Picture of AI Healthcare

Clinical Informatics Lab

    FosterGossDr. Foster Goss, DO, MMSc, FACEP

    Dr. Goss completed his residency training at Albert Einstein Medical Center and then pursued a National Library of Medicine sponsored fellowship and Masters in Medical Science in Biomedical Informatics at Harvard Medical School and Clinical Decision Making fellowship at Tufts Medical Center. His research focuses on natural language processing, decision analysis and information retrieval, as it relates to patient safety, communication and coordination of care. Dr. Goss has been involved in many of the innovation efforts at UCHealth including introduction of Clinical Pathways (AgileMD) and Real-Time Benefit/indication-based prescribing tools (SwiftRx). 

    He was the founder of a startup, CareLoop, a healthcare company focused on improving patient safety, communication, and accountability across transitions in care. This company was recently acquired by DispatchHealth. 

    His work has led to multiple publications, grants, and presentations at national conferences.

    Active Projects

    1. Improving Allergy Documentation and Clinical Decision Support in the EHR
      Developing advanced clinical decision support for improving the accuracy of allergy alerting leveraging NLP and big data. 
    2. Artificial Intelligence to interpret Electrocardiograms
      Developing machine learning tools to automatically read and interpret ECGs.
    3. Indication based prescribing to improve antibiotic stewardship for common ED infections. 

    Other areas of interest and applications

    • Patient Safety
    • Quality of Care
    • Natural Language Processing
    • Machine Learning
    • Adverse Drug Events
    • Speech Recognition
    • Electronic Health Records
    • Clinical Decision Support
    • Clinical Decision Making/Decision Analysis
    • Decision Trees/Monte Carla Simulation/Modeling
    • Standard Terminologies

    Collaborators

    2018
    AHRQ R01
    Improving Allergy Documentation and Clinical Decision Support in the EHR
    Developing advanced clinical decision support for improving the accuracy of allergy alerting leveraging NLP and big data. 
    Role: Co-I, Site PI

    2016
    AHRQ 1R21HS024541-01 
    NLP to identify and rank clinically relevant information from EHRs in the acute care setting
    The goal of this project is to develop a natural language processing tool identify contextually relevant information from the EHR that is relevant to a patients presenting problem. 
    Role: PI

    2015
    NLP to improve the accuracy and quality of dictated medical documents
    AHRQ/R01 – Partners Health Care, University of Colorado Hospital
    Role: Co-investigator 
    The proposed work provides a new, systematic approach to address the consistently high number of dictation errors in medical documents, with the motivation to improve the accuracy of clinical documentation and an ultimate goal to improve the quality, safety and efficiency of care. The proposed work also reduces costs by saving physicians and transcriptionists’ time that they would have spent proofreading dictated documents.

    2013
    National Emergency Medicine Chief Complaint Ontology
    American College of Emergency Physicians (ACEP) Section Grant – Beth Israel Deaconess Medical Center, University of Colorado Hospital, Brigham and Women’s, Summa Health System, University of Nebraska
    Role: Team member 
    This project developed a national, standardized chief complaint ontology for the Emergency Department that can be utilized by any emergency department with an ED information system. A standardized chief complaint vocabulary will allow administrators as well as researchers to accurately and systematically represent the reason for visit in the emergency department as structured data that will facilitate comparison of patients within an institution and across institutions. A structured chief complaint can then be used to facilitate 1) Clinical Care and ED Operations 2) Quality Assurance, Improvement, and Measurement, 3) Education, 4) Surveillance, and 5) Research.

    2012
    Encoding and processing patient allergy information in EHRs
    AHRQ/R01 - Partners Healthcare and University of Colorado Hospital
    Role: Co-investigator 
    The major goal of this study is to build a knowledge base and NLP tool for representing allergy information such that free-text entries are automatically identified, extracted and encoded from clinical notes (ED/inpatient). This will allow important drug-allergy checking to occur and prevent adverse drug events.

    In the News

    • How will artificial intelligence affect health care?

      Feb 21, 2019
      The explosion of big data promises potential breakthroughs in disease treatments, but, just as in the development of new drugs, scientists and clinicians must exercise caution in how they apply algorithms and other technologies, according to a CU Anschutz panel of experts.
      Full story

    • Errors Common in Notes Produced by Speech Recognition Software

      Jul 6, 2018
      Although computerized speech recognition (SR) may lighten physicians' documentation load, a new study published online July 6 in JAMA Network Open calls into question the accuracy of this time-saving software. The data reveal an error rate of more than seven words per 100 in unedited SR-generated documents, including clinically significant errors in one of every 250 words that could affect care.
      Full story

    • Combatting the Opioid Crisis

      Aug 25, 2017
      In 2015, 97.5 million people in the United States used prescription pain relievers.1 That’s 36.4% of the population. Of those, 12.5 million misused the medication and 33,091 died from an overdose.2 Emergency physicians see the impact of the opioid crisis every day.
      Full story

    • Food Allergies Found in Less Than 4 Percent of Americans

      Jun 1, 2017
      Some 3.6 percent of Americans — fewer than one in 25 — have at least one food allergy or intolerance, according to a new report. Those numbers are lower than many earlier estimates.
      Full story

    • 3.6 Percent Of Americans Found To Have Food Allergies Or Intolerances

      Jun 1, 2017
      Researchers are giving us new insight into the problem of food allergies and intolerances. A new study out of Brigham and Women's Hospital finds 3.6 percent of Americans are dealing with those problems.
      Full story

    • Study Offers Hard Data on Food Allergies

      May 31, 2017
      Anecdotal evidence of food allergies abounds, but just how common are these allergies and intolerances? In a new study, investigators from Brigham and Women’s Hospital combed through medical records from more than 2.7 million patients, identifying more than 97,000 with one or more documented food allergy or intolerance.
      Full story

    • UCHealth brings Facebook-like feed to its ED through CareLoop

      Apr 10, 2017
      According to patients, the application of a social feed in the University of Colorado Hospital emergency department helped improve overall communication.
      Full story

    • Is Speech Recognition Viable in the ED?

      Apr 1, 2017
      There's a lot going on in emergency departments (EDs), most of it centered on saving lives. Documentation may rank pretty low on the staff's list of priorities, but that doesn't diminish its significance. To make note-taking more convenient, some hospitals have turned to speech recognition for help. Is it effective?
      Full story

    • UCHealth Denver pilots social media-style ER updates

      Mar 30, 2017
      Denver-based UCHealth deployed new tech into UC Anschutz Medical Campus' emergency department, Foster Goss, DO, founder of the digital health company CareLoop, told Becker's Hospital Review via email.
      Full story

    • UCHealth Denver pilots 'Facebook feed' for patient communication in its ER

      Mar 24, 2017
      The hospital tested CareLoop technology and found that both clinicians and patients responded very positively to the social-style method for keeping up to date on care.
      Full story