Presentation (CTI)

OBJECTIVES: To examine how diagnostic uncertainty shapes additional testing, treatment deviation, and resource consequences in coronary stenosis decisions, using primary clinician-reported data and established health services research methods relevant to health economics and outcomes research.
METHODS: We analyzed a physician-vignette study conducted in Wuhan, China. Seventy-four cardiologists each evaluated the same 10 coronary stenosis cases, yielding 740 physician-case observations. Quantitative coronary angiography was used to define a 70% stenosis benchmark for medication versus percutaneous coronary intervention (PCI). Physicians reported visual stenosis estimates, perceived diagnostic uncertainty, and management decisions among medication, fractional flow reserve (FFR), and PCI. We used logistic regression to examine factors associated with FFR use as additional testing and with treatment deviation among no-FFR decisions. A Wuhan base-case resource framework assigned pathway-specific procedural costs and estimated incremental resource loss across the full decision sequence.
RESULTS: FFR was selected in 300 of 740 observations (40.5%). Greater subjective uncertainty and closer visual proximity to the 70% threshold were independently associated with FFR use (both p<0.001 in the main model). Among 440 no-FFR observations, any treatment deviation was associated with larger estimation error between visual assessment and quantitative angiography (coefficient 0.61, p<0.001), whereas subjective uncertainty was not significant. Mean incremental resource loss increased across uncertainty tertiles from CNY 1,916 to CNY 3,418 and across threshold-proximity tertiles from CNY 1,727 to CNY 3,671, indicating higher expected resource consequences in threshold-adjacent decisions.
CONCLUSIONS: In this primary-data vignette study, coronary stenosis management under diagnostic uncertainty followed a branching decision process rather than a single treatment choice. Uncertainty and threshold proximity mainly promoted testing escalation through FFR, whereas downstream treatment deviation without FFR was driven more by estimation error. These findings link diagnostic uncertainty to care-process variation, potentially low-value resource use, and measurable resource consequences in an Asia-Pacific setting.