A Qualitative Study Exploring the Role of Hindsight Bias in the Process of Reviewing Clinical Practice Prior to Adverse Incidents

[摘要] Aims. To explore the effect of hindsight bias on retrospectivereviews of clinical decision making prior to adverse incidents toinform future approaches to incident investigations.Methods. We have undertaken focus groups with doctors of varying grades across the North West of England and North Wales. Avignette based on a real-life case from the publicly available NHSEngland Homicide Independent Investigation report database waspresented to each group in one of three versions which differed interms of the ending of the vignettes (i.e. suicide, homicide, noadverse incident). Using a semi-structured interview approach,the group participants were encouraged by the facilitators toreflect on issues relating to risk and risk management. All groupswere provided with the same vignette which initially made no reference to the outcome and asked to comment on matters of riskand risk management. Halfway through the discussion, one of thethree outcomes was disclosed, and further group discussion washeld. The recorded interviews were transcribed and thematic analysis was undertaken using an adapted Framework Method.Results. Preliminary results (n = 10) indicate that participantsidentified the potential for significant harm, particularly to others,and identified evidence of key psychopathological and historicalcorrelates to support assertive management of risk and admissionto hospital.Whilst knowledge of the outcome did not lead to participantschanging their favoured management plans, it did alter how they appraised the case and led to participants constructing “narrative”explanations for the outcome given. The level of conviction participants held for their management plan reduced when theirexpectations about the outcome were confounded.Participants presented with the suicide outcome vignettedescribed their difficulties appraising risk to others and theirover-sensitivity to that risk. Participants faced with the ‘noadverse outcome’ vignette perceived the original managementplan far more favourably in hindsight. The groups that were presented with the homicide outcome vignette initially focused onboth risks to self and others as well as the perceived need for further information. Following knowledge of the outcome, there wasa tendency to highlight parts of the letter pertaining to risk toothers which they previously had not given as much attention.Conclusion. The initial analysis of our data confirms the findingsfrom previous studies that hindsight colours the appraisal ofadverse events. However, this study is novel in that it describesthe nature of the thought processes underpinning the influenceof hindsight on appraisals of risk.

[学科分类] 计算机科学（综合）