1. Introduction
Anyone who has sat in a crowded outpatient waiting room, watching the clock while a queue barely seems to move, has an intuitive sense of what this paper is about. Patient flow, the term researchers use for how people move through registration, consultation, diagnostics, and discharge, has become one of the more persistent headaches in health service management, and not for lack of attention (Akenroye et al., 2023). When that movement stalls, the consequences are not merely inconvenient. Patients wait longer than they should, hospitals become crowded beyond what their physical layout can comfortably absorb, and the quality of care that ultimately gets delivered tends to suffer as a result (Bakker & Tsui, 2017). Perhaps unsurprisingly, the frustration is not one-sided; clinicians and administrative staff report similar strain, caught between rising operational costs, tightening resources, and patients who are, understandably, growing less patient (Feng et al., 2021).
Much of the operations-research literature has approached this problem through queueing theory, a mathematical framework that, at its core, tries to describe how arrivals, services, and delays interact within a system (Wang & Liu, 2021). It is not a new idea, and it is not exotic; anyone who has stood in a bank line has lived it. What queueing models offer healthcare administrators, though, is a way to see bottlenecks before they become crises, to anticipate where a system will strain under load, and to reallocate staff or space accordingly (Hu et al., 2021). This kind of forecasting has proven especially useful in emergency departments, outpatient clinics, and diagnostic units, precisely because these are the settings where patient arrivals are hardest to predict and easiest to underestimate (Van Hulzen et al., 2021).
At more or less the same time, a rather different toolkit has been gaining ground: data-driven analytics, built on the growing availability of electronic health records and real-time monitoring infrastructure (Manktelow et al., 2022). Where queueing theory offers a formal, somewhat abstract model of how a system behaves, analytics offers something more immediate, a live read on what is actually happening on the floor right now, and, increasingly, a forecast of what is likely to happen next (Tavakoli et al., 2022). Paired with predictive algorithms, these systems have started to reshape how hospitals schedule staff and allocate beds, often catching problems that a purely theoretical model might miss (Fan et al., 2019). In a country like the United States, where patient volumes are large, providers numerous, and digital systems frequently sit in silos rather than talking to one another, this kind of visibility is not a luxury so much as a necessity (He et al., 2018). Emergency departments in particular continue to report overcrowding severe enough to compromise both safety and efficiency (Hu et al., 2017), while outpatient scheduling remains, in many settings, more art than science, with appointment patterns that work against rather than with available capacity (Ahsan et al., 2019). Digital solutions have been rolled out to address these gaps, admittedly, but adoption alone has rarely been sufficient; without a strategic framework tying technology to underlying theory, the improvements tend to be partial at best (Bard et al., 2014).
What is somewhat curious, given how much attention each method has separately received, is how rarely they have been studied together. Prior work has shown that queueing theory alone can trim waiting times by a modest but meaningful five to nine percent when properly applied (Li et al., 2021), and data-driven analytics has, on its own terms, earned a reputation for sharpening forecasting and resource decisions (Zhai et al., 2022). Yet the bulk of this literature treats the two as parallel tracks rather than an integrated system (Wartelle et al., 2022), a divide that, by most accounts, has kept healthcare organizations from realizing whatever additional benefit their combination might offer (Tao & Liu, 2019).
This study, then, sets out to address that gap directly, examining whether and how queueing theory and data-driven analytics, considered together rather than apart, might be used to optimize patient flow across healthcare settings. In doing so, it aims to give administrators something more actionable than either framework offers alone: a clearer sense of where the two genuinely reinforce one another, and a modest, evidence-grounded case for treating them as complementary investments rather than competing priorities.
