No Show Prediction at Outpatient Clinic - Observatory of Public Sector Innovation

This is a large-scale national initiative employing artificial intelligence techniques to proactively predict which patients are most likely to miss their appointments in outpatient clinics.
It uses machine learning techniques to process, analyse, and train data in the electronic medical record (EMR) system, encompassing patient, clinic, and appointment history.
The no-show rate at KAMC has decreased by 10%.

Innovation Summary

Innovation Overview

Within the Ministry of National Guard's Health Affairs (MNGHA) roadmap to employ Artificial Intelligence technologies to provide innovative health solutions, and improve the quality of services provided to patients and enhance the process of digital transformation in the field of health care. The department of Information Technology, in cooperation with the departments of Outpatient Clinics and Patient Services, presented a new experience that involves employing artificial intelligence to predict in advance which patients are most likely to miss an appointment in outpatient clinics. Outpatient clinics are one of the most important and largest channels in providing services to patients and visitors at MNGHA. They are considered the largest umbrella for medical specialties and the basic nucleus of medical services provided to employees of MNGHA and their families. In recent years, many tasks and plans have been successfully completed to improve performance by reconsidering the use of available resources and setting corrective measures to improve some areas, including patient no-show. Patients not attending their appointments are one of the main problems that affect the use of resources and pose risks to the quality of health care services provided.
The project is linked to the Kingdom of Saudi Arabia's strategy for digital transformation in the health sector, which contributes to providing high-quality health services to patients. The aim of this project is using artificial intelligence to develop a prediction model for predicting no-show visits using machine learning techniques based on the data in the electronic medical record. The application of machine learning techniques to data extracted from the electronic medical records (EMRs) of health care system provides a data‐driven approach to predictive No-Show in outpatients clinics. The developed predictive model embedded into BESTCare systems (the name of EMR system) for clinical practice. The emergence of advanced models, such as the use of machine learning techniques, allow risk prediction and offer promising results for improving clinical outcomes. The result showed decreasing in rate of No-Show by 10% after the implementation.
Overall, Ministry of National Guard Health Affairs has come a long way in its digital transformation journey, and the most important pillar of this transformation is the employment of advanced technologies such as Artificial Intelligence to raise the quality of health services and increase patient satisfaction. In addition, to improving operational efficiency and using health facility resources at a high level of effectiveness. MNGHA focus efforts on the field of Artificial Intelligence, and confirms the great role played in achieving the highest level of medical care and the level of maturity in its data and digital infrastructure.

Innovation Description

What Makes Your Project Innovative?

A new experiment involves the use of artificial intelligence to build a model predicting which patients are most likely to be absent from outpatient clinics, based on data from electronic medical records. To cope with increasing demand and compensate for patient no-shows, addressing patient absence is one of the most interesting and challenging tasks for healthcare providers. The project aims to add value in two main ways. First, it focuses on better management of hospital resources, including doctors and nurses, ensuring they provide care and attention to patients, which healthcare providers consider a fundamental pillar. Second, it aims to enhance the level of service by offering crucial information to support decision-making, an area with significant growth potential for the healthcare sector. All of this promotes more personalised care and complements professional decision-making.

What is the current status of your innovation?

After successfully implementing the prediction model and linking it with the electronic medical file system (BESTCare), we have achieved sufficiently accurate results to move on to the next level. In the fourth quarter of 2023, the Ministry of National Guard Health Affairs intends to operate the predictive model to cover all healthcare facilities in all regions, enabling care providers and patient services specialists to take appropriate action for patients expected not to attend. This stage required specific settings, ensuring these systems were designed with a scalable framework to accommodate future expansions across the Ministry of National Guard Health Affairs facilities throughout the Kingdom, considering the need for infrastructure that supports horizontal scaling and high availability.

Innovation Development

Collaborations & Partnerships

MNGHA has collaborated with several national entities to reach a digital maturity level that facilitates the implementation and utilization of smart digital solutions. National Data Management Office (NDMO) is one example of the active collaborations that allowed MNGHA to adopt the guidelines and standards of data governances leading towards 2030 vision.

Users, Stakeholders & Beneficiaries

• Patients and health care providers, including physicians and specialists in patient services and care
• Administrative positions and decision makers.
• Technologists who want to use data to achieve highly efficient business results and make improvement decisions through prediction.
• Scientists and scientific research teams.

Innovation Reflections

Results, Outcomes & Impacts

A real-world implementation of the no-show model reflects the prediction results in a meaningful way to support decision-making. The developed predictive model has been adopted in practice and embedded into the BESTCare system. The project achieved its optimal goal by decreasing the no-show rate by 10%. It is worth noting that ignoring a medical appointment costs the organisation substantial sums of money, amounting to an average of 1,315 Saudi riyals per appointment. Approximately 2,833,825 Saudi riyals were saved as a result of reducing wasted resources due to unused appointments and patient no-shows after implementing the model. Additionally, the access to care rate increased from 12.54% to 21.19%.

Challenges and Failures

Increasing the demand for outpatient services in Saudi Arabia, while the appointment booking system suffers from high no-show rates, is a challenging undertaking. Reducing no-shows is essential for healthcare providers to utilise resources effectively, decrease the financial burden, and provide timely treatment to patients who need care.

Healthcare providers explore various approaches to minimise no-show rates, employing a variety of strategies. However, these strategies often involve manual processes or difficult-to-enforce policies, resulting in a low impact.

Conditions for Success

Providing high quality of health care that is readily accessible, cost effective and meets the needs of the communities we serve. Provide our patients, their families and our community with extraordinary healthcare service; to ensure high quality, compassionate treatment; and to deliver care beyond their expectations. For this we developed a clear understanding of emerging technologies' potential, identifying the right use cases, building the necessary infrastructure and talent, and ensuring ethical and responsible use of AI.

Replication

The project can be replicated in any other facilities in different regions and countries and in different areas of healthcare. The project's scalability is testified by the Saudi experience itself, in which the project has been, and is planned to, successively implemented in more hospitals.

Lessons Learned

Data-driven organizations have the ability to build data strategies and implement data analysis to maintain the highest standards of performance.
Use of Emerging Technologies (AI and ML algorithms) are useful with Problems, which are difficult to solve with traditional methods. It has proven to be a superior prediction of no-show. Thus, effective optimizing of health resource usage.
Predict the likelihood of no-shows will guide the decision to implement more reliable strategies for scheduling appointments and better care.
The use of machine learning techniques represents an important opportunity for healthcare providers to impact healthcare outcomes by enhancing advanced analytical capabilities.