ASSIST: Development of a Simplified Clinician-Patient Hybrid Reporting Outcome Measure for Remote Diagnosis of Surgical Site Infection.

Remote assessment of surgical site infection(SSI) lacks sensitivity for the diagnosis of SSI, but current evidence has not evaluated whether a combination of photographs and questionnaires improves diagnostic accuracy. This study aims to develop a remote diagnostic measure to identify SSI. A two-phase mixed methods study was conducted. In phase I, five clinicians reviewed the Bluebelle wound healing questionnaire(WHQ) on a five-point Likert scale of agreement for inclusion in a remote measure. Discussion generated a hypothesis as to which items should be included. In phase II, a cohort study, whereby clinicians evaluated patient's wound images and patients completed the WHQ, were reviewed for scale structure. Principal component analysis (PCA) with scree plot examination and maximum likelihood of estimation (MLE) for one, two and three factors were evaluated. Internal consistency was assessed with Cronbach's α. Phase I: hypothesis generation estimated a measure containing between 10 and 12 items would include all relevant items without ambiguity or redundancy. Phase II: a combined sample of 570 responses provided clinician reviewed images and patient responses. PCA suggested that a 12-item measure with a combined variance of 60.2% would have the best model fit. Cronbach's α was high at 0.841. One included item was highlighted as potentially ambiguous in phase I (wound pain), providing an additional model with this removed. MLE for one, two and three factors suggested measures with 8, 10 and 11 items, respectively. Total variances were low at 29.7%, 39.8% and 41.4% and Cronbach's α were high at 0.838, 0.827 and 0.823, respectively. Three potential models for a remote diagnostic measure were identified. Each is shorter than alternative available measures, which have not been designed for combined use, ensuring this is easy to use. Further evaluation for reliability and diagnostic accuracy is needed to validate a final measure that can be implemented in clinical practice.