Sharpening clinical decision support with machine learning
Artificial intelligence (AI) and machine learning hold promise for the future of health IT. AI today ingests, processes and acts upon information intelligently, to automate and complete tasks; machine learning facilitates continuous adaptation and evolution of algorithms and processes, as a human might learn from repeatedly completing activities.
There are countless applications for these technologies within medicine, but one of the most exciting areas is how machine learning can transform clinical decision support from systems laden with meaningless alerts to intelligent workflows and precision best practices driven by relevant patient history.
Algorithms enable more timely, practical and precise clinical decision support
Clinical decision support is based on evidence and research. It’s often hard coded into systems to deliver more standardized, best-practice care for patients. Today, clinical decision support embedded in the electronic health record (EHR) will remind clinicians to record height and weight to calculate BMI, or to perform medication reconciliation at each office visit.
Best practices should change with evidence-based literature, and it can take a long time for new findings to affect patients. Sadly, some estimates suggest that it can take up to 20 years before research findings become part of widely accepted clinical practice.
Another challenge clinicians face in adoption of clinical decision support is the failure to deliver guidelines within unique context of each patient. Patient factors – such as allergies, comorbidities or interaction of drugs – might signal that some of the recommended items within the guidelines won’t be effective.
Each patient needs something a little different. For instance, guidelines recommend 30 minutes of aerobic exercise five days a week to patients, but the plan for a patient who has multiple comorbidities – such as individual with angina and Type 2 diabetes in addition to COPD – may need tailoring, taking into account physiologic markers for safe exercise tolerance.
Machine learning can enable clinical decision support based on multi-system analysis to understand which patients are at highest risk of a negative outcome, or to optimize treatment in real-time. Algorithms can parse available historical and current information to inform clinicians which patients are at risk for specific outcomes, or even adjust treatment regiments according to most up-to-the-minute changes.
For example, caregivers can fine-tune or optimize insulin pumps according to historical patterns for each diabetic patient. Machine learning can ultimately enable clinicians to develop more precise, personalized and effective care plans, and do so earlier on the disease progression timeline.