Timeliness of lung cancer care and area-level determinants in Victoria: A Bayesian spatio-temporal analysis.

Abstract

Background: It has been reported that the timeliness of lung cancer care varies significantly across different regions. According to the Victorian Lung Cancer Registry (VLCR) report, the timeliness of lung cancer care in Victoria has changed over time. Therefore, we aimed to quantify the extent of these spatial inequalities over time and to identify area-level determinants contributing to these changes.

Methods: The study analysed lung cancer cases reported to the VLCR between 2011 and 2022. Bayesian spatio-temporal Conditional Autoregressive (CAR) models were fitted, incorporating spatial random effects, temporal random effects, as well as spatio-temporal interactions. The best-performing model was selected using the Deviance Information Criterion (DIC). For the final best-fit model, the adjusted Relative Risks (aRR) and their 95% Credible Interval (CrI) were reported.

Results: Over half (51.24%) of lung cancer patients experienced treatment delays, while approximately one-third (30.98%) encountered diagnostic delays. Moderate spatio-temporal variations were observed in both delayed diagnosis and treatment. In the final best-fit model for treatment delay, an increase in the percentage of smokers was significantly associated with a higher risk of treatment delay (RR = 2.13, 95% CrI: 1.13, 4.20).

Conclusions: Identifying high-risk areas provides useful information for policymakers, helping in the reduction of delays in lung cancer diagnosis and treatment.

Impact: This study has revealed spatio-temporal inequalities in diagnostic and treatment delays, providing valuable insights for identifying areas that should be prioritized to ensure timely care for lung cancer.