Diagnostic and Treatment Intervals for Colorectal Cancer and Associations With Overall Survival and Disease Stage: An Australian Data Linkage Study.

INTRODUCTION Timeliness in colorectal cancer diagnosis and treatment is crucial, however recommendations are often not evidence based. This study aimed to assess associations between the length of diagnostic and treatment intervals and overall survival and stage of disease in colorectal cancer using Australian data. PATIENTS AND

Methods

Data from multiple sources were probabilistically linked to ensure coverage across the healthcare system. Data sources included 2 general practice electronic medical record databases, a hospital-based colorectal cancer clinical registry from 3 health services, and health service administrative data from 3 hospitals. Overall survival was assessed using multivariable Cox proportional hazards regression models; risk of advanced stage was assessed using multivariable logistic regression. Restricted cubic splines were used to allow for nonlinear associations.

Results

Six discrete intervals could be assessed, and 4 had sufficient samples for analyses of outcomes (range 87-9049). Three models, assessing the diagnostic and the treatment intervals showed a nonlinear association with both very short and long intervals associated with poorer outcomes. The nadir of these curves was approximately 150 days for the diagnostic interval for survival and risk of advanced stage, and 28 days for the treatment interval for survival.

Conclusion

There is an association between time taken to diagnose and treat colorectal cancer and poorer outcomes, consistent with international studies using methods to account for nonlinear associations. The U-shaped association between time and outcomes can be used to guide maximum waiting time targets and confirms the need for early detection interventions as a strategy to improve outcomes.