QUESTION: The authors, in an article for the JAMA section on the rational clinical examination, consider the evidence on whether and how to use clinical breast examination as a cancer screening technique.
BACKGROUND: Breast cancer is a common disease, particularly in older women. The authors note that by age 70 the annual incidence of breast cancer is one in 200 women. Breast cancer survival is strongly influenced by the stage of the disease at the time of diagnosis. As a result, it is important to consider how best to screen for this disease.
In recent years there has been considerable attention in the clinical literature and in the popular media paid to the screening strategies of breast self-examination and of screening mammography, but somewhat less to the potential role of the breast examination by the healthcare provider. In actual clinical practice, the same woman may be the recipient of any, none, or all of these screening modalities. The best way to combine these screening strategies, particularly in the case of the older woman, remains a subject of some uncertainty and controversy.
DATA SOURCES: Data were obtained from a MEDLINE search of the English-language literature for 1966 through 1997 and additional articles as identified by the authors.
STUDY SELECTION CRITERIA: In their evaluation of the effectiveness of clinical breast examination, the authors included both controlled trials and case-controlled studies in which clinical breast examination was used as a component of the screening. Study of breast examination technique considered both clinical studies and studies using silicone breast models.
DATA EXTRACTION: The combined data from the trials included information on approximately 200,000 women who received a breast cancer screening intervention (mammography and/or clinical breast examination). However, none of the studies made the direct comparison of a group receiving clinical breast examination as a sole intervention with a control group that did not receive any screening. Data on the utility of clinical breast examination were partially derived from studies where that screening modality was used in combination with mammography.
MAIN RESULTS: A number of trials of cancer screening have demonstrated a reduction in mortality from the use of mammography and clinical breast examination as combined screening strategies compared with no screening, with the inference that the reduction in mortality comes from the earlier detection of breast cancer. The percentage of the detected cancers that are detected in the trials by clinical breast examination despite having been missed on mammography varies across the trials from a low of 3% of the detected cancers to a high of 45%. It is speculative whether the marginal contribution of clinical breast examination to the mortality reduction in these screening trials corresponds to the percentage of cancers detected by clinical breast examination alone.
In most of the clinical trials, the technique of breast examination reportedly was not well described. It is unclear therefore how much the technique of breast examination used varied within and among the clinical trials. Data from studies using examinations of breast models made of silicone demonstrated that test performance accuracy correlated with a lengthier breast examination, better breast examination technique, and perhaps with examiner experience.
The report includes data from six comparator studies and from two demonstration projects. Of the six comparator studies, four compared a screened population with an unscreened population and two compared different intensities of screening strategies. None of the eight clinical trials was directed to a geriatric population and in fact older women were excluded by upper age entry criteria from the six comparator studies. (The upper age limit for study entry in the six comparator studies varied from 49 to 64.)
CONCLUSION: The authors drew on the pooled results of these eight studies to conclude that clinical breast examination has a sensitivity of 54% (95% confidence interval, 48.3–59.8) and a specificity of 94% (95% confidence interval, 90.2–96.9). The authors conclude that screening clinical breast examination should be done for women age older than 40.