• Politics
  • Diversity, equity and inclusion
  • Financial Decision Making
  • Telehealth
  • Patient Experience
  • Leadership
  • Point of Care Tools
  • Product Solutions
  • Management
  • Technology
  • Healthcare Transformation
  • Data + Technology
  • Safer Hospitals
  • Business
  • Providers in Practice
  • Mergers and Acquisitions
  • AI & Data Analytics
  • Cybersecurity
  • Interoperability & EHRs
  • Medical Devices
  • Pop Health Tech
  • Precision Medicine
  • Virtual Care
  • Health equity

AI Model Analyzes EHR Data, Identifies Child Leukemia Patients for Studies


The data could help researchers design real-world studies in patients with a range of different cancers.

data points

A computable phenotype analyzed electronic health record (EHR) data and accurately classified patients with pediatric leukemia or lymphoma, according to the findings of a study published in the journal Pediatric Blood and Cancer.

The algorithm showed 100% sensitivity and 99 to 100% specificity, investigators found.

“Accurately identifying patient cohorts is key to designing better research,” said study lead Charles Phillips, M.D., a pediatric oncologist at Children’s Hospital of Philadelphia (CHOP). “Because not every patient in large data sets would be appropriate for a clinical study, having a tool to separate signals from the noise will help researchers leverage data to design pragmatic, real-world studies in patients with a range of different cancers.”

Researchers could use the data to better evaluate nausea medicines or to detect factors that influence infections, he added.

The research team analyzed EHR-derived data in PEDSnet, a national pediatric clinical research network, to develop and evaluate a computable phenotype algorithm. They built the algorithm to identify patients with pediatric leukemia and lymphoma who received treatment with chemotherapy.

The data came from three large pediatric hospital systems: CHOP, Children’s Hospital Colorado and Seattle Children’s Hospital. Data included diagnoses, procedures, medications, laboratory tests and provider specialties.

The researchers created a computable phenotype that automated their search algorithm to check off a series of boxes. First, it checked off whether a patient had at least three visits to a pediatric hematologist-oncologist (27,450). Then it checked whether the patients had at least one leukemia or lymphoma diagnosis, which cut the number of patients to 4,535.

Further screening required three specialist visits, at least two diagnostic codes and at least two administrations of chemotherapy. This cut the number of patients to 1,825.

That group of patients became the computable phenotype curated cohort.

After analyzing the medical records of the cohort in a masked review, the computable phenotype showed 100% sensitivity and 99 to 100% specificity in accurately classifying the patients as having leukemia or lymphoma.

“This algorithm can accurately and efficiently narrow down the number of medical charts researchers need to review to identify a patient cohort for subsequent clinical studies,” Phillips said.

Additional studies are necessary to refine the algorithm to meet study-specific needs. But Phillips said “it offers a potential new tool to clinical researchers in improving outcomes for children with leukemia or lymphoma, who represent about 40% of all U.S. pediatric cancers.”

Get the best insights in digital health directly to your inbox.


AI Model Passively Detects Cardiac Arrest through Smart Speakers

Adventist Health to Integrate Cancer Risk Tool Directly into EHR

Machine Learning Accurate in Identifying Patients at Risk of Developing Psychosis

Related Videos
Image: Ron Southwick, Chief Healthcare Executive
George Van Antwerp, MBA
Edmondo Robinson, MD
Craig Newman
Related Content
© 2024 MJH Life Sciences

All rights reserved.