Anthony DeCristofaro, Director, Center for Healthcare Quality and Analytics
Children's Hospital of Philadelphia
Children’s Hospital of Philadelphia improves its quality of care and patient outcomes through the use of data and advanced analytics. As part of an internal grant to streamline reminder systems to alert patients of upcoming appointments, a multidisciplinary team of family consultants, improvement advisors, providers, practice managers, statisticians and data scientists formed a project aimed to reduce missed clinical appointments. Missed appointments occur when a patient cancels with minimal lead time or does not arrive at a scheduled appointment (no-show). A no-show contributes to significant disruption for clinical staff and providers, hinders productivity and efficiency, underutilizes medical resources, and interrupts critical treatment for patients. The goal is to give staff the opportunity to have ample time to prepare for a missed clinical appointment.
To develop an analytics-based solution, a dataset using Electronic Heath Records data was created from 624,752 patients and 6,275,398 appointments. The complete dataset comprises 138 possible factors or variables. Using new technologies in the areas of machine learning, a tailored algorithm models the missed appointment risk and calculates the likelihood of patients attending or missing an appointment.
The appointment data collected from the Epic system is from January 1, 2013 until December 31, 2016 with four practice groups (primary care, surgical specialty care, medical specialty care). Of those patients, 79.61% were classified as no-shows who were correctly identified as such (sensitivity score). Additionally, the system has correctly classified patients who actually showed with AUC of 0.81, 0.82, and 0.83 for primary care, surgical care, and specialty care, respectively. The results of this model, in conjunction with the Epic system, have the opportunity to positively impact patients and providers because it proactively identifies barriers to attend appointments, intervenes prior to missed appointments, improves appointment adherence and follows up on clinical recommendations.