Deviation Between Incidence Rates of Screenable Cancers and the Proportion of Late-Stage Disease

Screening for cancer decreased during the COVID-19 pandemic, leading to fewer cancer cases and less than expected early-stage diagnoses. Due to this decrease, the subsequent literature suggested a tidal wave of increased incidence and a larger proportion of late-stage disease.

The National Cancer Database (NCDB) was used to identify screenable cancer cases from lung, female breast, and colorectal cancer from 2010 to 2023. These were standardized to calculate annual incidence rates per 100,000. A linear regression model of 2010-2022 incidence rates was used to calculate predicted incidence compared with observed incidence in 2023 (most recent NCDB data). Late-stage disease proportions (III/IV versus I/II) were also compared. Sensitivity analyses were conducted using alternative modeling periods and restricting to stable facility participation in NCDB.

A total of 2,235,175 lung, 3,090,471 breast, and 1,448,861 colorectal cancer patients were analyzed. In 2023, the observed incidence was lower for lung 63.282 versus 68.914 (95% predictive interval 62.677-75.151), breast 174.055 versus 193.417 (95% PI 178.613-208.221), and colorectal 43.348 versus 45.517 (95% PI 41.266-49.769), although only breast was significantly lower. Observed and predicted late-stage disease proportions were similar for all three screenable cancers: lung: 56.0% versus 59.0%, breast: 11.6% versus 11.6%, and colorectal: 55.6% versus 56.8%. Findings were similar in sensitivity analyses.

Despite the broadly accepted expectation that the incidence of screenable cancers would increase after the initial wave of the COVID-19 pandemic, the rates decreased, particularly for breast cancer. Furthermore, the expectation that late-stage cancers would rise significantly was not observed.