Category: Diabetes/Prediabetes/Hypoglycemia

Monitor: 19

19 - DEVELOPMENT AND VALIDATION OF A MACHINE LEARNING MODEL TO PREDICT DIABETES MELLITUS DIAGNOSES IN A MULTI-SPECIALTY CLINICAL SETTING

Thursday, Apr 25
1:00 PM – 1:30 PM

Objective :

The prevalence of undiagnosed type 2 diabetes mellitus is high, with estimates of over 7 million patients in the United States and nearly 175 million cases globally.  This study evaluated a machine-learning model for screening electronic health records (EHR) data to identify potential patients with undiagnosed type 2 diabetes mellitus. 


Methods :

A supervised jungle binary classifier machine-learning model was created using the most recently reported de-identified patient data, excluding glycemic measures, sourced from a large multi-specialty clinic’s EHR.  Patient data was segmented (50% spit) into training and validation datasets.  The training dataset was used to train the model to identify patients at high risk of type 2 diabetes mellitus (T2DM) from nine EHR measures: 1. Age, 2. Gender, 3. Race, 4. Body Mass Index, 5. Blood Pressure, 6. Creatinine, 7. Triglycerides, 8. Family History of Diabetes, and 9. Tobacco Use.  Using a two-class decision jungle algorithm, each patient was assessed against a binary diagnosis scenario (diabetic or not diabetic), with diagnosis of T2DM defined as a random glucose > 140 mg/dL and/or HbA1c > 6.5%. The validation set was then used to determine the model’s predictive accuracy.


Results :

The initial sample size totaled 618,022 subjects. Incomplete data resulted in exclusion of 532,303 subjects. The remaining 85,719 subjects were segmented into equal training and validation datasets. After training the model on the training dataset, T2DM was identified in the validation dataset with positive predictive value (precision) of 0.686 and negative predictive value (recall) of 0.65.  Area under the curve (AUC) and F-score for predictive accuracy were 0.72 and 0.77, respectively.     


Discussion :

This study demonstrates the feasibility of using a machine-learning assessment tool to identify patients at high risk of T2DM from evaluation of demographic, clinical, and non-glycemic laboratory parameters commonly found in electronic health records.  After additional work to refine the model and improve predictive accuracy, this machine-learning tool may prove valuable in helping clinical practices screen large patient populations to identify individuals in need of diabetes screening. 


Conclusion :

Initial results indicate that a jungle binary classifier machine-learning model can be developed to create a screening tool to accurately identify patients at high risk of undiagnosed T2DM who would benefit from glycemic screening.   

SHORT URL FUNCTION-->

Chaitanya Mamillapalli

Endocrinologist
SPRINGFIELD CLINIC
Springfield, Illinois

Endocrinologist, Director for the DIabetes learning center and Bone health center at Springfield clinic.
Dr.Mamillapalli is passionate about improving patient outcomes, by developing healthcare technology software and apps to help providers in making the best decisions based on the up to date available medical evidence.

Shaun C. Tonstad

President
Clarion Group, Inc.
Springfield, Illinois

Shaun Tonstad is president of Clarion Group, Inc., a technology and software development consultancy. Among his responsibilities, Mr. Tonstad oversees development of consumer-oriented healthcare products, including several which achieved successful venture rounds and exits. His areas of interest include cloud computing, machine learning, and mixed reality computing. Mr. Tonstad has received top awards in international software development competitions, including Microsoft Dream.Build.Play and the Connected Systems Developer Competition. He holds a B.S. in Computer Science from Colorado State University.

Daniel J. Fox

Director of Clinical Research
Director
Springfield, Illinois

Dr. Fox is a translational scientist dedicated to high-quality and efficient "bench-to-bedside" research. For over 15 years, he has worked in highly-regulated academic, industrial, and military research environments.

Geoffrey W. Rutledge

Chief Medical Officer
HealthTap, California

Geoffrey W Rutledge, MD PhD Chief Medical Officer, HealthTap
Dr Rutledge is a double-board certified physician who practiced and taught medicine for more than 25 years. He earned a PhD in medical computer science from Stanford, was an NIH-funded researcher, and served on faculty at Harvard, Stanford, and UCSD medical schools. Before HealthTap, he created the first consumer health website and PHR at Healtheon/WebMD, was SVP of clinical transformation at First Consulting Group, CMIO at San Mateo Medical Center, and EVP, Product Development and Chief Medical Officer at Epocrates.
Geoff is an avid pilot of hang gliders and experimental aircraft, SCUBA diver, and photographer. He also enjoys bicycling and electric unicycling.

David B. Graham

Senior Vice President and Chief Information Officer/Chief Medical Information Officer
Memorial Health System
Springfield, Illinois

David B. Graham, MD, Senior Vice President and Chief Information Officer/Chief Medical Information Officer of the Memorial Health System, and Clinical Associate Professor, Family & Community Medicine for Southern Illinois University School of Medicine. Dr. Graham is an experienced clinician and informaticist with over 24 years of experience. He continues active clinical practice and teaching as clinical associate professor for SIUSM’s Department of Family and Community Medicine. Dr. Graham is Board Certified in Family Medicine and Clinical Informatics.

Dr. Graham provides strategic and operational leadership for all facets of information technology and clinical informatics. He joined the Memorial in January 2008 as the Chief Medical Information Officer and expanded his duties to his current role in January 2009. Dr. Graham’s interests emphasize process improvement and high quality, patient-centered healthcare through information technology systems and support. He serves the dual role of Chief Medical Information Officer and Chief Information Officer in work with physicians, residents, nurses, and all clinical units as the leader for information technology development and implementation.

Prior to joining Memorial and SIU, Dr. Graham graduated from the Feinberg School of Medicine at Northwestern University in 1995 and completed his Family Medicine Residency at the Oregon Health and Sciences University in 1998. He emphasized clinical informatics, practice-based research and curriculum design in his Faculty Development Fellowship at the University of Washington in 1999. Dr. Graham practiced in rural Eastern Oregon for several years prior to joining the faculty at the University of Colorado in 2004.

Michael Jakoby

Associate Professor of Medicine and Chief
Division of Endocrinology, SIU School of Medicine
Springfield, Illinois

Dr. Jakoby is the Division Chief and an Associate Professor of Medicine in the SIU School of Medicine Division of Endocrinology.

Chaitanya Mamillapalli

Endocrinologist
SPRINGFIELD CLINIC
Springfield, Illinois

Endocrinologist, Director for the DIabetes learning center and Bone health center at Springfield clinic.
Dr.Mamillapalli is passionate about improving patient outcomes, by developing healthcare technology software and apps to help providers in making the best decisions based on the up to date available medical evidence.