Skip to main content
Back to Library
Prompt Engineering Guide

Mastering Medical report summary
on DeepSeek V3

Stop guessing. See how professional prompt engineering transforms DeepSeek V3's output for specific technical tasks.

The "Vibe" Prompt

"Summarize this medical report in a way that's easy to understand for a patient. Make sure to capture all the important details."
Low specificity, inconsistent output

Optimized Version

STABLE
You are a highly skilled medical AI assistant tasked with summarizing complex medical reports for patients. Your goal is to provide a clear, concise, and easy-to-understand summary, ensuring all critical information is retained, while omitting technical jargon and irrelevant details. Here's the medical report: [MEDICAL_REPORT_TEXT] Follow these steps to generate the summary: 1. **Identify the Patient's Chief Complaint/Reason for Visit:** Extract the primary reason the patient sought medical attention. 2. **List Key Diagnoses:** Identify all confirmed diagnoses mentioned in the report. 3. **Summarize Significant Findings:** Extract objective findings from physical examinations, lab results, and imaging studies that are crucial for understanding the patient's condition. 4. **Detail Treatment Plan:** Outline the recommended treatments, medications, procedures, and lifestyle changes. 5. **Explain Prognosis/Next Steps:** Describe what the patient can expect (e.g., recovery time, follow-up appointments, potential complications). 6. **Consolidate and Simplify:** Combine the extracted information into a coherent, patient-friendly narrative. Use simple language. Avoid medical acronyms or explain them clearly if unavoidable. Focus on answering the patient's likely questions (e.g., 'What do I have?', 'What should I do?', 'What can I expect?'). 7. **Review for Clarity and Completeness:** Ensure the summary is accurate, covers all essential aspects, and is easily digestible by a non-medical professional. Provide the summarized report in the following format: **Patient Summary:** - **Reason for Visit:** [Chief Complaint] - **Main Diagnoses:** [List of Diagnoses] - **Important Findings:** [Key Findings] - **Treatment Plan:** [Details of Treatment] - **What to Expect:** [Prognosis/Next Steps]
Structured, task-focused, reduced hallucinations

Engineering Rationale

The optimized prompt leverages several best practices for instructing large language models. It starts with a clear 'System Persona' ('You are a highly skilled medical AI assistant...'). It then explicitly defines the 'Goal' and 'Constraints' (clear, concise, easy-to-understand, no jargon, critical info retained). The most significant improvement comes from the 'Chain of Thought' steps (1-7), which guide the model through a logical reasoning process, breaking down the complex task into manageable sub-tasks. This significantly improves the quality and structure of the output. Finally, it specifies the 'Output Format,' ensuring consistency and ease of parsing for downstream applications or direct patient consumption. This structured approach forces the model to think sequentially and extract specific information, rather than just generating a free-form summary, leading to more accurate and relevant results. The explicit 'Review' step further prompts the model for self-correction.

0%
Token Efficiency Gain
The summary correctly identifies the patient's chief complaint.
All primary diagnoses are accurately listed in the summary.
The summary omits technical medical jargon or explains it clearly.

Ready to stop burning tokens?

Join 5,000+ developers using Prompt Optimizer to slash costs and boost LLM reliability.

Optimize My Prompts