Assessment of the sunk-cost effect in clinical decision-making

► U.S. Clinicians did not demonstrate a sunk-cost effect when making a hypothetical clinical decision. ► Instead, clinicians demonstrated active avoidance and overcorrection for sunk-cost effects. ► Absence of sunk-cost effect held regardless of type of prior investment or provider demographic characteristics. ► A surprising number of clinicians would recommend a patient continue treatment that is ineffective. ► Recommendation for continuation of ineffective treatment reflected unrealistic optimism about future treatment efficacy.