New York state in late May had more than 300,000 cases of the coronavirus. Health officials, with little or no data about what to expect in terms of hospitalizations could have been overwhelmed in a number of ways with all the unknowns about what to expect from patient admissions and how they would respond to treatment.
Health officials deployed a platform to monitor, aggregate, measure, analyze and report data on a continuous basis during the crisis to help understand and manage patient inflow and hospitalization.
The platform, the Cloudticity Healthcare DataHub built on Amazon Web Services, allows users to see clinical data in realtime to determine trends and manage inventories, such as ventilators and ICU beds and discover via machine learning, correlations like co-morbidities, demographics, etc.
The platform was developed for the health-care industry and thrust into deployment when the coronavirus pandemic hit.
“We initially developed this product for a very large hospital system with on premises data link that was far too expensive to maintain,” said Cloudticity CEO Gerry Miller. “We quickly retooled it to focus on COVID cases, and in six days from the first call, we had all the contracts signed and had deployed an operational data lake.”
That was in New York City, which was being overrun by the coronavirus. “When the pandemic began, there wasn’t a lot of information about what was going on for hospitalizations,” Jim Kirkwood, division director of Healthcare Innovation at the New York Department of Health, said in a webinar. “There wasn’t a lot of information on how long people would be staying, which turned out to be very long in a lot of cases.”
The Department of Health wanted to understand frequency and length of hospitalization, and collect information on comorbidities, symptoms and clinical history of patients and needed a more automated process than they had to do it.
They needed clinical lab information about admissions, discharges and transfers as well as the observational results. The department also needed a single data store for positive patients. “One of the questions we’re looking at, the big one, of course, is the frequency of hospitalization that’s going on in the state and really understand what’s going on there,” Kirkwood said. “And understanding where folks are going in the hospital and where within the state.”
The Healthcare DataHub has no bandwidth limitations and can ingest a huge amount of data from disparate sources and customize that data for analysis, Miller said.
One of the results was a diversion of the workforce where needed. The data analysis helps health-care officials understand which patients, even those without the coronavirus, are likely to need a case worker upon admission instead of discharge to see them through the whole process and reduce the number of patients that are readmitted.
The product was initially developed to better understand the readmission process and which patients were being readmitted and for revenue recapture. If a Medicare patient has to be readmitted within 30 days of the original hospitalization that hospital won’t be reimbursed.