BMC Health Serv Res. 2026 Feb 5. doi: 10.1186/s12913-026-14093-1. Online ahead of print.
ABSTRACT
BACKGROUND: This study explores the “wave-off” mechanism in healthcare, in which physicians implicitly discourage patient revisits to manage high workloads. Understanding this mechanism is critical for balancing workload management and patient-centered care, as it highlights how physician discretion shapes patient behavior and operational efficiency.
METHODS: We analyze 200,426 outpatient records from a cardiology department to investigate the relationship between physician workload and the likelihood of patient revisits. We introduce a novel metric, Sample Entropy, to quantify patient “stickiness,” or the propensity to revisit. We examine physician behaviors-such as referrals for diagnostic examinations and medication prescribing-to identify strategies used to modulate patient flow. Statistical analyses assess the impact of workload on revisit patterns and the mechanisms driving the wave-off effect.
RESULTS: Higher physician workloads are significantly associated with reduced short-term patient revisits. Physicians manage workload by prescribing more medications and referring fewer diagnostic examinations, laboratory tests, particularly for patients with higher stickiness as measured by Sample Entropy. These behaviors alleviate short-term workload pressure but risk long-term inefficiencies, potentially reflecting bounded rationality in clinical decision-making.
CONCLUSIONS: The wave-off mechanism illustrates a trade-off between immediate workload relief and long-term operational performance. While it helps physicians address short-term capacity constraints, it may undermine optimal long-term patient care and system efficiency. This study highlights the operational and behavioral implications of physician-driven demand modulation, including the subtle yet consequential phenomenon of physician-reduced demand, in which patients are implicitly discouraged from revisiting due to workload-driven adjustments in care delivery.
PMID:41645147 | DOI:10.1186/s12913-026-14093-1
