Utilising linear sequential unmasking to minimise bias in ARCP

Background

Research has demonstrated that the order in which data is presented plays a critical role in decision-making processes and outcomes.¹ It is well-documented that people tend to remember the initial information in a sequence better — and be more strongly impacted by it — than subsequent information in the sequence. Such effects have been shown in various domains, from food tasting to jury decision-making.

The Annual Review of Competency Progression (ARCP) is the process of reviewing the trainee's evidence of achievement over a period of training to decide progression.² It is not an assessment but a summative judgement of the doctor's performance and development throughout that period. ARCP panels utilise the data contained in the trainee's portfolio to decide progression.

Method

Linear Sequencing Unmasking-Extended (LSU-E)3 is a process used to minimise bias by regulating the flow and order of information. LSU-E provides three criteria for determining the optimal sequence of exposure to task-relevant information: biasing power, objectivity, and relevance.

We believe we can minimise cognitive bias in the ARCP process by utilising LSU-E principles. We propose an approach that controls data better before and during an ARCP. We have organised all data discussed in the ARCP depending on its relevance, objectivity and biasing power. Based on that, we agreed on the correct order of how these data are discussed.

We are not clear how much bias affects outcome for doctors in training, but we believe that anything that can mitigate bias will be positive for doctors in training. Additionally, we are not sure of the process practicality. Theoretically, less information is discussed, so less time is spent per trainee. We can answer these questions by piloting the process for different specialties and modifying it based on panels' feedback.