The variability and the complexity of the data needed for clinical care requires clinicians to accurately and efficiently recognize COVID-19 amongst individuals, ranging from asymptomatic infection to multiorgan and systemic manifestations. COVID-19, like sepsis, involves different disease etiologies that span a wide range of syndromes (e.g., initial, inflammatory, hyperinflammatory response). Because patients can present with mild, moderate, or severe symptoms, clinicians must both identify the disease stage and optimal treatment. The factors that trigger severe illness in COVID-19 patients are not completely understood. Like other complex, challenging diagnoses, clinicians in the trenches struggle to diagnose and treat patients using data available in the electronic health record (EHR). In our current NIH NLM R01 ?Signaling Sepsis: Developing a Framework to Optimize Alert Design?, we created sepsis specific enhanced visual display models that outranked preference and performance when compared with the usual care of fragmented, non-directed information gathering. For this supplement, we propose the design and development of COVID-19 diagnosis and clinical management enhanced visual display models to support clinicians? recognition of critical phases in COVID-19 diagnosis and treatment decisions. In order to create the models, we will identify relevant diagnostic and treatment data elements that will include clinical characteristics, laboratory results, and radiology results (e.g., chest CT). Our project will survey emerging models of COVID-19 and its stages, and ensure our models are congruent with best practices that emerge as our knowledge as a medical community evolves. The models provide an EHR based method to mine clinical data to identify the presence of COVID-19 which supports the variety of ways in which COVID-19 presents, availability of data elements, accuracy of diagnostic tests, and the highly infective nature of the disease.
Specific Aim 1 : To identify emerging patient-specific clinical features of COVID-19 and testing analytics to present critical information for COVID-19 diagnosis and clinical management. Elements include the characteristics listed above (e.g., symptoms, co-morbidities) plus COVID-19 specific test results, including data specific to the tests? positive and negative predictive values.
Specific Aim 2 : To develop an EHR embedded CDS tool using our COVID-19 enhanced visual display models using synthesized information obtained through the NLM parent grant and Specific Aim 1. Evaluate the technical feasibility and usability of the novel COVID-19 CDS tool. Why It Matters: During a pandemic, there?s no room for ambiguity as clinicians are required to comb through the EHR. The ability to better visualize and interpret EHR data supports optimal diagnosis and clinical management. Our enhanced visual display models will support clinicians as they evaluate demographic factors, underlying conditions, and comorbidities that identify patients at higher risk of morbidity and mortality and will therefore drive better clinical management.
In our current NIH NLM R01 ?Signaling Sepsis: Developing a Framework to Optimize Alert Design?, we created sepsis specific enhanced visual display models that outranked preference and performance when compared with the usual care of fragmented, non-directed information gathering. For this supplement, we propose the design and development of COVID-19 diagnosis and clinical management enhanced visual display models to support clinicians? recognition of critical phases in COVID-19 diagnosis and treatment decisions. The models provide an EHR based method to mine clinical data to identify the presence of COVID-19 which supports the variety of ways in which COVID-19 presents, availability of data elements, accuracy of diagnostic tests, and the highly infective nature of the disease.
| Miller, Kristen; Mosby, Danielle; Capan, Muge et al. (2018) Interface, information, interaction: a narrative review of design and functional requirements for clinical decision support. J Am Med Inform Assoc 25:585-592 |
| Capan, Muge; Hoover, Stephen; Miller, Kristen E et al. (2018) Data-driven approach to Early Warning Score-based alert management. BMJ Open Qual 7:e000088 |
| Miller, Kristen; Capan, Muge; Weldon, Danielle et al. (2018) The design of decisions: Matching clinical decision support recommendations to Nielsen's design heuristics. Int J Med Inform 117:19-25 |
| Long, Devida; Capan, Muge; Mascioli, Susan et al. (2018) Evaluation of User-Interface Alert Displays for Clinical Decision Support Systems for Sepsis. Crit Care Nurse 38:46-54 |
