How effective are breast cancer screening programmes by mammography? Review of the current evidence

Abstract

Randomised controlled trials showed that breast cancer screening by mammography reduces breast cancer mortality in women over age 50 by 25–30%. However, it was not clear if this effect would persist outside the controlled trial environment or even could be enhanced.

We review the current evidence of the impact of long-standing breast cancer screening programmes (Australia, Canada, Denmark, Finland, Iceland, Italy, the Netherlands, Spain, Sweden and the United Kingdom) on breast cancer mortality.

The decrease observed in women invited to screening ranges from 16% to 36%. Breast cancer mortality reductions range from 24% to 48% in women having attended at least one screen after correcting for selection bias. Although evaluation design, time period studied, participation rates achieved differ, the trend in mortality reduction is consistent. Adjuvant therapy is estimated to contribute about one third to this decrease. We conclude that mammography screening programmes implemented for at least 10 years achieve a similar, but not greater mortality reduction as the randomised controlled trials. However, it may take some more years before the full impact of these mammography screening programmes can be assessed.

Introduction

Based on the results of eight randomised controlled trials studying the efficacy of mammography screening to reduce breast cancer mortality and numerous meta-analyses, showing a 25–30% decrease in breast cancer deaths, several industrialised countries started implementing national mammography screening programmes in the 1980s and 1990s. A Cochrane review published in 20011, 2, 3 challenged the results of some of the randomised controlled trials leading to a major controversy on the benefits of mammography screening.^c Several extensive reviews then re-asserted that randomised controlled trials had shown that breast cancer screening by mammography in women age 50–69 at regular intervals can reduce breast cancer mortality.5, 6, 7, 8 Furthermore many papers were published in 2002 and 2003 that countered specific issues raised in the Cochrane review.9, 10, 11, 12, 13, 14, 15, 16, 17

However, it was not clear how much of this effect would persist outside the controlled trial environment. A first review of the effectiveness of mammography screening programmes was done at an international level by IARC in 2002.¹⁸ Analysing studies on the impact of service screening mammography published between 1990 and 2004, another review concluded that breast cancer mortality reductions observed were consistent with those observed in the randomised trials.¹⁹ In 2006, the Advisory Committee on Breast Cancer Screening in England considered efficacy of mammography screening to be underestimated in the randomised controlled trials.²⁰ In the same year, a health technology assessment requested by the Quebec Ministry of Health (Canada) re-examined the scientific evidence on which screening mammography programmes are based.²¹ This extensive review also concluded that modern breast cancer screening programmes may achieve greater reduction in breast cancer mortality than was found in the screening trials, without providing conclusive evidence.

This paper reviews the recent evidence on the impact of breast cancer screening programmes on breast cancer mortality. We focus on programmes evaluated after at least 10 years of implementation, as statistically significant results of breast cancer screening can be expected 10–15 years after the introduction of the screening.