101. Evaluating the quality of reporting of melanoma prediction models.
- Author
-
Jiang MY, Dragnev NC, and Wong SL
- Subjects
- Humans, Validation Studies as Topic, Melanoma, Models, Statistical, Skin Neoplasms
- Abstract
Background: Multivariable prediction models combine patient data points to provide actionable estimates of outcomes. Prediction models for melanoma are important for guidance in the midst of the rising incidence and evolving treatment options. This study evaluates the quality of reporting of prediction models using the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) checklist., Methods: We performed a systematic literature search to identify publications describing development and/or validation of melanoma prediction models. For each study, reviewers assessed compliance with 22 TRIPOD items. We also assessed a model's predictive ability (area under the curve) compared with TRIPOD adherence., Results: We originally identified 67 articles, of which 27 met inclusion criteria. No study completely followed the TRIPOD checklist, and median overall adherence was 61%. Authors were least likely to report participant characteristics, title, and abstract in accordance with the TRIPOD checklist. Linear correlation between a model's area under the curve and TRIPOD checklist adherence was not statistically significant, r = -0.09 (P = .34)., Conclusion: Current reporting of melanoma multivariable prediction models does not meet standards. Although there is room for improvement in how melanoma models are reported, our findings do not indicate a significant relationship between the model's performance and adherence to the TRIPOD checklist., (Copyright © 2020 Elsevier Inc. All rights reserved.)
- Published
- 2020
- Full Text
- View/download PDF