Curse of the Anchor

Most people work in jobs that do not expose them to rapidly changing situations requiring immediate responses with limited or incorrect information. The emergency services are different. Dynamic and dangerous situations with threat to life and personal safety can arise completely unexpectedly on shift, often in challenging environments due to crowds, access, weather, or poor lighting. Decisions taken can have tragic outcomes and very long reaching repercussions which will be subject to intense scrutiny retrospectively by people who have never been anywhere near that degree of risk exposure. 

In such incidents, under stress and with pressure of time to intervene, anchor bias is one of the main confounders. In my work both in the ambulance service and in a very busy major trauma centre, I have fallen victim to this problem in which the first piece of information forms the basis of the incident narrative to the exclusion of clear contradictory evidence. For example, the 999 call was for a collapse, but the patient is badly injured after a fall down the stairs. The potential medical causes of collapse are prioritised over a careful initial search for injuries (despite the slippers being obvious at the top of the stairs) until a fracture is noticed and then the assessment is reset and taken in a different, more appropriately focused direction. This is related to, but fundamentally different from, confirmation bias where information is sought to confirm a pre-existing internal personal or institutional belief system.

The trick to prevent anchor bias resulting in a tragedy is to pause, slow down and review before committing to action (“take a beat”) and to have a healthy scepticism about the quality of the original information or diagnosis. Acquiring this skill takes training, deliberate practice (often in simulated situations), and experience. I have learned to do it (most of the time) by having made significant mistakes in the past. Organisations can raise awareness of the risk through internal governance structures that review incidents critically and allow colleagues to experience the dilemmas faced by the first responders. Training and operational procedures can be adjusted as a result of review, but anchor bias is extraordinarily powerful under pressure, and in my experience is a much greater threat to safety than confirmation bias relating to personal or cultural issues.