113 Exploring the potential impact of new diagnostic labels for melanoma in-situ: an online randomised experiment

<h3>Objectives</h3> This study aimed to explore whether alternative diagnostic labels to communicate a hypothetical melanoma in situ diagnosis influenced management choice, perceived risk of invasive melanoma and melanoma death, and level of anxiety among Australian adults aged 40 years or older. The alternative labels were chosen by our Clinician and Consumer Co-Investigators through a series of iterative online questionnaires. The final two labels that were chosen were: ‘low-risk melanocytic neoplasm’ and ‘low-risk melanocytic neoplasm, in-situ’. These were intended to convey the low, but not zero, risk of developing an invasive melanoma and of dying from melanoma. <h3>Method</h3> The study was designed as a parallel randomised online hypothetical experiment, with participants randomised 1:1:1 to one of: ‘melanoma in-situ’ (control), ‘low-risk melanocytic neoplasm’, or ‘low-risk melanocytic neoplasm, in-situ’. The co-primary outcomes were (i) choice between no further surgery or further surgery to ensure clear histological margins greater than 5 mm, and (ii) choice between patient-initiated clinical follow-up when needed (patient-led surveillance) and regular routinely scheduled clinical follow-up (clinician-led surveillance). Secondary outcomes included diagnosis anxiety, perceived risk of invasive melanoma and of dying from melanoma, and management choice anxiety (after surgery choice and follow-up choice). We will make pairwise comparisons across the three diagnostic label groups using regression models (univariable and multivariable). Using multivariable regression models in an intention-to-treat approach to analysis, we made pairwise comparisons across the three diagnostic label groups. <h3>Results</h3> Of 1688 people randomised (50.1% male), there were 562 in the melanoma in-situ (control) group and 563 in each of the alternative label groups; covariates were balanced across randomised groups. Extreme anxiety was reported after the hypothetical diagnosis in 47 (8.4%) of the melanoma in-situ (control), 3% (17/563) of the low risk melanocytic neoplasm, and 3.9% (22/563) of the low risk melanocytic neoplasm, in-situ groups (p&lt;0.001). Perceived risk of dying from melanoma was 46% (control), 29% (low risk melanocytic neoplasm) and 31% (low risk melanocytic neoplasm, in-situ)(p&lt;0.001). No further surgery was chosen by 43% (239/562) in control, 52% (295/563) of low risk melanocytic neoplasm, and 51% (289/563) of low risk melanocytic neoplasm, in-situ groups (p&lt;0.001). Patient-initiated follow up was chosen by 21% (119/562) in control, 24% (136/563) of low risk melanocytic neoplasm, and 23% (130/563) of low risk melanocytic neoplasm, in-situ groups (p=NS). <h3>Conclusions</h3> 1. Both new labels reduced anxiety, perceived risk of developing invasive melanoma or of dying from melanoma, and increased the probability of choosing no further surgery for a completely excised lesion with narrow margins. They had minimal impact on choosing choosing patient-initiated follow up. 2. The addition of ‘in situ’ does not appear to add value to the alternative diagnostic label. Alternative terminology, that better communicates the difference between low risk in situ lesions and low risk invasive lesions, needs to be developed and tested. 3. Perceived risk across all randomised arms was much higher than actual risk; additional interventions beyond a change in diagnostic label are likely needed. Risk communication interventions developed according to health literacy principles, may have synergistic effects to a new diagnostic label. 4. Research is needed to explore effects of new diagnostic labels on the management choices offered by treating clinicians to their patients.