Disaggregating the mortality reductions due to cancer screening: model-based estimates from population-based data

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Cancer screening mortality impact is typically reported as a single hazard ratio or percentage reduction.
• This aggregate measure overlooks the crucial influence of screening round frequency, timing, and the temporal distribution of averted deaths.

Purpose of the Study:

• To develop and illustrate a model that incorporates the number and timing of screening rounds.
• To disaggregate observed mortality reductions by time and screening history.
• To project the impact of alternative cancer screening regimens.

Main Methods:

• A regression-type model was developed to include screening timing and frequency as determinants of hazard ratios.
• The model was fitted to breast cancer mortality data from Funen, Denmark, considering age and follow-up year.
• Data included mammography screening invitations for women aged 50-69 every two years.

Main Results:

• A standard proportional hazards model yielded an overall hazard ratio of 0.82 (18% average reduction).
• A non-proportional hazards model incorporating timing showed reductions varying from 0% to 30%.
• Largest reductions were observed in cohorts receiving more invitations and with sufficient follow-up time.

Conclusions:

• The timing and frequency of cancer screening rounds significantly influence mortality reductions.
• Current methods may underestimate the true impact of screening by ignoring temporal factors.
• The proposed model allows for more accurate disaggregation and projection of screening benefits.