Defining Survival Epidemiology: Postdiagnosis Population Science for People Living with Disease

This article proposes survival epidemiology as a distinct branch of epidemiology focused on postdiagnosis populations, arguing that prevention-based estimates often fail to apply after diagnosis. It outlines methodological requirements, data needs, and practical implications for research, guidelines, and communication.

Raphael E. Cuomo· Journal of Clinical Epidemiology· 2025· DOI 10.1016/j.jclinepi.2025.112122

The core problem

Epidemiology has traditionally focused on explaining who becomes ill, but many clinical and public health decisions occur after diagnosis. This article introduces survival epidemiology as a conceptual and methodological umbrella for studying how people live longer and better with disease. The central argument is that estimates derived for disease prevention should not be assumed to apply in the postdiagnosis state. The purpose is to define the scope of this branch, identify the methodological and data requirements for credible postdiagnosis inference, and outline practical implications for research, guidance, and communication.

The article synthesizes evidence across cardiovascular, renal, oncologic, pulmonary, and hepatic conditions to demonstrate that associations measured for incidence frequently fail to reproduce, and can sometimes reverse, among patients with established disease. Diagnosis functions as a causal threshold that introduces distinct biases and time scales, including selection on disease (collider stratification), time-dependent confounding, immortal time bias, and reverse causation. Credible analysis therefore requires designs that emulate randomized trials, explicit alignment of time zero with clinical decisions, strategies defined as they are used in practice, and appropriate handling of competing risks, multistate transitions, and longitudinal biomarkers through joint modeling. Data rarely captured in general cohorts are essential after diagnosis, including disease stage, molecular subtype, prior lines of therapy, dose intensity, adverse events, performance status, and patient-reported outcomes.

Introduction

Epidemiology has traditionally focused on explaining who becomes ill, but many clinical and public health decisions occur after diagnosis. This article introduces survival epidemiology as a conceptual and methodological umbrella for studying how people live longer and better with disease. The central argument is that estimates derived for disease prevention should not be assumed to apply in the postdiagnosis state. The purpose is to define the scope of this branch, identify the methodological and data requirements for credible postdiagnosis inference, and outline practical implications for research, guidance, and communication.

The principal recommendations are to estimate prevention and postdiagnosis survival effects in parallel for the same exposure-disease pairs, to report effect heterogeneity by stage, subtype, treatment pathway, and time since diagnosis, and to align reporting with clinical decision points. Major conclusions are that journals should expect authors to make clear whether claims pertain to prevention or survival and to report target-trial elements; guideline bodies should, when evidence allows, distinguish prevention and survival questions separately rather than extrapolating across states; funders and training programs should prioritize methods and data standards specific to postdiagnosis inference, including survival-focused curricula and reporting guidance; and public communication should mirror this split in carefully framed, clinician-mediated messages that avoid oversimplified narratives, so that people living with disease receive accurate, context-specific advice. Aligning methods, data, and guidance around postdiagnosis realities can improve treatment tolerance, functional outcomes, and the clarity of clinical counseling.

Praktisi kesehatan, peneliti klinis, dan pembaca Research Digest.

Opening member content…