Industry
Client Overview
Mid-West Regional Health (MRH) is a respected non-profit hospital system with 15 hospitals and over 200 clinics. Like many healthcare providers, they were facing a crisis of clinician burnout. Their internal surveys showed that physicians were spending over three hours per day on administrative tasks, with clinical documentation in the EHR being the primary culprit. This "pajama time" doctors finishing notes late at night was leading to decreased job satisfaction and increased risk of errors. MRH's leadership knew they needed an innovative solution to give their doctors time back for patient care.
Client Testimonial
"The impact of CIS's AI scribe has been nothing short of transformational. Our clinician satisfaction scores have improved by 30%, and we've seen a measurable improvement in the quality of our clinical notes. CIS was more than a vendor; they were a true partner. They took the time to understand our clinical workflows and the sensitivities around patient privacy. Their CMMI Level 5 discipline was evident in every step of the project, from security design to pilot execution." - Virginia Woolf, Chief Information Officer, Mid-West Regional Health System
Problem
The manual process of typing or dictating patient encounter notes into the EHR was time-consuming, inefficient, and a major source of frustration for clinicians. Existing dictation services were slow and often inaccurate, requiring extensive correction. MRH needed a solution that could automatically and accurately capture the patient encounter and generate a structured, compliant clinical note directly in their Cerner EHR.
Key Challenges
-
01
Clinician Adoption : Doctors are busy and skeptical of new technology that might disrupt their workflow. The solution had to be seamless and demonstrably better than the status quo.
-
02
Accuracy and Nuance : Medical conversations are complex, filled with jargon, accents, and interruptions. The AI needed to be highly accurate in its transcription and understanding.
-
03
EHR Integration : The generated note had to be inserted into the correct fields within the Cerner EHR system automatically, without requiring manual copy-pasting.
-
04
Data Privacy : The solution involved capturing sensitive patient-doctor conversations, making HIPAA compliance and data security the absolute top priority.
Our Solution
CIS partnered with MRH's IT and clinical leadership to develop a custom, Generative AI-powered "Ambient Scribe" application. The solution was designed with a clinician-first and privacy-first mindset.
Ambient Listening Mobile App : We developed a simple, secure mobile app that physicians could activate at the start of a patient encounter. The app used a high-fidelity audio capture module and streamed the encrypted audio to a secure server in real-time.
Proprietary NLP & GenAI Pipeline : We built a sophisticated AI pipeline on a HIPAA-compliant Azure environment. The pipeline consisted of:
- A fine-tuned speech-to-text model for high-accuracy transcription of medical conversations.
- A speaker diarization module to separate doctor and patient speech.
- A custom-trained Large Language Model (LLM) that would take the transcript and generate a structured clinical note in the SOAP (Subjective, Objective,Assessment, Plan) format.
"Human-in-the-Loop" Verification : The generated note was not automatically filed. It was presented to the physician in a review screen for verification and editing. This critical step ensured clinician oversight and accountability, which was key for building trust and mitigating risk.
Cerner Integration via Smart on FHIR : We built a Smart on FHIR application that allowed the verified note to be seamlessly and securely committed to the patient's record in Cerner with a single click, populating the correct fields automatically.
Implementation & Execution
-
Feasibility & Model Training
We collected and de-identified thousands of existing clinical notes and dictation transcripts from MRH to fine-tune our LLM on their specific language and formatting.
-
Secure Infrastructure & App Prototype
We built the HIPAA-compliant Azure infrastructure and developed the initial version of the mobile app and the physician review interface.
-
Integration & Workflow Design
We worked with MRH's Cerner integration team to build and test the Smart on FHIR application, ensuring data flowed correctly and securely.
-
Pilot Program with Champions
We launched a pilot with a small group of 20 "physician champions" from different specialties (primary care, cardiology, orthopedics).
-
Iteration Based on Feedback
We used feedback from the pilot to rapidly improve the model's accuracy across specialties and refine the UI to make it even more intuitive.
-
System-Wide Rollout & Training
Following the successful pilot, we developed a training program and began a phased rollout of the application across the entire health system.
Positive Outcome
1. 45-Minute Reduction in Daily Documentation
On average, physicians using the AI scribe reported saving 45 minutes per day on documentation.
2. 30% Improvement in Clinician Satisfaction
In post-rollout surveys, overall job satisfaction among physicians increased by 30%, with many citing the AI scribe as the primary reason.
3. Improved Note Quality
The AI-generated notes were found to be more complete and structured than the previous dictated or typed notes, leading to better data for downstream analytics and billing.
4. 90% Adoption Rate
Within six months of the full rollout, 90% of eligible physicians were actively using the system, a testament to its usability and value.
Why Choose Us
-
Verifiable Process Maturity
Our structured, phased approach (from feasibility to pilot to rollout) was critical for a project with such high stakes.
-
Fortress-Grade Security
We designed the solution with end-to-end encryption and data segregation, meeting the stringent security requirements of MRH's board.
-
Deep Domain Expertise
Our understanding of clinical workflows and the importance of the "human-in-the-loop" was key to gaining clinician trust.
-
Measurable ROI Focus
The project was framed around the clear ROI of reducing burnout and improving efficiency.
-
100% In-House Experts
Our team of AI engineers, cloud architects, and UI/UX designers worked in concert to deliver a polished, integrated solution.
-
Flexible Engagement PODs
We provided a dedicated POD that felt like an internal innovation team to MRH.
-
Radical Transparency
MRH's CIO and clinical leads were involved in weekly demos and sprint planning sessions.
-
Future-Proof Technology
The solution was built using the latest in Generative AI and cloud technology, positioning MRH as a leader in digital health.
-
Guaranteed Peace of Mind
We de-risked a cutting-edge AI project for a large, cautious organization and delivered transformational results.
Conclusion
This case study illustrates CIS's ability to tackle one of the biggest challenges in healthcare: clinician burnout. By combining cutting-edge Generative AI with a deep understanding of clinical workflows and enterprise-grade security, we delivered a solution that provided immense value to both the hospital system and its most valuable asset: its physicians.