Use of Predictive and Simulation Models to Develop Strategies for Better Access Specialists Care

Use of Predictive and Simulation Models to Develop Strategies for Better Access Specialists Care

Siang Li Chua (Changi General Hospital, Singapore) and Wai Leng Chow (Changi General Hospital, Singapore)
Copyright: © 2020 |Pages: 28
DOI: 10.4018/978-1-7998-0047-7.ch007

Abstract

No-shows are patients who miss scheduled Specialist Outpatient Clinic (SOC) appointments. No-shows can impact patients' access to care and appointment lead time. This chapter describes a data-driven strategy of improving access to specialist care through first developing a stratified predictive scoring model to identify patients at risk of no-shows; second, studying the impact of a dynamic overbooking strategy that incorporates the use of the no-show prediction model using discrete event simulation (DES) on lead time. Seventeen variables related to new SOC appointments for subsidized patients in 2016 were analyzed. Multiple logistic regression (MLR) found eight variables independently associated with no-shows with area under receiver operation curve (AUC) 70%. The model was tested and validated. DES model simulated the appointment overbooking strategy as applied to the top highest volume specialties and concluded that lead time of Specialty 1 and 2 can be shortened by 27.5 days (49% improvement) and 21.3 (33%) respectively.
Chapter Preview
Top

Background

Predicting No-Show

Many studies concurr that a no-show does not occur arbitrarily and instead, aimed to predict it. Almost all no-show studies had analyzed the relationship between no-shows and appointment records through univariate and bivariate analysis. Multiple Logistic Regression (MLR) is the most commonly used predictive method (56 out of 105 studies) for this type of analysis (Dantas et al., 2018). Other statistical and predictive methods used include Gradient Boosting Algorithm (Elvira, Ochoa, Gonzalvez, & Mochón, 2018), stochastic programming (Samorani & LaGanga, 2011), database exploration technique (Samorani & LaGanga, 2015) and association rule mining (Glowacka, Henry, & May, 2009). Hybrid methodologies such as empirical Markov modeling with MLR (Goffman et al., 2017) and empirical Bayesian inference with MLR (Alaeddini, Yang, Reddy, & Yu, 2011) have also been used.

Key Terms in this Chapter

CHAS: Community Health Assist Scheme is a scheme by the Ministry of Health that enables Singapore citizens from lower-to-middle income households to receive subsidies for medical and dental care at participant General Practitioner and dental clinic near their home

Specialist Outpatient Clinic (SOC) operating hour: 8.30am to 6pm on Monday to Friday

Regional Health System (RHS): RHS were set up to strengthen partnerships across care providers in each region: both public and private, and across care settings. Each RHS is anchored by a Public hospital working closely with community hospitals, nursing homes, home care and day rehabilitation providers, as well as polyclinics and private General Practitioners (GPs) within the geographical region. The RHS functions as a coordination platform to organize service providers. It delivers patient-centric care through the integration of services and processes. Such integration helps patients navigate across providers within the same region more easily, enabling them to manage their own care needs ( Agency for Integrated Care, 2016 ).

Outpatient Appointment System: (OAS): OAS is a web-based system to manage outpatient appointments. The system allows administrators to set up the service providers, calendar, appointment slot; call center operators to search for earliest appointment and book appointments; and managers to query and generate reports.

Visit Status: There 4 types of visit statuses: Actualized, No-Show, Cancelled and Walk-ins. Actualized refers to a patient who attended the scheduled appointments. No-show, similar to non-attendance and missed appointments, refers to a patient who missed the schedule appointment without cancellation notice to the service provider. An appointment can be cancelled by either doctor or patient. Walk-in refers to a patient who is served without appointment.

Appointment Lead Time: Duration in days, from an appointment request till appointment date.

Visit Type: There are 4 visit types: New Case (NC), Follow-up (FP), Ancillary New Case, and Ancillary Follow-up. The former 2 types required doctor services while the latter 2 types require non-doctor services.

SAF: Singapore Armed Forces

Appointment Center Operating Hour: 8.30am to 8pm on Monday to Friday, 8.30am to 12.30pm on Saturday and Sunday. Closed on Public Holidays.

Polyclinics: Subsidized primary care outpatient clinics fulfilled by the government.

Complete Chapter List

Search this Book:
Reset