A Three-Level Integrated Decision Support System for Surgical Scheduling: Elective Planning, Emergency Management, and Real-Time Bed Occupancy Integration

This paper describes a decision support system prototype for surgical scheduling in Italian NHS healthcare facilities. The system addresses a gap in the existing literature by proposing an integrated three-level architecture that simultaneously manages elective surgical planning, emergency surgical demand, and real-time inpatient bed occupancy — three problem domains that existing scheduling systems treat independently or in pairwise combinations. The central methodological contribution is a hierarchical two-tier framework for scheduling decision support. The first tier — clinical and regulatory compliance — operates as a non-negotiable precondition: it guarantees adherence to PNGLA urgency classifications (A>B>C>D, with mandatory maximum waiting times), clinical appropriateness guidelines by specialty, and regional tariff regulations (DGR Lazio n. 1186/2024 abatement rules). The second tier — strategic optimisation — operates exclusively within the space defined by the first tier, balancing the facility's quarterly case mix targets against the real post-abatement economic value of each scheduled procedure. Clinical priority is guaranteed by construction and cannot be traded off against economic value, even implicitly. The emergency management module introduces a hybrid buffer architecture: static when a dedicated emergency theatre is planned in the operating calendar, dynamic when no dedicated emergency slot is scheduled. When emergency volume exceeds available capacity, the system reschedules elective procedures to the earliest available session based on clinical criticality, not economic priority, preserving the integrity of the elective plan. The bed occupancy module receives patient census data from the Admission, Discharge, and Transfer (ADT) system three times daily, covering ordinary inpatient wards, intensive care units, and operating theatres. It produces a projected occupancy trajectory — incorporating both planned elective activity and emergency admission history — that serves as a hard constraint in the elective scheduling engine. The system is applicable to any Italian NHS healthcare facility; the primary difference between public and NHS-accredited private facilities is the annual budget cap that governs the latter's production volume. The system is in active development and has not yet been validated on live operational data. A methodological realignment conducted in April 2026, in which a validation narrative built on an internally inconsistent backtest was withdrawn and replaced with a sounder prospective evaluation framework, is documented in detail as an instance of the epistemic rigour required in healthcare decision support development.