Effective resource management using machine learning in medicine: an applied example

Area of Science:

• Healthcare Informatics
• Applied Machine Learning
• Health Services Research

Background:

• Healthcare is increasingly digital, generating vast amounts of data.
• Adoption of machine learning and data analytics in daily healthcare delivery has been slow.
• This study presents machine learning application to optimize laboratory testing programs.

Purpose of the Study:

• To demonstrate the benefits of machine learning in healthcare.
• To optimize a laboratory testing program for community patient sample processing.
• To showcase practical collaboration between clinicians and machine learning engineers.

Main Methods:

• Developed a prototype transport scheduling platform using machine learning techniques.
• Simulated the platform's efficiency and cost impact using historical data from Canterbury District Health Board.
• Focused on urgent lab sample processing in the community to reduce hospital emergency presentations.

Main Results:

• The simulation demonstrated procedural efficiencies.
• Potential for annual cost savings between 5% and 14% was identified.
• Key advantages included a forward job list for the laboratory, expected time-to-result, and a streamlined transport request process.

Conclusions:

• Healthcare has numerous opportunities for improving care delivery using large datasets.
• Machine learning techniques can significantly improve the efficiency of community patient lab sample processing.
• This work exemplifies practical applications of machine learning in healthcare settings.