The three-month referral threshold as an independent prognostic marker of cancer-specific survival in advanced oncological disease

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Background: Guidelines recommend the early integration of palliative care in advanced cancer; however, “earliness” remains heterogeneously defined, with no temporal consensus, and timing studies are vulnerable to immortal-time bias and do not model competing risks. Objective: To evaluate a pre-specified three-month threshold as a prognostic marker of cancer-specific survival (CSS) and as a potential auditable indicator of timely access to palliative care. Methods: A retrospective observational cohort study of 147 consecutive adults with advanced locoregional or metastatic cancer, admitted to a Romanian palliative care center (2018–2022; median follow-up 26 months), was conducted. The six-stratum analytical framework comprised multivariable Cox regression, Fine–Gray models, Grambsch–Therneau coefficients, Imai–Keele–Tingley mediation, restricted mean survival time (RMST) with inverse probability of censoring weighting, and validation through landmark and left-truncation analyses. Results: Fifty-nine patients (40.1%) were referred within the first three months. Referral within three months was independently associated with a reduced risk of cancer-specific death in locoregional disease (adjusted hazard ratio [HR], 0.49; 95% confidence interval, 0.28–0.86) and metastatic disease (HR, 0.58). Fine–Gray models confirmed the association (subdistribution HR 0.52). The RMST benefit was + 18.7 months (locoregional) and + 4.3 months (metastatic), persisting across landmark and truncation analyses. Conclusion: The three-month threshold is independently associated with significantly better CSS and is robust across six analytical approaches. As the design is observational, the findings remain associative and hypothesis-generating. Relevance for patients: The proportion of patients referred within the first three months can be computed from routine documentation and, subject to external validation, could serve as an auditable indicator of access to palliative care.