Attribution bias underlying burns-induced anxiety symptoms

Introduction: Burn injuries are a debilitating cause of morbidity and mortality associated with the long-term impact of psychological factors on quality of life. Accurate assessment of the differential impact of burn sequelae and anxiety is often complicated by the overlap between psychological and somatic symptoms in burns patients. The Beck Anxiety Inventory (BAI) is one validated psychometric tool for anxiety assessment. The primary objective of this study is to investigate whether utilising the BAI as a tool to assess for anxiety in burns patients is biased due to the confounding of symptoms of anxiety with the physical sequelae of a burn injury.

Methods: This is a single-centre, prospective, cross-sectional study. The study was conducted in accordance with the UK Good Clinical Practice guidelines (CAPP reference number 506). Patients were recruited over a three-month period from November 2016 to February 2017 and were offered a modified BAI questionnaire to complete. Patients were asked to indicate to what degree they attributed each symptom to their physical injury or their psychological state on a visual analogue scale (VAS).

Results: 50 patients, comprising 33 females (66%) and 17 males (34%), participated in the study with a median age of 33.5 years (range: 20-88). Date of injury spanned May 1991 to January 2017. Percentage of the total body surface area (% TBSA) affected by burn ranged from 1 to 86%. Patients attributed eight of the 21 self-report items within the BAI as being more physical than psychological in origin. The results reveal a statistical significant difference in patient VAS scores between physical (mean: 34.16, 95% CI: 29.04-39.28) and psychological (mean: 61.2, 95% CI: 56.33-66.17) BAI items, with p<0.0001. In addition, patients with a facial burn injury were more likely to report 'face flushed' (Mann-Whitney U Test, Z=-2.11, p<0.05) and patients with a hand burn injury were more likely to report 'hands trembling' (Mann-Whitney U Test, Z=-2.52, p<0.05).

Conclusions: This feasibility study found preliminary evidence suggesting that the BAI may, in part, represent misattributed symptoms of cutaneous injury from burns. However, whilst our findings suggest an attribution bias, there is not enough evidence from this data to comment on whether its use should be restricted in burns patients. Further research is needed to formally quantify convergent and divergent validity through structured interviews. In addition, further research using other self-report tools of anxiety in burns patients would be useful to corroborate the prospect of biased and confounded anxiety scores.