AWS for Industries

How Contrast AI Is automating physician documentation with AWS HealthScribe

In today’s healthcare landscape, the burden of administrative tasks, particularly electronic health record (EHR) documentation, consumes an average of 4.5 hours per day for physicians. This time commitment detracts from patient care and is the number one driver of physician burnout—an issue affecting 48% of healthcare workers. It is leading to a projected shortage of up to 86,000 physicians by 2036.

Contrast AI, guided by Dr. Kai Skallerud (the founder and CEO), are leveraging Amazon Web Services (AWS) HealthScribe to reduce documentation time—saving doctors 10 hours a week. It allows them to devote more attention to patient care, which enhances health outcomes and improves both clinician and patient satisfaction.

Opportunity

Dr. Skallerud’s journey to becoming a doctor began with a simple desire: to help people. However, he quickly realized there were significant obstacles hindering the connection between doctors and their patients. One of the most formidable challenges is the overwhelming burden of charting.

Recognizing the need for change, he founded Contrast AI to support doctors by alleviating the administrative burdens that detract from their mission. “When I realized that the excessive burden of medical charting and administrative tasks was the main contributor to staff burnout and a major source of healthcare spending, I knew I had to do something about it,” shared Dr. Skallerud.

Contrast AI’s mission is to transform the clinical documentation process by minimizing the administrative burdens that healthcare providers face. Dr. Skallerud recognized the inefficiencies of traditional EHR systems, where manual data entry and administrative tasks consume valuable time. Therefore, Contrast AI set out to create a solution that allows physicians to focus more on patient care rather than paperwork.

One key example of their automation efforts is the generation of SOAP (subjective, objective, assessment, plan) notes. Typically, physicians would have to manually compose these detailed medical notes based on a patient conversation. However, by leveraging the output from AWS HealthScribe, a fully managed service powered by Amazon Bedrock, Contrast AI is able to automatically transform the natural conversation into a well-documented SOAP note. It can be directly sent to the medical record after being reviewed and approved by the physician.

This automation reduces the time physicians have to spend at the keyboard composing notes, allowing them to focus more on patient care. By automating repetitive tasks, minimizing the number of mouse clicks and reducing data entry actions required, Contrast AI helps alleviate the daily administrative load that often leads to physician burnout.

However, developing the capabilities to accurately capture, transcribe, and summarize ambient clinical conversations into draft documentation presents considerable challenges. Enabling AI-powered ambient listening, transcription, and natural language summarization requires extensive resources. These include access to AI experts, vast amounts of annotated healthcare data, and significant computational power. To produce accurate and trustworthy clinical documentation that physicians can rely on, Contrast AI needed a scalable, robust, and cost-effective AI solution.

Contrast AI evaluated several options to accommodate their needs for advanced clinical documentation capabilities. Ultimately, the engineering team chose to work with AWS HealthScribe due to its robust set of features tailored for healthcare. The innovative features clearly show a deep understanding of healthcare workflows, and offer the scalability and cost-effectiveness required to build an enterprise-grade clinical documentation solution.

Becca McHenry, VP of Engineering at Contrast AI, helped design the solution, define the requirements, and assess the different providers. “We chose AWS HealthScribe because it helped us get to a complete product quickly, delivered above market results, and was easy to get started with,” shared McHenry.

The integration of AWS HealthScribe was completed in just a few days. “The ease of use and scalability of AWS offerings have been crucial for us,” says McHenry. “We were able to get our solution up and running quickly, and the ability to easily adjust our compute and storage resources as needed has been invaluable.” This quick setup allowed Contrast AI to focus on refining and continuously improving the user experience, making sure that the platform remains intuitive and efficient for healthcare providers.

