MicroHealth, LLC - Company of the Year
Company: Microhealth, LLC, Vienna, VA, USA
Company Description: MicroHealth, LLC is an 8(a) SDVOSB Small Business that provides customers with the right information to promote better decision-making to improve our Nation’s health and wellness. We harness the combined knowledge and expertise from our seasoned team of health professionals, researchers, informaticists and technologists who specialize in research, health information management, and technology.
Nomination Category: Company / Organization Categories
Nomination Sub Category: Company of the Year - Health Products & Services - Medium-size
Nomination Title: MicroHealth, LLC
Tell the story about what this nominated organization has achieved since 1 January 2017 (up to 650 words). Focus on specific accomplishments, and relate these accomplishments to past performance or industry norms.
The Department of Defense (DoD) and the Department of Veterans Affairs (VA) both have Electronic Health Records (EHRs) with decades of information stored as free text clinical narratives. Though there has been much progress in structuring the data for analysis and abstraction, there are still many terabytes of valuable information. A large amount of unstructured data exists in EMR legacy systems. Unstructured data has the potential to provide a wealth of information to improve outcomes and provide an improved patient experience. Recently the Joint Program Commission (JPC-1) awarded funds for a proof of concept (PoC) evaluating technology to process unstructured clinician notes into a higher value, consumable, summarized and searchable format. The objective of the PoC is to leverage Natural Language Processing (NLP) and Machine Learning (ML) to structure free text clinical notes from DoD and VA systems and create a user interface (UI) to view and search processed notes.
To unlock decades of textual data, MicroHealth is supporting Madigan Army Medical Center (MAMC) and the Interagency Program Office (IPO) in transforming unstructured text into normalized structured data. The MAMC/IPO PoC built a Joint Data Repository (JDR) containing medical records from DoD, VA, and Managed Care Support Contractor (MCSC) data sources. The JDR data is linked to the NLP software to process the medical records.
Using NLP and ML, millions of clinical records are processed and predictive insights (e.g., ejection fraction, concussion history) are extracted to augment original data sources. This extracted data is then presented to clinician end users via a Private Key Infrastructure (PKI) authenticated UI to enable search and reporting functions for selected use case(s).
Selected PoC use cases include:
-A clinical provider needs to know which patients have an implanted device subject to manufacturer device recall; search by device type, manufacture, model or date
-A clinical provider wants to search for implanted device(s) for a given patient, using patient ID, to authorize a medical procedure (e.g., MRI-compatibility)
-A clinical provider wants to search a patient’s record for a specific concept(s) (e.g., clinical diagnoses)
Application of NLP and ML to free text clinical data assists providers with clinical decision support (CDS) by providing all relevant patient information to providers at the point of care, which aids in diagnosis and treatment.
For example, a manufacturer of a commonly used cardiac pacemaker has issued a recall due to battery issues. The Defense Health Agency’s (DHA) medical facilities, of which MAMC is one, may have implanted the recalled device in one or more patient(s). MAMC must identify all patients with the recalled device and contact them regarding suggested fixes to the identified issue. Leveraging NLP with ML technology, providers can receive a printable list of all instances of the use of the identified pacemaker in free text within MAMC patient clinical records. The report pulls back patients by name, Primary Care Manager (PCM), contact information, and facility where care is currently provided, as well as any MCSC who may have implanted the device. In this example, we apply NLP and ML to extract information on devices from the free text of patient records, index extracted terms to allow for concept/device information search or data aggregation to generate reports, then compile the contextual information (e.g., demographic details, medical facility, contact information) into an exportable file the end user can use to execute device recalls.
In bullet-list form, briefly summarize up to ten (10) of the chief accomplishments of this organization since the beginning of 2017 (up to 150 words).
-Inc. 5000 Fastest Growing Companies List (#1593)
-Inc. 5000 Best Place to Work Award
-International Stevie Award (Bronze): Woman of the Year
-International Stevie Award (Bronze): Customer Service Team of the Year
-International Stevie Award (Bronze): Management Team of the Year
-Virginia Fantastic 50 List (#33)
Of the following measures of success, which ONE do you want the judges to most appreciate about your organization's story of achievement since the beginning of 2017? Product Innovation