KFSH doubles its AI-supported emergency demand forecast to 24 hours

King Faisal Specialist Hospital and Research Centre says it can now forecast emergency department demand 24 hours ahead, giving its teams a longer window to prepare staff and capacity. In an October 21 announcement, the Saudi hospital also reported a roughly 42% reduction in the time patients wait from arrival to physician assessment.

More visits, coordinated responses

The operational backdrop is rising demand. KFSH reports that average daily emergency department visits increased by approximately 8.5% compared with 2022. It says patient-flow measures improved despite that increase, placing the new forecasting capability within a wider effort to manage patients’ movement through emergency care rather than a standalone AI deployment.

That effort runs through the Crisis Management and Patient Flow and Capacity Command Centre. The centre brings operational teams together with real-time oversight, data-driven decisions and an expanding use of AI. Since its launch, its team has carried out more than 300,000 interventions to monitor and support emergency department operations, according to the hospital.

The gains span several waiting points

One of the hospital’s largest reported improvements predates the latest announcement. Median boarding time—the wait for an inpatient bed after admission—fell from more than 13 hours in 2023 to 3.2 hours by mid-2025. KFSH describes that as a 75% reduction. This measures a different part of the emergency-care journey from the wait to first see a physician.

The hospital also reports improvements at earlier stages and across the full emergency department stay. Its figures distinguish the initial arrival-to-triage interval from the subsequent wait for physician assessment, rather than treating every delay as the same measure:

• Arrival to triage: average time improved by approximately 24%.
• Triage to physician assessment: time improved by 47%.
• Total emergency department length of stay: improved by around 30.5%.

The roughly 42% reduction covers the combined interval from arrival to physician assessment. It is not the boarding-time result or the reduction in total length of stay. The announcement presents these as performance indicators for the command centre’s broader work, which includes operational coordination and direct team interventions alongside technology.

Earlier preparation, not automated care decisions

The present change is a longer view of expected demand. KFSH says combining predictive analytics with live operational monitoring and coordinated intervention allows it to prepare staffing and capacity earlier, before bottlenecks intensify. The stated purpose is to move patient-flow management away from responding to congestion after it develops and toward preparing for it in advance.

The hospital draws a clear boundary around that role: clinical teams remain responsible for triage and care decisions. Technology supports the operational coordination surrounding those decisions. The announcement concerns forecasting department demand and managing patient flow, not transferring responsibility for assessing or treating individual patients to an AI system.

How accurately the longer-range forecast predicts demand remains unanswered: the announcement gives no forecast-error measure. It also does not isolate AI’s contribution to the reported waiting-time gains. Those are hospital-reported results from a combined operational program, not evidence that doubling the forecast window caused the earlier improvements.