Smarter scheduling, shorter waits: New research aims to transform how Canadian health organizations manage capacity

Across Canada, health organizations routinely struggle to meet mandated wait time targets, the benchmarks that define timely access to care for patients. Behind every missed target lies a co-ordination problem. Decisions about how many resources to allocate, when to schedule patients and how to meet varying demand are typically made separately, even though they’re deeply interconnected. This is why Professor Jonathan Patrick has received a Natural Sciences and Engineering Research Council Discovery Grant for his project, “Scheduling and Capacity Planning Using Approximate

Despite growing investment in health-care technology and data systems, the tools to manage patient flow and allocate clinical resources have trailed behind. Most scheduling approaches rely on static models that can’t adapt to shifting demand, variable service times or domino effects of decisions made across interconnected services. This leaves health organizations constantly reactive rather than proactive.

A unified framework for scheduling and capacity planning

Current approaches to patient scheduling and capacity planning fall short because they treat these decisions separately, leaving health systems ill-equipped to adapt to the uncertainty and complexity of real-world care. Patrick is developing a unified patient scheduling and capacity planning approach. This framework links capacity allocation, scheduling decisions and wait time targets in one integrated model.

At its core, the research uses approximate dynamic programming, a computational technique well-suited to the kind of sequential, uncertain decision-making in health-care operations. Key innovations include making scheduling models more responsive to changing conditions and incorporating machine learning and neural networks to handle the non-linear complexity of capacity planning across networks of care.

Why this research matters

The implications of this work extend well beyond a single clinic or methodology. For the academic community, it advances operations research by combining approximate dynamic programming with machine learning to better reflect the complexity of real health systems. This will create new knowledge that blends optimization, artificial intelligence and health-care management.

For practitioners and policymakers, the payoff is more immediate. Health administrators will gain data-driven tools to allocate resources more efficiently, reduce bottlenecks and ensure patients receive timely care. In a country where wait times are a policy priority and a persistent challenge, research that meaningfully improves how health systems plan and schedule care could benefit patients, providers and the public alike.