|
|
| |
| The Effective Management of Machine Learning Using Electronic Health Record Data for Earlier Diagnosis of Type 2 Diabetes
Download PDF
|
Katharine Blancett and Dr. Bruce Lazar, MBA, DM
|
Abstract:
Type 2 diabetes is a chronic disease that affects many people worldwide and, if left untreated, can lead to more severe health issues. Specific factors can increase a person's risk of developing type 2 diabetes; screening tests are essential to identify these factors. Machine learning can use electronic health record data to help detect risk factors earlier. The systematic literature review explored strategies that healthcare informatics leaders use to implement and effectively manage machine learning using electronic health record data for earlier diagnosis of type 2 diabetes. Medical Literature Analysis and Retrieval System and Cumulative Index to Nursing and Allied Health Literature Plus with full text were used to search for research-related articles to help answer the research question while using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses as a guide. Twenty relevant articles enabled results to answer the research question. The five primary themes that appeared from the literature analysis of the articles were artificial intelligence and machine learning have predictive capabilities to help manage type 2 diabetes; data mining can be used to find other diseases commonly found in patients with type 2 diabetes, machine learning can look at laboratory data to early screen for type 2 diabetes, machine learning can help develop more personalized treatments plans for patients with type 2 diabetes, and machine learning can look for medications already on the market that might help treat type 2 diabetes. All five themes used electronic health record data. The findings implied that machine learning could be a valuable tool to help diagnose type 2 diabetes early, but more research needs to be done to determine which specific data points will lead to the best results.
|
|
|
| | |