Venous thromboembolism (VTE), consisting of deep venous thrombosis (DVT) and pulmonary embolism (PE), is a deadly disease with 30-day and one-year case fatality in the general population of 14% and 29%, respectively. This concern has increased with the recent pandemic as complications due to VTE have been associated with 40% of deaths of patients with Coronavirus Disease 2019 (COVID-19). The overarching aim of this proposal is: 1) to refine and validate two clinical prediction rules (CPRs) for assessment of VTE in patients presenting to or admitted from the emergency department (ED) with COVID-19 and 2) to design, perform usability testing, and implement two clinical decision support (CDS) tools based on these CPRs within our novel, innovative service-oriented architecture (SOA)-based complex CDS system (CDSS). Wells? criteria in combination with D-Dimer is a CPR that helps healthcare providers exclude PE and the need for unnecessary computed tomography pulmonary angiography (CTPA). In addition, acutely-ill hospitalized medical patients are at moderate-to-high risk for developing VTE events during the hospital stay. Hospitalized COVID-19 patients are at especially high risk of VTE, with recent reports suggest a prevalence of DVT as high as 46% in medical wards and PE as high as 42% in ICU settings. These data are consistent with VTE risk associated with previous viral pneumonias such as H1N1 which have shown a 23-fold increased risk for VTE. Thrombo-prophylaxis shows significant benefits for in-hospital period, but must be used appropriately as it carries risk for morbidity and mortality related to bleeding. VTE risk in admitted patients is difficult to assess and physicians rely on CPRs such as IMPROVE-D-Dimer (DD) for patient care. Our parent grant (1R18HS026196-01A1) proposed the use of two validated CPRs (Wells? criteria and the IMPROVE model) within the context of a novel CDSS, building on a SOA that will function across multiple electronic health record (EHR) platforms. We will build upon our original idea by refining and validating the Wells? criteria in the ED and IMPROVE-DD risk model in hospitalized COVID-19 patients and then implementing the CDS by utilizing the groundwork set forth within the platform we are currently developing for the parent grant. This work is significant because CPRs for VTE assessment are essential but have not been established or validated in COVID-19 patients. We are uniquely suited to perform this work at Northwell Health as we are at the forefront of fighting the pandemic, having treated over 150,000 COVID-19 patients with subsequent performance of high-level big data analysis on this cohort and are well positioned for appropriate validation and subsequent implementation of VTE risk assessment. The long-term goal of this project is implementation of evidence-based CDS tools on a widely disseminated platform to meaningfully inform universal VTE management of COVID-19 patients.
This proposal focuses on building clinical decision support (CDS) tools for use at the point of care with Coronavirus Disease 2019 (COVID-19) positive patients to assist with the diagnosis and treatment of blood clots, medically referred to as thromboembolic events. We will leverage Northwell Health?s massive dataset of COVID-19 patient data ? one of the largest in the world ? to refine and validate two existing blood clot clinical prediction rules (CPRs) in the COVID population as this is of utmost public relevance with the current pandemic. We will then take the refined CPR models and implement them as integrated web apps that can be launched from within the flow of care in electronic health record software from multiple vendors, using open standards-based technology.
