Emerald Health LLC - Electronic Referral System and EMR Software

Blog

31 July
2018
By admin | 0 comments

Clinical Context & Synopsis Generation from EHR

Clinical workload is increasing. Clinicians are seeing more patients and radiologists are having to read more scans in a day. Also, as the electronic health record system gets more sophisticated and detailed, we have access to unprecedented amounts of data. This is a scenario where doctors have more patients to see, more data to consume and better results to aim for. It’s a challenge and an opportunity.

 

EHR’s are getting smarter. They are no longer the dumping grounds for every bit of digitized patient data. It is a mineable repository of extremely valuable information shrouded in redundant noise. In the early days of EHR, it was difficult to sift through hordes of lab values and clinical/rad/path reports and have a nice summary of what’s been going on with the patient. A physician would have had to spend an inordinate amount of time to do so before every encounter. When HL7 FHIR came about, mining data from EHR became possible. It opened the floodgates for developers and informaticists to work with physicians to make EHR more user-friendly, providing itemized, relevant information as needed in a clinical workflow.

The system has to be reliable in the sense that it doesn't miss out key information from a parent source

The main concept of clinical context generation is that the care provider should have ready access to all the relevant information pertaining to the patient he/she is-bout to conduct the care of. The technology that made it possible is Natural Language Processing (NLP), which is a methodology concerned with automated interpretation and generation of human language. It depends on ontology libraries to help mine data through FHIR to create synoptic reports for real-time consumption by the physician. An example of that would be that if patient XYZ is having an MRI scan of his knee for meniscus tear surgery, a synoptic report would be created using keyword searches and text mining into the clinic notes, prior radiology reports, lab reports, prior procedure notes, etc.. The report would use NLP to create human language format and embed all the information that is ranked based on hierarchy of relevance. The radiologist would then be able to access just that report, as opposed to all the patient chart.

 

The system has to be reliable in the sense that it doesn’t miss out key information from a parent source, weights it correctly so that it makes it at the right degree of importance in the synopsis, and that it is succinct yet covering all key aspects of that patient’s history. A system has to practice through a lot of context generation exercises to be able to get to that level. A nuanced feedback approach will fine tune it to ultimately provide a highly intelligent report that greatly. improves clinical efficiency.

 

Such reporting systems can serve as historical milestones in the patient’s chart in the EHR, organizing until it becomes a diligently maintained streamlined system. User experience will play an important role in creating a system that works in the background and generates such milestones (synopses), which are easy to access, fast to process, palatable to consume and overall, eases the burden of the clinician rather than further complicate the EHR.

Share your Story. Choose your Platform

Leave a Reply

Your email address will not be published. Required fields are marked *

Latest Videos

Recent Tweets

Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text.

http://www.google.co.in/?gws_rd=cr
3 months ago

Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text.

http://www.google.co.in/?gws_rd=cr
3 months ago