Health systems around the nation are using EHR data to gain valuable insights into treating COVID-19 patients and develop predictive models about what their organization will need to combat the pandemic in the future.
New Hyde Park, N.Y.-based Northwell Health, at the epicenter of the pandemic in New York, was able to analyze medical record data for patients hospitalized in the state's health system. The Northwell Health COVID-19 Research Consortium, with support from the Feinstein Institute for Medical Research, examined the outcomes of 5,700 Northwell patients treated for COVID-19 and the 2,634 who were hospitalized.
The system collected patient demographic information, home medications, triage visits, initial lab tests, initial electrocardiogram results and diagnoses during hospitalization. The researchers also examined data on treatment and outcomes that was available in the EHR. Insights from the research:
· 88% of the patients on medical ventilation died
· 21% of all patients treated died
· 14% of the patients spent time in the ICU
· 12% of patients were put on ventilators
· 3% of patients received kidney therapy
· Patients were discharged after four days on average
The health system developed a Journal of the American Medical Association paper with the findings and outlined methodology there to help other organizations track data of their populations.
Cleveland Clinic, in partnership with SAS, used its data to create a predictive model that forecasts patient volume and supply availability. The analytic models are designed for healthcare organizations to create the best-case, worst-case and most-likely scenarios, which can be adjusted in real time.
The models allow hospitals to make decisions on ICU beds, personal protective equipment and ventilators. Cleveland Clinic used the model to project a worst-case scenario and decided to build a 1,000-bed surge hospital on its education campus to support COVID-19 patients that didn't need an ICU.
Parkview Health, along with several other hospitals with Epic EHR, is using Epic's artificial intelligence system to predict when patients with the coronavirus will become critically ill, according to a STAT report. The health system is analyzing data from its nearly 100 cases and realized that 75 percent of the patients hospitalized with a score of 38 to 55 on the nursing assessment scale of zero to 100 spent time in the ICU. As a result, the system was able to manage resources and transfer at-risk patients to larger hospitals.
UChicago Medicine implemented a virtual screening questionnaire through its EHR to help patients who are concerned they have symptoms of COVID-19. The screening categorizes patients into three categories, and from there, UChicago Medicine can determine which patients need to be tested.
Worcester-based UMass Memorial Medical Center deployed a new function in its EHR system that allows clinicians to message medical questions to an infection control provider. Messages that providers send can be attached to patient names, which allows the on-call provider to access the patient's medical record and help provide treatment recommendations.
More articles on healthcare:
Google Cloud aims to boost health system interoperability, data analytics with API release: 5 details
The difference-makers during the surge: 6 CIOs on top priorities and key advice amid COVID-19
5 ways CIOs are preparing for teams to return from remote work