Integration of automated predictive analytics into electronic health records: can spine surgery applications lead the way using SMART on FHIR and CDS Hooks?

Abstract We are increasingly moving towards precision and personalized medicine for spine surgery. Predictive models have been developed to assist clinicians in predicting a patient's survival, discharge disposition and infection following surgery. However, the developed algorithms generally require the use of stand-alone tools that can be time-consuming and cumbersome to use. Incorporating these algorithms into the electronic health record and clinical workflows will therefore be critical for unlocking these algorithms’ potential. This review summarizes the current state of predictive algorithms in spine surgery and proposes a way forward for optimally integrating them into clinical care. The proposed improvements could also be translated to other subspecialties in orthopedics and neurosurgery.