Quality Assurance

Quality assurance (QA) in mammography is defined as those planned and systematic activities that monitor and improve the early detection of breast cancer and the evaluation of breast disease. Those activities include the employment, training, and continuing education and experience of qualified personnel. They also include the selection of appropriate mammography equipment, acceptance testing and regular evaluation of equipment performance, and the evaluation of positioning and compression. QA also includes the evaluation of patient interactions, reporting of results, diagnostic accuracy, patient tracking, and follow-up.

QA activities may be subdivided into two major categories: quality-control (QC) procedures and quality administration procedures. QC includes the technical components of QA: equipment selection, equipment performance evaluation and routine equipment monitoring, technique factor selection, and evaluation of breast positioning and compression. Quality administration includes monitoring methods that assess interactions and communications between the mammography provider and the patient, and between the interpreting physician and the referring physician. Quality administration also includes steps that assess the skills of the interpreting physician by comparing screening or diagnostic results with patient outcomes and other administrative monitors of quality.

6.1 The Current Status of Quality Assurance in the United States

During the 1980s, the quality of mammography improved through the replacement of conventional x-ray units used for mam-mography by dedicated mammographic units and by the improvement of image receptors designed specifically for screen-film mammography (Bassett et al., 1992). During the mid-1980s, it was commonly believed that the use of a dedicated mammography unit with appropriate screen-film image receptors was adequate to insure high-quality mammographic images at low radiation dose.

During the latter half of the 1980s, several studies revealed that the use of dedicated mammography equipment alone was insufficient to insure the production of consistently high-quality images at uniformly low radiation doses. The Nationwide Evaluation of X-Ray Trends conducted in 1985 (NEXT-85) evaluated radiation dose and image quality at 232 mammography sites in the United States. The NEXT-85 study found a wide variation in image quality and radiation dose from site-to-site (Conway et al., 1990; Reuter, 1986). Similar results were found in a survey of 29 dedicated screen-film mammography sites in the Philadelphia area in 1986 (Galkin et al., 1988). The study found that the film processors at 41 percent of sites varied in film mid-density by more than ±0.10 over a 15 d period, suggesting that short-term processor variations might be a common source of variation in mammographic image quality.

Data collected during the first six months of the ACR-MAP, which began in August 1987, confirmed the wide variations in image quality and dose observed in the NEXT-85 study (Hendrick, 1990; Hendrick et al., 1987). Data collected as part of ACR-MAP site applications indicated that most sites were not performing QC tests at adequate frequencies. For example, on ACR applications collected during 1987 and 1988, approximately one-half of sites claimed to perform daily processor sensitometry and less than one-third of sites claimed to perform, at least, monthly evaluation of image quality using a phantom (Hendrick et al., 1998). Data collected from ACR-MAP applicants over the first 6 y of the program indicated increased performance and improved performance frequencies of QC tests at mammography sites. In 1992, 88 percent of sites stated that they were performing daily processor sensitome-try and 61 percent of sites stated they were performing, at least, monthly evaluations of image quality using a phantom (Hendrick et al., 1998). A 9 y study of film processing in radiology conducted by the Center for Devices and Radiologic Health found a high rate of underprocessing among hospitals (33 percent in 1987) and private practices (42 percent in 1989), but a surprisingly low rate of underprocessing among mammography sites (seven percent in 1988). This improved and significantly lowered the rate of poor processor performance in mammography (seven percent underpro-cessing in 1988 versus 18 percent underprocessing in 1985) was attributed to increased attention to QC practices at mammography sites (Suleiman et al., 1992).

The improvement in QC practices over time can be attributed to a number of factors including the advent of ACR-MAP in 1987, publication of the ACR Mammography Quality Control Manual (ACR, 1999); publication of AAPM Report No. 29 on Equipment Requirements and Quality Control for Mammography (AAPM, 1990); the ratification of the ACR Standards for the Performance of Screening

Mammography (ACR, 1990b); the ACR-MAP requirement that accredited mammography sites perform QC tests according to the ACR QC manuals beginning in January 1992, and the passage of MQSA in October of 1992 and its subsequent implementation.

Even though QC practices have improved over the last decade and a half, there is still room for improvement.

