BACKGROUND: The Patient and Observer Scar Assessment Scale (POSAS) v2.0 is a widely used instrument to evaluate postsurgical scars. Its two subscales respectively investigate the patient's and clinician's opinion on the scar quality. However, psychometric studies of the POSAS have indicated that its metric performance is suboptimal, and structural adjustments may be appropriate. The authors aimed to verify through Rasch analysis the measurement properties of the POSAS v2.0, and propose eventual structural refinements for an easier and more confident use of the scale in clinical practice and research.
METHODS: Consecutive patients admitted to two rehabilitation centers for postsurgery rehabilitation over a 2-year period underwent scar assessments with the POSAS v2.0. We performed Rasch analysis to examine the scale's dimensionality, rating categories, item fit, reliability indices, local item independence, and differential item functioning.
RESULTS: The study population consisted of 115 patients. The 10 response options of the POSAS showed malfunctioning, and thus were collapsed, forming a parsimonious five-level rating scale, which helped to improve the measurement accuracy. After that, unidimensionality of both subscales was confirmed. Then, internal construct validity of the POSAS v2.0 was demonstrated (through item fit to the Rasch model). Reliability indices were high (≥0.80). No significant differential item functioning was detected concerning age or sex.
CONCLUSION: This study demonstrates the good psychometric properties of a simplified Rasch-based version of the Patient and Observer Scar Assessment Scale with five response options (POSAS v2.1) in patients with postsurgical linear scars, and provides insights for future refinement of the tool.