Risk Factors Associated with Hospital Unwarned Appointment Absenteeism: A logistic binary regression approach

Abstract

One of the main problems faced by health institutions is the unwarned absenteeism of patients in medical appointments. Patients’ no-shows, without prior notice, can result in loss of revenue for health centres and increasing waiting lines. Hence, there is a need to predict the non-attendance of patients to improve health institutions’ management performance. In this paper, a brief literature review was carried out to understand which factors can be related to patients’ absenteeism, and which forecasting methods are often applied to discover patterns in health datasets. As the logistic binary regression model has been proved to be effective on that matter, it was applied to a real hospital data set comprising information on 98.511 patients, with a corresponding 645.576 appointments, in a period between 2018 and 2020. Results indicate a significant effect on the chance of appointment attendance of patient age, patient gender, patient Marital Status, number of previous appointments, appointment month, precipitation levels, Lead time, and the number of previous no-show appointments.