A Simulation Modelling Study of Referral Distribution Policies in a Centralized Intake System for Surgical Consultation.

Delays beyond recommended wait times, especially for specialist services, are associated with adverse health outcomes. The Alberta Surgical Initiative aims to improve the referral wait time-the time between a referral is received at the central intake to the time a specialist sees the patient. Using the discrete event simulation modelling approach, we evaluated and compared the impact of four referral distribution policies in a central intake system on three system performance measures (number of consultations, referral wait time and surgeon utilization). The model was co-designed with clinicians and clinic staff to represent the flow of patients through the system. We used data from the Facilitated Access to Surgical Treatment (FAST) centralized intake referral program for General Surgery to parameterize the model. Four distribution policies were evaluated - next-available-surgeon, sequential, "blackjack," and "kanban." A sequential distribution of referrals for surgical consultation among the surgeons resulted in the worst performance in terms of the number of consultations, referral wait time and surgeon utilization. The three other distribution policies are comparable in performance. The "next available surgeon" model provided the most efficient and robust model, with approximately 1,000 more consultations, 100 days shorter referral time and a 14% increase in surgeon utilization. Discrete event simulation (DES) modelling can be an effective tool to illustrate and communicate the impact of the referral distribution policy on system performance in terms of the number of consultations, referral wait time and surgeon utilization.

Intended and unintended consequences of strategies to meet performance benchmarks for rheumatologist referrals in a centralized intake system.

RATIONALE: Timely assessment of a chronic condition is critical to prevent long-term irreversible consequences. Patients with inflammatory arthritis (IA) symptoms require diagnosis by a rheumatologist and intervention initiation to minimize potential joint damage. With limited rheumatologist capacity, meeting urgency wait time benchmarks can be challenging. We investigate the impact of the maximum wait time guarantee (MWTG) policy and referral volume changes in a rheumatology central intake (CI) system on meeting this challenge. METHODS: We applied a system simulation approach to model a high-volume CI rheumatology clinic. Model parameters were based on the referral and triage data from the CI and clinic appointment data. We compare the wait time performance of the current distribution policy MWTG and when referral volumes change. RESULTS: The MWTG policy ensures 100% of new patients see a rheumatologist within their urgency wait time benchmark. However, the average wait time for new patients increased by 51% (178-269 days). A 10% decrease in referrals resulted in a 76% decrease on average wait times (178-43 days) for new patients and an increase in the number of patients seen by a rheumatologist within 1 year of the initial visit. CONCLUSION: An MWTG policy can result in intended and unintended consequences-ensuring that all patients meet the wait time benchmarks but increasing wait times overall. Relatively small changes in referral volume significantly impact wait times. These relationships can assist clinic managers and policymakers decide on the best approach to manage referrals for better system performance.

Alternative care providers in rheumatoid arthritis patient care: a queueing and simulation analysis.

Patients diagnosed with rheumatoid arthritis require lifelong monitoring by a rheumatologist. Initiation of the disease-modifying anti-rheumatic drug therapy within twelve weeks of the onset of symptoms is crucial to prevent joint damage and functional disability. We examine the impact of the engagement of alternate care providers (ACP) in alleviating delay due to limited rheumatologist capacity. Using queueing theory and discrete-event simulation, we model rheumatologist-only and rheumatologist-with-ACP system configurations as closed, multi-class queueing networks with class switching.Using summary data from an actual rheumatology clinic for illustration, we analyze various parameter conditions to aid clinic managers and policymakers in decisions concerning capacity allocations and feasible patient panel size that impact timeliness of care and resource utilization.Results not only confirm that a substantial increase in RA patient panel size with an ACP involved in the care of follow-up patients but also demonstrates the boundaries for feasible panel sizes and workload allocation.