Evidence for Anchoring Bias During Physician Decision-Making

Area of Science:

• Medical Decision Making
• Health Services Research

Background:

• Cognitive biases, such as anchoring bias, can influence physician decision-making.
• Limited large-scale evidence exists on the impact of these biases in clinical practice.

Purpose of the Study:

• To investigate if physicians are less likely to test for pulmonary embolism (PE) in patients with congestive heart failure (CHF) presenting with shortness of breath (SOB) when CHF is documented in the triage notes.
• To assess the association between documented CHF at triage and the timeliness and diagnosis of PE.

Main Methods:

• Cross-sectional study using national Veterans Affairs data from 2011-2018.
• Included 108,019 patients with CHF presenting with SOB.
• Analyzed the association between mention of CHF in triage documentation and PE testing rates, time to testing, and PE diagnosis.

Main Results:

• Patients with CHF and SOB were less likely to receive PE testing (4.6 pp reduction) and had delayed testing (15.5 minutes longer) if CHF was mentioned in triage.
• BNP testing increased (6.9 pp) in this group.
• A lower likelihood of acute PE diagnosis in the ED was observed (0.15 pp reduction), but no significant association with ultimate PE diagnosis.

Conclusions:

• Physician anchoring on initial triage information (CHF mention) may lead to reduced and delayed pulmonary embolism testing in patients with congestive heart failure.
• This bias can impact the diagnostic workup and potentially the timely diagnosis of PE.