How can health systems integrate predictive analytics into the everyday delivery of patient care? Harvard Business Review offer some strategies.
Here are four suggestions:
1. Figure out the clinical decision an algorithm will inform. Many algorithms will predict a patient's risk of readmission, for example, but providers want to know specific questions within that — when can I discharge my patient? Will my patient need a home care visit?
2. Use electronic health record data. Algorithms will only be helpful if backed by a strong layer of data. More clinical data will deliver better algorithm accuracy and potential for good outcomes.
3. Zone in on low-value decision points. Predictive analytics may allow physicians to only use expensive interventions for high-risk patients who can benefit.
4. Integrate analytics into workflow. If the algorithms don't fit seamlessly into existing workflows, physicians may be more likely to ignore the information.