Predicting the effectiveness of the Finnish population-based colorectal cancer screening programme

Area of Science:

• Oncology
• Public Health
• Biostatistics

Background:

• Colorectal cancer (CRC) screening effectiveness evaluation requires long-term follow-up due to its long natural history.
• A temporal multi-state natural history model offers an alternative for predicting CRC screening effectiveness.

Purpose of the Study:

• To demonstrate the utility of a temporal multi-state natural history model for predicting CRC screening effectiveness.
• To estimate the impact of CRC screening on mortality and advanced CRC reduction.

Main Methods:

• Utilized a Markov model to analyze data from the Finnish population-based biennial CRC screening program (2004-2007).
• Estimated pre-clinical incidence, mean sojourn time (MST), and faecal occult blood test (FOBT) sensitivity.
• Simulated the effects of five screening rounds on advanced CRC and mortality reduction in a cohort of 500,000 individuals aged 60-69.

Main Results:

• Mean sojourn times (MST) for localized and non-localized CRC were 2.06 and 1.36 years, respectively.
• FOBT sensitivity estimates were 65.12% for localized and 73.70% for non-localized CRC.
• Predicted relative risk reduction for non-localized CRC and CRC death was 0.86 and 0.91, respectively, in the screened group.

Conclusions:

• The developed model predicted a 9% reduction in CRC mortality.
• The model also predicted a 14% reduction in advanced CRC based on preliminary Finnish screening program results.