The use of metaphors to talk about cancer experiences has attracted much research and debate, especially in the case of military metaphors. However, questions remain about what metaphors are used by different populations for different aspects of the cancer experience. This scoping review aims to answer them.

We searched PubMed, PsycINFO, CINAHL, Scopus and Web of Science databases. Eligible sources include peer-reviewed scientific research published in English between 2013 and 2023, investigating patterns of metaphor use from adult populations (age ≥18) for cancer-related topics, such as cancer itself, the general experience of being ill, treatment, and people and relationships.

Out of 1929 articles identified, 30 met the criteria, spanning over different populations. While most papers focused on cancer in general, some focused on specific cancer types, such as breast cancer. Both spontaneous and elicited data were collected in ten languages: mostly English (N=12), Swedish (N=3) and Arabic (N=3). The identified metaphors were subsumed under various broad categories, including particularly violence and journey. Other categories include education and non-human animate entity for the cancer itself, confinement and deprivation and cleanliness for the general experience of being ill with cancer, Poison and gardening for cancer treatment, and distance for patients’ social relationships.

It was found that metaphors help to identify how patients describe experiences of vulnerability and empowerment. To provide patient-centred care, clinicians and researchers should avoid blanket conclusions about helpful or unhelpful metaphors, but consider the ways in which different metaphors are used by different populations in different contexts.

Multinomial logistic regression models are popular in multicategory classification analysis, but existing models suffer several intrinsic drawbacks. In particular, the parameters cannot be determined uniquely because of the over-specification. Although additional constraints have been imposed to refine the model, such modifications can be inefficient and complicated. In this paper, we propose a novel and efficient simplex-based multinomial logistic regression technique, seamlessly connecting binomial and multinomial cases under a unified framework. Compared with existing models, our model has fewer parameters, is free of any constraints, and can be solved efficiently using the Fisher scoring algorithm. In addition, the proposed model enjoys several theoretical advantages, including Fisher consistency and sharp comparison inequality. Under mild conditions, we establish the asymptotical normality and convergence for the new model, even when the numbers of categories and covariates increase with the sample size. The proposed framework is illustrated by means of extensive simulations and real applications.