Adaptation and validation of the Health Anxiety Inventory (short version) for medical settings

Objectives: This study seeks to adapt the 18-item Health Anxiety Inventory - Short Version (SHAI; Salkovskis et al., 2002) for use in medical settings and to evaluate the adapted version's psychometric properties. It reports two studies. The objectives of the first study were to use expert opinion 1) to identify the limitations of the SHAI for administration in medical settings, 2) to formulate adaptations to items for use in medical settings through an iterative process of feedback and modification, and 3) to develop an adapted version of the SHAI for medical settings using an evidence-based consensus. The objective of the second study was to evaluate the psychometric properties of the adapted measure in samples with Multiple Sclerosis (MS), myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and chronic pain. Method: In study one, a Delphi study method was used to adapt the SHAI for medical settings. Eight experts with experience of living with a physical health condition and four experts by profession working within physical health were invited to review the SHAI and suggest adaptations through an iterative developmental process. Study two employed a cross-sectional mixed methods questionnaire design. Individuals with MS, ME/CFS and chronic pain were invited to complete a battery of questionnaires via an online survey, and to complete a second survey one week later, allowing for the psychometric properties of the adapted measure to be evaluated. Results: An evidence-based consensus was reached for an adapted version of the SHAI for medical settings (HAI-M) following study one. Consultation with an expert in the field of health anxiety and a person with personal experience confirmed face validity of the 12-item HAI-M. During study two, 315 participants took part in the online survey of which 80.6% fully completed the HAI-M and 62.86% fully completed the battery of questionnaires. 33.3% completed the follow up questionnaire one week later. All groups rated the HAI-M as more acceptable for assessing their health concerns than the SHAI and no significant differences were found between clinical groups in how they scored on the HAI-M. No significant differences were found between clinical groups in how they scored on the SHAI which was an unexpected finding. The HAI-M demonstrated high internal consistency (.875), good test-retest reliability (.812) and good convergent validity (.801). Divergent validity was also good (.515). The exploratory factor analysis suggests the presence of a 3-factor model and this model is discussed in relation to previous studies reporting the factor structure of the SHAI. Conclusion: This study provides preliminary evidence for a psychometrically sound health anxiety screening tool for use in medical settings. Recommendations are made for further research.