Health Navigator announces release of V 12.5
(July 2018) Version 12.5 includes a new integration with telehealth devices, improved triage and diagnosis engine performance, and other areas of expanded functionality.
The Health Navigator API content version 12.5h was released to production on Monday, July 2nd, 2018. This is the 5th update to the Health Navigator clinical content this year. Over the past six months we have made numerous improvements to the clinical content and functionality. Here are some of the great enhancements we have made.
We continue to expand the clinical vocabulary to support the beginning, middle, and end of digital health and telemedicine encounters.
- * Added 3 new Coded Chief Complaints: Head Bleeding, Mouth or Tongue Bleeding, and Tooth Loss or Knocked Out. There are now 483 distinct Coded Chief Complaints that can be used as the reason for visit at the beginning of a medical encounter.
- * Added 45 new diagnoses, including AV Reciprocating Tachycardia, Cannabinoid Hyperemesis Syndrome, Digital Myxoid Cyst, Fever without a Known Cause, Separation Anxiety, and White Coat Hypertension. There are a total of 2,724 distinct diseases and conditions generated by the Diagnosis Engine.
- * Added over 275 new German Alternate Terms and edits to existing German terms, based on client input.
The clinical vocabulary is now available in eight different languages.
Expanded Triage Engine Dispositions
We added 5 new dispositions to the Health Navigator Triage Engine:
* Labor & Delivery Unit
* Poison Center
* Dentist Today
* Dentist Today or Tomorrow
* Dentist 3 ‐ 7 Days
Would you like to see this in action? There are 3 new scenarios ( #24, #25, #26) available on the v12.5 TriageTest Scenarios spreadsheet (XLS) that we share with partners. Try these using the Triage Symptom Checker demo application on the Health Navigator portal (https://portal.healthnavigatorapis.com). There are now 14 different triage dispositions.
Our new API method ListTelehealthDevicesByConceptList returns a list of the telehealth devices (e.g., Ear Exam, Heart Rate, Temperature, Throat Exam) to prompt a patient to use during an eVisit intake process, based on the patient’s presenting symptoms or problems. This new method and capability answers the question: What are the best and most appropriate telehealth devices for a patient to use during a telemedicine intake process?
Here is a sample use case scenario: A telemedicine company encourages its patients to use telehealth intake devices such as Heart Rate, Throat Exam, during the telemedicine intake process. The telemedicine company wants to recommend just the needed assessment for a specific symptom. For example, a patient with throat pain would be encouraged to use the Throat Exam camera and also provide a Temperature.
Health Navigator continues to test (and tune) the NLP, Clinical Documentation Support, Diagnosis, and Triage Engines against large volumes of medical call center, emergency department, and telemedicine encounters.
* Telemedicine encounter data from Q1 2018 (4,500+ encounters) was abstracted and coded by our testing team; used for testing the Health Navigator engines; and analyzed. Results were then methodically reviewed by the Health Navigator Medical Review Board (MRB).
* Analyzed and reviewed 40,000 medical call center and telehealth encounters from January 2018 for which one or more Coded Chief Complaints were found. Primarily looking for incorrect matches. This type of testing helps ensure specificity, which for the Health Navigator NLP engine typically averages 99%.
* Reviewed approximately 4,000 usages from January 2018 of our NLP demo application where NO CCC’s were found. This includes single letter entries, weird entries (“scooby doo!”), practice entries, and grossly misspelled words… in addition to some “real NLP gold” in terms of classifiable and valid entries. This type of testing work improves NLP sensitivity, which typically runs 93-98% depending on the data source and spelling.
* Analyzed and reviewed 80,000 emergency department reason for visits and associated metadata using Health Navigator Natural Language Processing engine and clinical vocabulary. Results were summarized and published in May 2018 issue of Academic Emergency Medicine. View abstract and chart.