Executive Summary: Transforming Surgical Documentation with Generative AI
Auna Salud, a premier private healthcare provider in Peru, faced significant inefficiencies due to manual post-surgical documentation, which consumed substantial surgeon time and impacted data accuracy. Partnering with DFX5, Auna Salud implemented a Generative AI-powered voice-enabled solution leveraging Amazon SageMaker, Bedrock, DynamoDB, and AWS Step Functions. The innovative system reduced surgeons’ documentation time per procedure by 95%, improved CPT and CIE-10 coding accuracy by 85%, and dramatically enhanced medical record completeness and operational efficiency.
About the Customer: Leading Healthcare Excellence in Peru
Auna Salud operates an extensive network of hospitals and specialty clinics throughout Peru, delivering high-quality healthcare services at scale. The enterprise manages thousands of medical procedures daily, emphasizing patient safety, accurate billing, and regulatory compliance through precise clinical documentation.
Customer Challenge: Addressing Inefficient and Error-Prone Surgical Documentation
Surgeons at Auna Salud traditionally spent 1–2 hours manually documenting detailed operative reports after each surgical procedure, recording essential details such as procedure steps, vital signs, complications, and clinical codes (CPT/CIE-10). This manual approach led to:
- Reduced Clinical Productivity: Surgeons diverted valuable clinical time towards administrative tasks, negatively impacting their focus on patient care and contributing to burnout.
- Delayed Documentation: Lengthy and delayed documentation hindered billing cycles, continuity of care, and regulatory compliance.
- Data Inconsistencies and Errors: Reliance on memory and manual coding increased inaccuracies and omissions, affecting record quality and billing accuracy.
- Coding Complexity: Manual assignment of CPT and CIE-10 codes led to variability, compromising billing precision and downstream analytics.
- Limited Analytical Insight: Narrative documentation lacked structured data required for quality improvements and operational analytics.
Without automation, Auna Salud risked continued inefficiencies, financial impacts from inaccurate billing, regulatory scrutiny, and compromised patient safety
Why AWS: Robust AI Infrastructure for Healthcare Innovation
Auna Salud selected AWS due to its powerful managed AI capabilities, scalable infrastructure, and deep integration potential into healthcare operations. AWS provided critical components such as Amazon SageMaker for customized medical transcription models, Amazon Bedrock for sophisticated natural-language understanding, AWS Step Functions and Lambda for seamless workflow orchestration, and secure, scalable data management with Amazon DynamoDB and Amazon S3. AWS’s proven track record in healthcare use cases and data security was essential to Auna Salud’s decision.
Why the Customer Chose DFX5: Expertise in AI-Powered Clinical Documentation
Auna Salud engaged DFX5 for their deep expertise in deploying specialized Generative AI-driven healthcare documentation solutions. DFX5’s proven track record leveraging AWS AI services to streamline clinical workflows, improve documentation accuracy, and automate complex medical coding processes uniquely positioned them to address Auna Salud’s needs. The partnership was driven by DFX5’s capability to rapidly implement innovative solutions tailored explicitly to healthcare documentation challenges.
Partner Solution: Real-Time, AI-Enhanced Surgical Documentation System
DFX5 designed a sophisticated Generative AI-powered voice-enabled documentation solution leveraging AWS’s managed AI and cloud services:
- Real-Time Dictation and Transcription: Surgeons dictate procedural details via Amazon Connect immediately after surgery. Audio recordings route to a specialized Whisper model hosted on Amazon SageMaker, fine-tuned to accurately transcribe complex medical terminology.
- Intelligent Extraction and Coding: Amazon Bedrock utilizes powerful Generative AI models to extract critical procedural parameters, identify complications, and automatically assign precise CPT and CIE-10 codes from the transcribed notes.
- Validation and Structured Storage: GenAI-driven validation rapidly flags missing or inconsistent data for human review. Final structured records, including coding metadata, are stored securely in Amazon DynamoDB, with audio and transcript archives maintained in Amazon S3.
- Automated Workflow Orchestration: AWS Step Functions and Lambda orchestrate the entire end-to-end process seamlessly, from dictation capture to structured documentation finalization, minimizing human intervention.
Results and Benefits: Dramatic Efficiency and Accuracy Gains
The AI-powered documentation system delivered transformative improvements to Auna Salud’s clinical documentation processes:
- Reduced surgeon documentation time per procedure by 95%, from 1–2 hours to approximately 5 minutes, freeing surgeons to focus on patient care and clinical duties.
- Achieved an 85% accuracy rate for AI-generated CPT and CIE-10 codes upon review, significantly reducing manual coding efforts and billing inconsistencies.
- Enabled real-time, accurate clinical documentation, accelerating billing cycles, enhancing regulatory compliance, and reducing financial risks associated with delayed or erroneous medical records.
- Generated structured data for advanced clinical analytics, enabling continuous quality improvement and richer operational insights.
These enhancements substantially improved clinical productivity, operational efficiency, billing accuracy, and overall healthcare service delivery.
Next Steps: Expanding Generative AI Integration in Healthcare Operations
Auna Salud plans to expand the Generative AI solution’s capabilities to additional documentation workflows, leveraging continuous improvement in model accuracy and workflow efficiency. Ongoing enhancements will further solidify their leadership position in healthcare innovation, operational excellence, and patient care quality.
About the Partner: AI-Powered Healthcare Innovation by DFX5
DFX5, an AWS Advanced Consulting Partner, specializes in delivering advanced Generative AI applications tailored for healthcare and enterprise-scale documentation needs. By leveraging AWS’s sophisticated AI services—including Amazon Bedrock, SageMaker, and robust cloud infrastructure—DFX5 empowers healthcare organizations to streamline clinical workflows, improve data accuracy, enhance regulatory compliance, and achieve transformative operational efficiencies and patient care outcomes.