Clinical breast examination. A contentious issue in screening for breast cancer

Aust Fam Physician. 2000 Apr;29(4):343-6; discussion 348.


Background: Breast cancer is the commonest cancerous cause of mortality in Australian women. It is now well established that mammography is effective in reducing this mortality through screening. To this end all states of Australia have established breast cancer screening programs using mammography as their investigative modality. The purpose of this review is to determine the value of clinical breast examination as another screening tool available to general practitioners.

Objective: We identified and examined a large amount of literature relating to this subject through Medline search, the Cochrane Library and hand searching (using keywords 'breast neoplasm', 'screening', 'breast self examination' and 'clinical breast examination'). The review of articles was undertaken by both authors, one of whom has extensive experience as a breast physician.

Discussion: Clinical breast examination for symptomatic women shows a true positive rate or sensitivity of 85%, a false positive rate of 20% and specificity of 80% for detecting the probability of cancer. The positive predictive value of this test is 81% and the negative predictive value is 84%. The sensitivity would be lower for asymptomatic women.

Conclusion: Much has been written about screening for breast cancer. However, the absence of comprehensive studies leaves the evidence for CBE somewhat in dispute. In the absence of any harm from the procedure, it is difficult to support the currently accepted Australian evidence based guidelines which discourage the practice of CBE. It would seem more prudent given the disagreement in the literature and the acceptability of CBE to both GPs and patients to consider that there is value in using CBE as a screening tool.

MeSH terms

  • Adult
  • Aged
  • Breast Neoplasms / diagnosis*
  • Breast Self-Examination*
  • Evidence-Based Medicine
  • Female
  • Humans
  • Middle Aged
  • Reproducibility of Results