By integrating AWS HealthScribe APIs, Contrast AI was able to include advanced conversational and generative AI capabilities within ContrastHub. ContrastHub is a healthcare charting platform designed by Contrast AI. Through the integrated generative AI capabilities of AWS HealthScribe there is no need to manage the underlying machine learning infrastructure or train their own healthcare-specific large language models (LLMs). This accelerated the development process by months and allowed the team to focus on refining the user experience for healthcare providers.

In addition to clinical documentation and evidence linking, AWS HealthScribe provides conversation transcripts, identifies speaker roles, classifies dialogues based on their relevance, and extracts medical terms. These capabilities significantly enhance the documentation review process for doctors, allowing them to efficiently validate and finalize patient records with greater accuracy.

Medical record displaying two sections side by side. The left section, titled "Chief complaint" and "History of present illness," provides detailed information about a patient's follow-up for heart failure. It includes the patient's symptoms, recent lifestyle changes, and medical history. The right-side section, labeled "Transcript," shows a partial conversation between a doctor and patient discussing the patient's family medical history, particularly focusing on the patient's parents. The bottom of the left-side section lists a "Review of systems" with bullet points of current symptoms. The image provides a comprehensive view of a medical consultation, combining written records with a transcribed dialogue.Fig 1: Contrast AI’s ContrastHub platform showcasing the use of AWS HealthScribe on a fictional conversation between a Doctor and a Patient.

Solution

In the clinical setting, after a doctor obtains consent from the patient, the medical visit is recorded using the ContrastHub platform, either through the Contrast AI mobile app or web client. These recordings are securely uploaded to Amazon Simple Storage Service (Amazon S3), where they are processed using AWS HealthScribe.

With an API call, AWS HealthScribe converts the audio into structured clinical documentation summaries and transcripts. They are specifically formatted in the standardized history and physical (H&P) note template. The ContrastHub platform then delivers these insights directly to healthcare providers, enabling them to efficiently review and integrate the information into their clinical workflow without manual data entry.

One of the critical aspects of AI in healthcare is the need for explainability and traceability in clinical documentation. AWS HealthScribe supports this need by providing detailed metadata that links the AI-generated clinical document back to the original dialogue in the transcript. This evidence linking is essential for maintaining the integrity and accuracy of medical records.

“This traceability is crucial for our healthcare customers, as it ensures the integrity and accuracy of the medical documentation,” McHenry notes. The ability to trace the origin of each piece of information helps healthcare providers trust the AI-generated notes, knowing they can always refer back to the original patient-doctor conversation if needed.

Contrast AI also leverages Amazon Comprehend Medical to enhance clinical documentation by extracting key medical entities, such as diagnoses and medications. “Amazon Comprehend Medical has been instrumental in helping us turn the raw text of the notes into discrete, actionable data,” McHenry adds. This capability allows healthcare providers to quickly update patient records in the EHR system, streamlining administrative tasks and reducing the potential for errors.

Diagram illustrating Contrast AI's application architecture on AWS Cloud. The image shows a workflow from users recording conversations to processed notes being sent to an Electronic Health Record (EHR) system. Key components include: 1.	Contrast AI Application using Amazon Route 53 and ECS 2.	Storage Layer with Amazon S3 for audio files and Amazon RDS for data 3.	AI/ML Workflow featuring AWS HealthScribe and Amazon Comprehend Medical 4.	Users interacting through mobile app and web client 5.	Data flow arrows showing the path of information The right-side lists four numbered explanations of AWS services used, including S3, HealthScribe, Comprehend Medical, and RDS. The diagram demonstrates how conversation recordings are processed, analyzed, and stored using various AWS services to generate structured medical notes and extract relevant medical information.

Fig 2: Contrast AI’s Cloud Architecture on AWS, leveraging AWS HealthScribe for transcription and AI Summary, as well as Amazon RDS for data storage.

Looking ahead, Contrast AI is exploring additional ways to leverage AWS generative artificial intelligence (AI) and machine learning services. For example, by integrating large language models to enhance the interactive capabilities of their solution. “As we continue to evolve our product, we know we can rely on the expertise and support of the AWS team to help us identify and implement the best technologies to meet the needs of our healthcare customers,” says McHenry.

