAA Compliant LLM to accelerate your workflows by >10x.
Access the full power of Hathr AI’s HIPAA Compliant LLM to do a week’s worth of work in minutes.
“The Hathr AI team was able to help us write new prompts to redact HIPAA related data, and speed up whole new workflows – [Hathr AI] turned a week of work into about 20 minutes.”
Tom H.
Mental Health Professional
Large Language Models (LLMs) like ChatGPT, Claude AI, etc. allow people to talk to computers with a regular conversation.
Before, to make computers do what you wanted, you had to be able to code.
Now you can ask a question of the LLM like you're talking to a coworker - aka via "Prompts".
A *prompt* is the foundation of effective AI interactions. It’s the instruction or query you give to Hathr AI to produce meaningful and accurate responses. Think of it as a bridge connecting your intent with the AI’s capabilities. Crafting a well-structured prompt ensures that Hathr AI, a HIPAA-Compliant AI tool, delivers precise outputs for complex tasks like summarizing medical notes, coding for insurance, or analyzing healthcare data.
Prompt Writing Best Practices
1. Be Specific: Clearly outline the task and avoid ambiguous language.
2. Provide Context: Give all necessary background details for accurate results.
3. Iterate and Refine: Test and tweak your prompts for optimal performance.
4. Use Formatting: For complex requests, organize prompts with lists or sections.
5. Ask the AI tool to write its own prompt: AI can help write its own prompts the best – give it the parameters you want, or an existing prompt to optimize, and Hathr AI will give you a prompt to do your work!
What is the anatomy of a prompt?
Every Prompt should have these components:
1. Instruction: The core task you want the AI to perform (e.g., “Summarize this medical document”).
2. Context: Any background information the AI needs to understand your request.
3. Examples (Optional): Provide samples to guide the AI’s response.
4. Output Format: Specify the desired structure (e.g., bullet points, a report).
Use these elements to get the most out of your LLM. It’s a little bit of work to get the most out of a HIPAA Compliant LLM, but it will get you so much more in return.
Prompts for HIPAA Compliant LLMs (HIPAA Compliant AI)
Summarize Patient Notes
Prompt:
“Summarize the key points from the following patient progress notes. Include diagnosis, treatment plans, and follow-ups. Provide the summary in bullet points.”
Use Case: Quickly review patient updates during consultations
Summarize a Document
Prompt:
“Here is a discharge summary. Summarize it for the medical team, focusing on the patient’s condition, treatment provided, and recommended follow-up steps. Deliver the summary in a structured format: Introduction, Findings, Recommendations.”
Use Case: Condense lengthy documents for efficient team communication.
Extracting Valuable Information
Prompt:
“Extract the following information from this insurance authorization request: patient details, requested procedure, insurance company requirements, and approval status. Return the details as a table.”
Use Case: Streamline insurance documentation workflows.
Draft Mental Health Reports
Prompt:
“Draft a mental health progress report for this client’s session. Include a summary of the session, progress on goals, and recommendations for the next steps.”
Use Case: Support therapists in efficient documentation.
Analyze Medical Notes and Coding for Insurance
Prompt:
“Analyze these medical notes to identify billable procedures and diagnoses. Translate the information into appropriate ICD-10 and CPT codes for insurance claims.”
Use Case: Simplify the coding process for accurate reimbursements.
Streamlining Therapy Notes
Prompt:
“Summarize therapy session notes for clinical use, highlighting key discussion points, therapeutic techniques used, and goals for the next session.”
Use Case: Save time while maintaining detailed, compliant records.
Preparing for Medical Authorizations
Prompt:
“Draft a letter for a pre-authorization request for a medical procedure. Include patient details, justification for the procedure, and references to supporting medical documentation.”
Use Case: Expedite the approval process for treatments.
Create Patient-Friendly Summaries**
Prompt:
“Translate the following medical report into a patient-friendly summary, avoiding jargon. Explain the condition, treatment options, and next steps in a conversational tone.”
Use Case: Enhance patient understanding and engagement.
Generating Patient Follow-Up Plans
Prompt:
“Based on the attached consultation notes, create a detailed patient follow-up plan. Include the recommended timeline, tests, and any lifestyle adjustments.”
Use Case: Deliver clear, actionable follow-up plans for patients.