Predictive Analytics Deployment to Forecast Infectious Disease, Care Needs
Highmark Health hopes to predict staffing needs, bed capacity, and healthcare utilization using Kinsa’s AI-based illness forecast tool.
Source: Getty Images
By - Pennsylvania-based Highmark Health will deploy an artificial intelligence (AI)-based early warning system from insights solutions company Kinsa to help predict healthcare utilization, staffing needs, and bed capacity during infectious disease spikes.
According to the press release, the tool is designed to forecast the spread of existing infectious diseases weeks or months in advance and the emergence of novel outbreaks up to two weeks before they occur using a combination of data related to symptom prevalence, geography, and illness spread. Using the technology, health systems can leverage these insights to predict care needs following an outbreak.
"Kinsa possesses a unique combination of expertise in both data science and epidemiology," said Richard Clarke, PhD, senior vice president and chief analytics officer of Highmark Health, in the press release. "This blended approach of artificial intelligence and medicine not only informs when and where illness will spread, but also helps organizations like Highmark Health to reduce clinical burn-out and accurately anticipate the needs of our customers so we can provide a remarkable health care experience."
Allegheny Health Network (AHN), Highmark Health’s integrated delivery network, will deploy the early warning tool to help model staffing needs and bed capacity in the emergency department (ED) and intensive care unit (ICU).
"In the current health care labor environment, anticipating our staffing needs is critical to our ability to provide optimal care for our patients," said Brian Parker, MD, chief quality and learning officer at AHN, in the press release. "Kinsa's real-time respiratory illness data and predictive modeling capabilities will help us to stay weeks ahead of the curve when it comes to potential viral surges, and near-future patient flow. We can be proactive, instead of reactive, when it comes to preparing for both potential ED volumes and making changes to our operations."
He further explained that forecasting the need for these changes in operations is critical in the event of a local infectious disease surge because hospital staff could get sick as ED and ICU volumes rise, resulting in fluctuations in staffing levels.
"When caregivers get sick, we have to shift our personnel," said Parker. "Knowing which communities will be affected first will allow us to allocate our resources in a much more strategic way."
Efforts to more effectively monitor infectious disease outbreaks have seen renewed interest in the wake of the COVID-19 pandemic. Some public health stakeholders have turned to AI and data analytics to fill in the gaps and enhance monitoring.
In April, the Centers for Disease Control and Prevention (CDC) announced an extension of its 10-year partnership with software company Palantir Technologies Inc. to address outbreak response and disease surveillance through the Data Collation and Integration for Public Health Event Response (DCIPHER) Program.
DCIPHER is a cloud-based platform used by public health organizations to help facilitate data interpretation and inform public health decisions. Under the extended partnership, groups using DCIPHER can leverage Palantir's Foundry platform to support additional data integration, conduct analysis, and create operational workflows for public health needs.
