1597 / 1816
To produce a clinically useful dose-response
relationship we need:
•
Large studies of women in the general population
–
patients who have comorbidities or complicated
problems are not usually entered into trials
•
Long follow-up
•
Data on non-fatal as well as fatal events
•
Sufficient detail on RT to calculate cardiac doses
•
Information on pre-existing risk factors for heart
disease