jargone

jargone demystifies medical records to address the costs linked to provider-patient miscommunication.

Comment 1

Inspiration

Healthcare consumers recall less than 50% of the information presented during their interaction with physicians. Away from the hospital or physician office, consumers may access their patient portal for a visit summary. However, as the Center for Disease Control and Prevention (CDC) stated that “9 out of 10 have difficulty following up routine medical advice because it is often incomprehensible to an average population.” Studies suggest that the asymmetry of information and miscommunication between provider and patient leads to reduced patient satisfaction, decreased medication adherence, increased co-morbidities and unnecessary admissions, and poor health outcomes. This issue was our inspiration to develop jargone, a platform to simplify doctor-patient communication.

What it does

jargone simplifies doctor-patient communication by demystifying medical terminology. Integrating with electronic health records, our online platform overlays on consumers’ medical record to translate medical jargon into easy-to-digest information. Our thesis states that improved communication between patient and provider supports care plan and medication adherence, adoption of preventive health behaviors, and patient satisfaction. We seek to test this thesis and supporting research through our pilot study.

How we built it

Our system would be implemented as part of an online patient portal that presents patients with more transparency regarding their medical data. We designed our prototype using Flask, a Python web framework, and implemented the view with Bootstrap/HTML/Javascript components. Our service is able to detect important information from convoluted electronic health records and physician notes, displaying this information to the user in a clear and organized manner. The information obtained from this process is further refined and simplified using a medical dictionary API to make it understandable for patients. We demonstrated proof of concept of our prototype using medical notes from an actual mole biopsy (patient consent was given).

Challenges we ran into

As a natural part of the brainstorming process, we spent an extended time determining our final concept. In addition, our original plan was to use the Iodine API. However, this was not available, and the team required a different approach. Ultimately, we utilized another medical dictionary API to provide a similar product.

Accomplishments that we're proud of

Over the course of the weekend, our team coalesced around the jargone vision. Although we began with a different concept, our diverse backgrounds (design, engineering, business, medicine) pushed our original concept to a new level. During the process we leaned on our respective strengths and unique perspectives to develop a financially-viable and impactful platform. In a short debrief this morning, each member of our team expressed enjoyment with our team.

What we learned

Through our brainstorming session, we learned the value of constructive conflict in challenging ideas to create a better product. Our team was conscious to give each member a voice during debate/discussion to ensure a common direction and goal. With no clear competitors in the space, our team needed to be resourceful in our research and framing of the situation. Finally, we witnessed the importance of humor to decompressed and bring the team back together after a few hours of work.

What's next for jargone

By training machine learning models with the sentences that doctors deem most important, we hope to better provide automatic extraction of key takeaways that doctors want to emphasize. We plan to combine this information with sophisticated medical dictionaries and refine the resulting text in a way that is more comprehensible for the patients (MVP 1.0). For example, with patients who suffer from chronic heart failure doctors often recommend they drastically improve diet and exercise. By detecting key words and phrases in a patient’s data, our model will be able to automatically include such phrases in the important information section of a patient’s portal. Then we hope to work with burgeoning Personal Health Records (PHR) startups to test our model and refine the technology (MVP 2.0+). Ultimately, we are planning to conduct a pilot study in collaboration with a Health System to prove the delivery of our value propositions and start reaching out to customers, insurance companies and health systems.

Built With

Share this project:

Updates