Quality administration of mammography practices is a newer concept than QC. While effective quality administration has been conducted and described in several model mammography practices (AHCPR, 1994; Bird, 1989; Linver et al., 1992; Sickles, 1990; 1992b), it is more recently becoming widespread among United States mammography sites. Prerequisites to an effective quality administration program are standardized reporting and recording of mammography results, a patient follow-up system, and a method to monitor outcomes of screened women or patients who receive both positive and negative results.

Comparison of screening results among sites, additionally, requires similarity of screened populations and a reporting and monitoring system standardized across mammography sites. Standardized reporting systems and computerized monitoring and follow-up systems have recently been introduced in the United States mammography market (ACR, 1998; Kopans, 1992a), but data comparing identically compiled results from different sites using a standardized reporting and tracking system have not yet been reported.

6.2 Essential Elements of an Effective Quality-Control Program

An effective QC program should begin with the selection of appropriate equipment for mammography and the use of qualified personnel, including the interpreting physician, radiologic technologist, and medical physicist, each of whom must participate actively in mammography QC.

An interpreting physician experienced in mammography should be designated to oversee, monitor and motivate the QC program at each mammography site. A radiologic technologist who is experienced in mammography and trained in mammography QC should be designated as the mammography QC technologist, performing the regular technologist QC tests. One technologist should be designated so that tests are performed consistently; the primary QC technologist should then train another technologist to perform the tests in a similar manner when the primary QC technologist is absent. A medical physicist experienced in mammography and mammography QC should perform acceptance testing of new mam-mography equipment, perform annual equipment surveys, and review the site's ongoing QC program and records.

6.2.1 Selection of Mammography Equipment

The type of equipment used for mammography is crucial to obtaining images of consistently high quality (see also Section 2). A dedicated x-ray unit, designed specifically for mammography, is a requirement for both screening and diagnostic mammography. The unit should be equipped with low-attenuation parallel-plate compression devices, a foot-activated motorized compression drive, image-receptor holders and removable grids for both 18 x 24 cm and 24 x 30 cm image-receptor sizes, and AEC. The x-ray generator should be capable of <10 percent kilovolt peak ripple (<20 percent exposure ripple) to minimize excess patient dose. The system should be capable of generating an x-ray output of at least 7 mGy air kerma (800 mR) per second at the entrance surface of the breast in contact mode at 28 kVp. The system should be able to sustain this radiation output rate for at least 3 s (ACR, 1993). For diagnostic mammography, the system should have both large and small focal spots, and the small focal spot should be used for magnification mammography. Diagnostic mammography equipment should be equipped to obtain coned, compressed views in magnification mode.

Rather than specifying maximum focal-spot sizes, which has been traditional but problematic for measurement, there is growing consensus that system specifications should be given in terms of the limiting spatial resolution of the system. For example, a suggested performance specification is that the limiting spatial resolution should be measured using a high-contrast resolution bar pattern oriented parallel to the plane of the image receptor, centered left-to-right and at the chest wall, and 4.5 cm above the breast support surface. In this location, the limiting spatial resolution should be no less than 13 cycles mm1 (lp mm1) with the pattern oriented with bars parallel to the anode-cathode axis, and no less than 11 cycles mm1 (lp mm1) with the pattern oriented with bars perpendicular to the anode-cathode axis. Such a performance specification would eliminate the difficulties inherent in accurately measuring focal-spot sizes using different devices and different measurement methods (ACR, 1993; Kimme-Smith, 1992).

More detailed specifications for mammography x-ray units are available in two documents [i.e., AAPM Report No. 29 lists both general and specific mammography equipment requirements (Yaffe et al., 1990), and the ACR Recommended Specifications for New Mammography Equipment lists recommended specifications for newly manufactured screening x-ray equipment (ACR, 1993; Yaffe et al., 1995)].

6.2.2 Selection of Screens and Films

Fluorescent screens and films used in mammography should be those designed specifically for mammography. Screens and films should be matched to one another for spectral characteristics. Typically, green light-emitting screens are used with green lightsensitive films in mammography. Single-emulsion films should be used only with single-screen cassettes, with the emulsion of the film facing the fluorescent screen. However, higher speed receptors may be useful for magnification mammography where their faster speed helps reduce the effects of patient motion and breast radiation dose, both of which tend to be greater in magnification mam-mography. Additional information and recommended specifications for mammography screen-film image receptors are available in the published literature (Law and Kirkpatrick, 1989; 1990) and in the ACR Minimum Specifications for Mammography Image Receptors (ACR, 1993; AHCPR, 1994).

