An application of computable biomedical knowledge to transform patient centered scheduling

Abstract

Introduction

Efficient appointment scheduling in the outpatient setting is challenged by two main factors: variability and uncertainty leading to undesirable wait times for patients or physician overtime, and events such as no-shows, cancellations, or walk-ins can result in physician idle time and under-utilization of resources. Some methods have been developed to optimize scheduling and minimize wait and idle times in the inpatient setting but are limited in the outpatient setting.

Methods

People and Organization Development, an internal group of organizational developers, led the development of a solution that selects the optimal group of appointments for a patient that minimizes the time between associated procedures as well as lead time built using a linear integer program. This program takes appointment requests, availability of resources, order constraints, and time preferences as inputs, and provides a list of the most optimal groupings as an output. Included in the methodology is the technical infrastructure necessary to deploy this within an electronic medical record system.

Implementation and Test Plan

A pilot has been designed to run this algorithm in a single department. The pilot will include training staff on the new workflow, and conducting informal interviews to gather qualitative data on performance. Key performance indicators such as schedule utilization, resource idle time, patient satisfaction, average appointment lead time, and average waiting time will be closely monitored.

Discussion

The model is limited in accounting for variability in appointment length potentially resulting in inaccurate schedules for healthcare providers and patients. Future states would incorporate certain visit types starting through machine learning techniques. Additionally expanding our data pipeline and processing, developing greater communication software, and expanding our research to include other departments and subspecialties, will enhance the accuracy and flexibility of the algorithm and enable healthcare providers to provide better care to their patients.