ESTIMATING AND COMPARING CANCER PROGRESSION RISKS UNDER VARYING SURVEILLANCE PROTOCOLS.

Area of Science:

• Biostatistics
• Cancer Research
• Epidemiology

Background:

• Cancer outcome data is typically collected at discrete time points, posing analytical challenges.
• Comparing patient cohorts with differing surveillance frequencies can lead to confounded results.
• Accurate risk assessment is crucial for effective cancer management and research.

Purpose of the Study:

• To develop a statistical framework for analyzing cancer outcomes observed at discrete times on a continuous scale.
• To address confounding factors in comparing cancer progression risks across cohorts with varied surveillance schedules.
• To accurately assess prostate cancer progression risks in active surveillance cohorts.

Main Methods:

• Utilized multistate and hidden Markov models.
• Represented events on a continuous time scale from discrete observation data.
• Applied the framework to compare prostate cancer progression risks across multiple active surveillance cohorts.

Main Results:

• The developed framework successfully models cancer events over continuous time.
• Differences in surveillance frequencies partially explained observed variations in progression risks between cohorts.
• Underlying cancer progression risks differences persisted across cohorts after accounting for surveillance frequency.

Conclusions:

• The statistical framework provides a robust method for analyzing time-to-event data with discrete observations.
• Accounting for surveillance schedules is essential for accurate cross-cohort comparisons of cancer progression.
• This approach enhances the reliability of findings in cancer research involving heterogeneous data collection schemes.