Screen-film combinations used for mammography should be capable of achieving a limiting spatial resolution (0.05 MTF) of at least 15 cycles mm1 (lp mm1). Many current screen-film combinations can achieve limiting spatial resolutions of approximately 20 cycles mm1 (lp mm1) or more.

6.2.3 Selection of Film-Processing Conditions

Mammography films should be processed in a processor suitable for, or designed specifically for, mammography and processed under conditions optimized for the mammography film. The processor should be operated with the chemistry, replenishment rate, processing time, and temperature specifically recommended by the film manufacturer for the film used. Typically, higher replenishment rates are required for mammography films due to the higher densities attained (ACR, 1993). A prolonged delay between film exposure and processing is not desirable due to loss of resultant film speed, a loss that can range from 6 to 46 percent over a 24 h period, depending on the mammography film (Kimme-Smith et al., 1991). If delayed processing is necessary, the length of delay should be kept constant and as short as possible (ACR, 1993).

6.2.4 Quality-Control Procedures

Regular quality-control (QC) procedures are essential to ensuring consistent mammography equipment performance. QC procedures may be subdivided into those tests conducted by the mammography QC technologist and those procedures conducted by the medical physicist.

Procedures that should be conducted by the QC technologist, recommended minimum frequencies, and the purpose of each test are listed in Table 6.1. Complete descriptions of the technologist QC tests are contained in the ACR Mammography Quality Control Manual (ACR, 1999) including required test equipment, step-by-step procedures, data recording forms, action limits, and corrective actions that should be taken if action limits are exceeded (ACR, 1999). Each site should have a technologist conducting QC tests according to the technologist's tests in the ACR (1999) manual.

Quality control (QC) requires consistent monitoring of quality. It is important to continue QC testing even if problems do not occur in the first few months of testing. QC is not just monitoring of quality, but also includes identifying that systems are "out-of-control" and taking appropriate actions when problems are identified. It is especially important that appropriate actions be taken when action limits are exceeded and before image quality and patient safety are compromised.

QC procedures that should be conducted by the medical physicist are listed in Table 6.2, and correspond to the tests for the medical physicist listed by ACR (1999). These tests should be conducted annually or after major equipment changes, including relocation of fixed mammography equipment. An independent evaluation of mammography image quality and artifacts by the medical physicist is important, since the medical physicist is in a position to compare image quality among a number of mammography sites. The medical physicist should also review the procedures and records of the technologist's QC tests at least annually and preferably on a more frequent basis, such as quarterly.

The ultimate responsibility for QC at each mammography site rests with the interpreting physician, who should take an active role in motivating, supporting and overseeing the QC activities of the radiologic technologist and medical physicist. The interpreting physician should insure that appropriately qualified and trained people are chosen for these important jobs, that adequate time and test equipment are made available to the QC technologist, and that QC records are properly maintained. The interpreting physician should review the QC records and reports of the QC technologist at

Table 6.1—QC tests for

the technologist listed in ACR QC manual (ACR, 1999).

QC Test

Frequency

Purpose

Processor

Daily

To insure consistent film processing

Darkroom cleanliness

Daily

To minimize film artifacts caused by dirt and dust

Mobile unit

Daily

To insure consistency

Screen cleanliness

Weekly

To free cassettes and screens of dirt and dust

Viewboxes and viewing conditions

Weekly

To insure that film viewing conditions are appropriate

Phantom images

Weekly

To insure that film density, contrast, uniformity and image quality are adequate

Visual checklist

Monthly

To insure that x-ray system lights, displays, locks and detents work properly

Repeat analysis

Quarterly

To determine the numbers and causes of repeated mammograms

Analysis of fixer retention

Quarterly

To determine the residual fixer in processed film, as a measure of film storage life

Darkroom fog

Semi-annually

To insure minimal film fogging

Screen-film contact

Semi-annually

To insure that each cassette maintains adequate contact between screens and film

Compression force

Semi-annually

To insure that motor-driven compression yields adequate, but not excessive, breast compression force

to o

0 0

Post a comment