798-799 / 1676
NTCP-models
Problems of available literature
1.
Retrospective assessment
prospective
–
Underestimation
–
Interpretation of available data
2.
Lack of multivariable models
multivariable
3.
NTCP-models based on RCT’s
cohort study
–
Highly selected (only: 1-5%)
4.
Evolution over time
follow up schedule
5.
Uniformity OAR definition
guidelines
6.
Small sample sizes
population based cohort