Authors' Reply: “An Analysis of the Timeline to Diagnosis and Treatment in Oral Cavity and Oropharynx Cancer”

We thank the authors for their thoughtful and constructive letter, which explicitly recognizes the methodological appropriateness and relevance of our study, while offering the opportunity to further clarify specific terminological aspects of the analysis (Gonçalves et al. 2026). The Cox proportional hazards model was used exclusively in an exploratory univariate stage, with the aim of identifying initial associations between clinicopathological variables and the analyzed time intervals, without any intention of causal inference or prospective risk interpretation, given that the intervals were measured retrospectively. The multivariable analysis was conducted using negative binomial regression, selected due to the discrete nature of the outcome (time in months), its right-skewed distribution, and the presence of overdispersion, evidenced by variance exceeding the mean. In this context, incidence rate ratios (IRR) should be understood strictly as ratios between expected times derived from a count model, conditional on the included covariates, rather than as measures of epidemiological incidence or risk. Throughout the manuscript, results were consistently interpreted as relative differences in the duration of the analyzed intervals, deliberately avoiding any interpretation in terms of incidence or risk, in accordance with the study design and the nature of the data (Martínez-Ramírez et al. 2025). Josefina Martínez-Ramírez: methodology, data curation, formal analysis, writing – original draft. Thaís Bianca Brandão: resources, writing – review and editing, validation. Alan Roger Santos-Silva: conceptualization, validation, funding acquisition, project administration, resources, writing – review and editing. Ana Carolina Prado Ribeiro: data curation, supervision, resources. The authors declare no conflicts of interest.