Impact of AWS HealthScribe Integration

The integration of AWS HealthScribe into ContrastHub has led to improvements in both efficiency and satisfaction for healthcare providers. Following are some of the key benefits that Contrast AI is experiencing:

  • Improved Charting Efficiency: With AWS HealthScribe, Contrast AI has helped healthcare providers achieve an 80% reduction in documentation time for each patient visit. Dr. Iliev, the medical director at Southdale pediatrics, a user of Contrast AI, noted, “Contrast AI has significantly improved my efficiency and work satisfaction, which has benefitted my patients as well.” Clinics can now process up to 1,000 patient records daily, compared to 200 previously, and access critical data within 24 hours instead of weeks.
  • Enhanced Patient Satisfaction: The streamlined documentation process has also resulted in enhanced patient satisfaction. With 99% of patients consenting to the use of Contrast AI’s technology, it is leading to shorter wait times, greater doctor-patient engagement, and an overall improvement in the patient experience. Dr. Iliev shared, “I find patients are very pleased to see technology being used to improve the patient-physician interaction.”
  • Increased Physician Satisfaction: Physicians using Contrast AI have reported a 93% increase in satisfaction with the documentation process. Contrast AI’s intuitive and easy-to-use interface has been a significant factor in this increase.

Contrast AI’s successful integration of AWS services demonstrates the power of cloud-based solutions in transforming medical documentation. By automating the generation of comprehensive medical notes and extracting key data, Contrast AI helps healthcare providers focus more on patient care and less on administrative tasks.

Conclusion

By using AWS, Contrast AI has transformed healthcare workflows, significantly reducing documentation time and costs. This collaborative effort has streamlined physician workflows, allowing for greater focus on patient care. By partnering with AWS, Contrast AI successfully migrated existing processes to the cloud within a week, achieving a seamless transition and substantial time savings.

Dr. Skallerud’s personal journey from experiencing physician burnout to creating Contrast AI underscores the transformative power of technology in healthcare. As Contrast AI continues to innovate, they look forward to further improving healthcare delivery through advanced technology.

As healthcare technology continues to evolve, AWS remains at the forefront of innovation, providing powerful tools like AWS HealthScribe to transform physician documentation.
Learn more about how AWS is revolutionizing healthcare workflows and improving patient care.

Check out more AWS Partners or contact an AWS Representative to know how we can help accelerate your business.

Further Reading

Dr. Kai Skallerud

Dr. Kai Skallerud

Dr. Kai Skallerud, MD, MBA is the founder and CEO of Contrast AI, a healthcare startup. With years of experience in clinical medicine, business, and finance Dr. Skallerud is a physician-innovator in training who is committed to advancing the healthcare industry to benefit patients and providers alike.

Clement Perrot

Clement Perrot

Clement Perrot is a Senior Solutions Architect and AI/ML Specialist at Amazon Web Services (AWS), where he helps early-stage startups build and leverage AI on the AWS platform. Prior to AWS, Clement was an entrepreneur who’s last two AI and consumer hardware startups were acquired.

Prasad Prabhu

Prasad Prabhu

Prasad Prabhu is a Principal AI/ML Specialist for Healthcare & Life Sciences at Amazon Web Services (AWS), where he focuses on helping customers understand and adopt generative AI and AI/ML services to address their business challenges. With nearly 20 years of experience in the technology industry, Prasad is specialized in building B2B enterprise software products and solutions.

Sarthak Handa

Sarthak Handa

Sarthak Handa serves as a Principal Product Manager at Amazon Web Services (AWS) AI/ML in Seattle, Washington, where his primary focus is on developing AI services that facilitate advancements in the healthcare industry. Prior to his work at AWS, Sarthak spent several years as a startup founder, building technology solutions for the healthcare and disaster relief sectors.