Here, one would also evaluate the sensitivity and specificity of screening mammography. The document also established goals for many of these outcome parameters as a guideline for facilities, noting, however, that outcome data vary based on the size of the practice or number of audits in the database and by the underlying characteristics of the population being screened. Although a more complete review of the document is not within the scope of this chapter, the interested reader is urged to review this set of guidelines.
The final and arguably most critical step toward quality mammography is interpretation. A major barrier toward this educational goal had been the absence of any standardized language for mammographic features. Through the efforts of the ACR, with support from various clinical colleges, including the American College of Surgeons (ACS) and the American College of Obstetrics and Gynecology, a multidisciplinary committee was established in the early 1990s to address the suboptimal and often confusing terminology used in mammography reports. The document produced by this committee is the Breast Imaging Reporting and Data System (BI-RADS). BI-RADS is designed to standardize mammographic reporting, reduce confusion in breast imaging interpretations, and facilitate outcome monitoring. The system is divided into four sections:
- Breast imaging lexicon. An illustrated review of all the findings seen on a mammogram with suggested terminology as well as guidance regarding whether the findings are worrisome for malignancy, are probably benign, or are definitely benign.
- Reporting system. Provides an organized approach to image interpretation and reporting. Several sample reports are included to illustrate the suggested format for different mammographic scenarios.
- Follow-up outcome monitoring. Describes minimum data to be collected and used to calculate important audit measures, allowing each radiologist to assess his or her overall performance in mammography interpretation.
- ACR National Mammography Database. A preliminary effort for collection of national data that will ultimately allow groups to adjust their thresholds by comparison with pooled national data.
There are six categories in BI-RADS that cover all possible initial and summary interpretations. Category 0 is “assessment is incomplete.” This category should be used on reports of screening mammograms that require further evaluation. After this evaluation is complete or in the case of screening examinations that do not require assessment, there are five final categories. Category 1 is a “negative” examination. It does not require comment, and the patient is encouraged to return for routine screening in 1 year. Category 2 is a “benign finding.” This is also a negative examination, but the interpreting physician may wish to describe a benign feature for subsequent reviewers. The patient is encouraged to return for routine screening in 1 year. Category 3 is for a “probably benign finding.” This category is reserved for findings that have an extremely low probability of malignancy (i.e., 2% or less). It is not intended as an intermediate category between benign and malignant, but rather to denote a finding that has such a low probability of malignancy that a short-term follow-up, usually 6 months, is more appropriate than further diagnostic interventions.
The interpreting physician, referring physician, and patient all have an interest in assuring compliance with this recommendation, because neglecting the short-term follow-up results in adverse consequences in a small percentage of cases. The next two categories are those that warrant a biopsy. Category 4, “suspicious abnormality - biopsy should be considered,” is for a lesion that has a definite probability of being malignant; category 5, “highly suggestive of malignancy - appropriate action should be taken,” is for a lesion that is very characteristic for malignancy.
At a time after significant progress has been made in lowering dose and improving image quality, a greater understanding of factors associated with higher mammographic accuracy should be a priority for greater data collection and research. With the advent of a standardized lexicon and a more complete understanding of the benefits of a medical audit, obtaining data on the status of interpretation in the United States is a necessary first step. Successful methods of education focused on interpretation skills must be developed and tested both in initial small groups and subsequently in actual practice. Finally, we must develop outcome measures of interpretive performance and identify where weaknesses exist.
Provided by ArmMed Media