The Agency for Healthcare Research and Quality (AHRQ) has launched a challenge competition focused on using predictive analytics to forecast healthcare utilization.
The challenge, which has a total prize pool of $225,000, aims to develop predictive analytics to estimate hospital inpatient utilization, including certain outcomes such as length of stay.
Traditional approaches in health services research rely on rigorous methods, the availability of recent data, and peer review to assure the highest quality of analyses, according to AHRQ. However, some decision-makers must make policy decisions quickly with the current information available and can’t wait for traditional research methods, said AHRQ Director Gopal Khanna in a video about the challenge.
“Predictive analysis and innovative use of data may offer a solution in those cases,” he said, by balancing the need for rapid information and academic rigor.
While other industries use predictive analytics and related methods successfully, the healthcare industry is lagging behind, he said.
The Centers for Medicare & Medicaid Services (CMS) announced on Wednesday details about a new Artificial Intelligence Health Outcomes Challenge with the potential for projects to win up to $1.65 million. It was created in partnership with the American Academy of Family Physicians and the Laura and John Arnold Foundation.
AHRQ predictive analytics challenge participants are encouraged to use data and predictive analytics to estimate hospital inpatient utilization and average lengths of stay in selected counties for the year 2017. AHRQ will provide applicants, who have executed the data use agreement required for participation in the challenge, access to customized analytic files that include information on hospital inpatient discharges for years 2011 to 2016.
According to AHRQ, applications will be evaluated based on reliability and validity. In general, reliability is assessed by how closely the model or method deployed predicts the actual utilization rates for 2017. Validity is assessed by how well the model performs on earlier years of data, AHRQ said.
The challenge is open to participants who apply independently or team with others, including health services and social science researchers, health IT developers and healthcare providers.
Applications for the challenge are due June 28 and winners are expected to be announced July 31, 2019.