Abstract
Over several decades, the incidence of melanoma and nonmelanoma skin cancers has markedly increased. This rise underscores the importance of examinations and biopsies for early detection. In practice, it is challenging to identify all significant disease while avoiding excess economic costs and patient morbidity from over-biopsying. In dermatology, the point of balance, while elusive, has been expressed as number-needed-to-treat (NNT, number-needed-to-excise/biopsy, NNE/NNB), introduced in 2004 by English et al and defined as the total number of biopsies divided by the number of biopsies positive for melanoma, or the inverse of absolute risk. English et al’s NNT (NNTE) has been used in reference to melanoma biopsy rates, including in a meta-analysis recently published by Petty et al trying to calculate an overall NNTE. However, before this, NNT (NNTL) was already a well-established statistic, introduced in 1988 by Laupacis et al and defined differently. NNTL is the inverse of the risk difference (absolute risk reduction). It describes the number of individuals that need to be treated for one additional improved outcome. Like NNTE, NNTL can be applied to the discussion of biopsy rates, albeit in a different manner. The key differences are that NNTL considers both non-biopsied melanomas and controls, while NNTE describes an absolute risk and not a risk difference. In this presentation, we demonstrate the differences between NNTE and NNTL to clarify the scientific definitions and avoid misunderstanding or misuse. Additionally, we highlight the uses of Youden’s index, number-needed-to-diagnose (NNDL), their relation to sensitivity/specificity, and their utility in the biopsy rate discussion.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call