The PHC appointment management enigma: Applying Real-time analytics and Artificial Intelligence to solve the case. (Preprint)

Abstract

Primary healthcare (PHC) services are strained by high patient volumes, leading to complex operations that require efficient management. Patients access services through booked appointments and walk-in visits. However, walk-in visits face longer waits due to prioritization of booked visits, compounded by a high rate of appointment no-shows (3-14% revenue loss) that disrupt resource allocation. These no-shows negatively impact operational efficiency and healthcare quality. At EHS PHC centers, handle over 140,000 visits monthly. At the baseline of our project, patients had an average wait time longer than 16 minutes, exacerbated by a 21% no-show rate for appointments. This situation demanded an efficient scheduling and resource management system to enhance patient experiences and operational efficiency.

- To Provide real-time actionable insights to all primary health center directors and administrators, including traffic overview, human resource distribution, and risk prediction of appointment no-shows within the first month of project implementation. -To Identify and address operational bottlenecks to streamline appointment management processes, aiming for a measurable improvement in efficiency within the first three months of project implementation. - To Reduce appointment no-shows by at least 20% in the first three months after project implementation through predictive analytics and targeted communication strategies. -To Reduce overall patient waiting times by at least 10% across all stations within the first three months after project implementation, utilizing real-time data and operational adjustments.

This project introduces a groundbreaking approach to enhancing primary healthcare efficiency by employing an interactive data application with an AI Solution aimed at reducing clinic wait times and appointment no-shows. Utilizing our Electronic Health Record system as the pivot for our innovation, we introduced an interactive data application for our PHC decision-makers. This application includes an artificial intelligence model with an 86% accuracy rate that can predict no-shows by analyzing historical data and categorizing appointments based on risk. It features a real-time dashboard for monitoring patient journeys and wait times across clinical stations alongside the AI's no-show predictions. Clinic coordinators use this dashboard to manage high-risk no-show appointments and optimize scheduling proactively. We studied the before and after statistics of PHC appointment dynamics to ascertain the impact of this data-driven, proactive program.

Our analysis revealed that implementing the No Show AI model significantly contributed to a 50.7% reduction in overall no-show rates within the Primary Health Care Department. This outcome, observed at the organizational level, achieved statistical significance with a p-value of less than 0.001. The Odds Ratio (OR) of 0.43 (95% confidence interval: 0.42 to 0.45) with a p-value of <0.001 indicates a strong effect of the intervention. We observed reduced average waiting times at all stations across the PHC journey. The reduction in waiting time from check-in to checkout has been significant after implementing this solution. We found a decrease in patient wait times of an overall average of 5.7% (from 54 minutes to 48 minutes) (p=0.00) and a maximum of 50% reduction in best-performing PHCs

The project significantly improves operational efficiency and patient satisfaction by conducting daily wait time assessments and making necessary adjustments, such as reallocating patients to different clinicians. It illustrates the impactful role of an integrated Electronic Health Record system and Artificial Intelligence in transforming primary healthcare delivery.