Background <p>Diagnostic errors remain a pressing challenge in health systems with uneven provider capacity and limited diagnostic standardization. In such environments, cognitive biases, particularly anchoring effect, may compromise diagnostic independence and reinforce structural disparities in care quality.</p> Methods <p>We conducted a randomized field experiment in western China using standardized patients (SPs) to examine how prior spectacle diagnostic prescriptions influence the behavior and accuracy of second-opinion optometrists. SPs visited optical providers in Shaanxi province, presenting either no prior prescription, a correct one, or an incorrect one. Diagnostic outcomes were evaluated against gold-standard prescriptions issued by an expert ophthalmologist.</p> Results <p>Exposure to prior prescriptions, especially inaccurate ones, significantly reduced diagnostic accuracy and process completeness. Providers given prior diagnoses were less likely to conduct key tests and spent less time on examinations, suggesting reliance on cognitive shortcuts. These findings provide field-based evidence of anchoring bias in real-world clinical settings.</p> Conclusions <p>Prior diagnostic information can shape second-opinion decision-making through cognitive anchoring, particularly in systems lacking strong institutional protocols. Addressing these biases through structured diagnostic procedures and provider training may enhance diagnostic accuracy and promote greater equity in vision care delivery.</p>