Mammographic screening: evidence from randomised controlled trials
BACKGROUND: All randomised breast cancer screening trials have shown a reduction in breast cancer mortality in the 'invited for mammography' screening arm compared with the 'control arm' for women aged 50 years and older at randomisation (overall 25%). However, individually published point estimates differ and concern has been raised about methodological quality and outcome measures. Materials and Methods Review of the evidence on breast cancer mortality reduction and discussion of the causes of difference in point estimates in the five Swedish and Canadian trials. A summary of the prerequisites for methodological quality and its available evidence from the trials is given. Data to support breast cancer mortality as a correct outcome measure are presented. RESULTS: There is no reason not to use breast cancer mortality as an outcome measure for trials intended to reduce breast cancer mortality, both from a clinical and a methodological point of view. Everything possible was performed in these trials in order to determine this outcome measure as accurately as possible. The fact that a few of the trials showed a relatively large breast cancer mortality reduction and others far lower reduction rates is irrelevant, if one does not consider the background situation in the region before the trial started, the design of the trial or quality of screening. CONCLUSIONS: There seems no reason to change or halt the current nation-wide population-based screening programmes. Nor is there any justifiable reason for negative reports towards women or professionals.
|, , , , , , , , , , , , , , , , , ,|
|Annals of Oncology|
|Organisation||Erasmus MC: University Medical Center Rotterdam|
de Koning, H.J. (2003). Mammographic screening: evidence from randomised controlled trials. Annals of Oncology. Retrieved from http://hdl.handle.net/1765/10194