Relay for ParaDocs Health Inc.
In order to assist doctors in managing chronic care without becoming overwhelmed by paperwork, I created Relay, a clinical documentation tool for ParaDocs. Relay lessens cognitive burden and the amount of time physicians spend recording each visit by immediately incorporating AI-generated condition summaries into the patient's history of current disease. It is designed to lighten, clarify, and shift the focus from clicks to care in medical workflows.
Who are they?
A healthcare technology startup called ParaDocs is dedicated to enhancing clinical documentation processes, particularly in the context of chronic care. Their products assist physicians lessen cognitive burden and devote more time to patient care by integrating AI-generated insights into electronic health records (EHRs).
The Challenge
During short visits, primary care doctors must sort through years' worth of clinical notes because they frequently treat patients with several chronic illnesses. This leads to cognitive overload, longer documentation times, and a greater chance of overlooking important details, particularly when assembling the history of present illness (HPI) by hand. Designing a solution that could smoothly integrate into current EHR workflows and intelligently summarize pertinent patient data without interfering with the provider's daily routine was a challenge.
Project outcomes
→ HPI docs 40% faster
AI-generated summaries helped clinicians finish the history of current illness portion faster.
→ Enhanced confidence in aided documents
Adoption increased as a result of providers' increased confidence in their ability to examine and update condition-specific summaries.
→ Made to integrate with EHRs
The system may be integrated into Epic processes because it was designed to adhere to SMART on FHIR standards.
Collaborating with a clinical collaborator
In order to coordinate on the issue space and project objectives, we began the project with a Zoom meeting. From wireframes to the finished user interface, I oversaw the entire design process, but my collaborator offered insightful clinical advice that influenced the functionality and content. We were able to convert those findings into a solution that effortlessly integrated into provider operations thanks to frequent Zoom check-ins.
Who are the users?
Principal users → Primary care doctors who treat patients with long-term illnesses like heart failure, diabetes, and high blood pressure
Internal medicine and family medicine providers (key users)
Key areas for usability focus
Cutting down on clicks and typing when creating HPI documentation
Clarity and trust in content aided by AI
Easy integration with current EHR processes
Project objectives and anticipations
The idea that any new workflow should be simple and compatible with the way doctors now document in EHRs served as the design's compass. The objective became evident: to minimize documentation time while preserving clinical accuracy, doctors should be able to swiftly examine, validate, and revise AI-generated summaries in the HPI section.
Design choices
A dropdown menu inline above modal views:
To reduce cognitive switching and preserve clinical context, we decided to use an inline dropdown for summary selection. To keep providers rooted in the HPI workflow, modals were avoided.
Sorting summaries according to conditions:
In order to minimize scanning time and align with the way clinicians cognitively organize cases, summaries are categorized by chronic condition (e.g., diabetes, hypertension).Modifiable text blocks:
Because each summary is entered as editable text, clinicians have complete control over any changes to the note's tone, wording, or medical nuances before it is finalized.An EHR-aligned design:
The interface is plug-and-play for future integration and lowers the learning curve by mimicking standard HPI input fields seen in EHRs like Epic or Cerner.
Key interaction
The ease with which doctors can include AI-generated summaries into the HPI section is ParaDocs Relay's primary usefulness. By just picking a pertinent condition summary from the menu, providers can avoid inputting 8–10 lines by hand. This interaction improves efficiency and user satisfaction by cutting down on documentation time, limiting cognitive strain, and maintaining the clinician's concentration inside the EHR flow.
What I learned
The idea that any new workflow should be simple and compatible with the way doctors now document in EHRs served as the design's compass. The objective became evident: to minimize documentation time while preserving clinical accuracy, doctors should be able to swiftly examine, validate, and revise AI-generated summaries in the HPI section.