For nearly one hundred years, chest roentgenography has been an important tool for the recognition, investigation, and evaluation of occupational lung disease. As the technique became more formalized, it became apparent that readers often disagreed on the presence, type, and extent of lung disease observed among dust-exposed workers (1). These observations prompted efforts to refine the radiographic film-reading process. A standardized scoring procedure was developed to document the various types and degrees of dust-induced abnormalities; this became the International Labour Office (ILO) classification (2). A process began of selecting radiographs as standard comparison films representative of certain abnormalities (3). Investigators identified factors that affected the reliability of readings, such as image quality (4), and recommended procedures to reduce variation in readings (5). With progressive improvement in techniques and procedures over the decades subsequent to its introduction, the ILO classification system emerged as a remarkably robust mechanism for assessing occupational disease. Classifications of radiographs show clear correlations with dust exposure, lung dust burden, lung pathology, and mortality (6-8). However, the full power of the ILO system is only realized with appropriate procedures. For example, in the Coal Workers' X-ray Surveillance Program, administered by the National Institute for Occupational Safety and Health (NIOSH), there are strict requirements for the x-ray film, exposures, and equipment to be used. Additionally, before films may be submitted under the program, sample images from each radiography unit must be evaluated and approved by NIOSH. ,To reduce the effect of between-reader variation, all final pneumoconiosis determinations for the Program are based on agreement of 2 or more readers, using a specified algorithm. Furthermore, to assure that physician readers are trained and proficient in the classification of dust-related changes in chest roentgenograms, a system of professional certification, the NIOSH B reader program, was established (9,10). The B-reader program has played a major role in advancing knowledge of radiograph classification in the United States. The program is intended to assist physicians interested in increasing their knowledge of the pneumoconiosis and related diseases, and in demonstrating their proficiency in reading radiographs using the ILO system. The program provides each candidate an opportunity to review the NIOSH self-study syllabus, and in partnership with the American College of Radiology, periodically offers comprehensive symposia on the radiology of the pneumoconioses (II). The B-reader certification examination is a rigorous evaluation of the physician's capability, at a single point in time, to identify and appropriately categorize radiographic changes of dust-related lung disease. It requires accurate categorization of a set of 125 test films within a period of 6 hours; historically, only about 50% of candidates are successful in passing the examination. There is a similar periodic recertification process. NIOSH has demonstrated ongoing interest in improving and updating the B-reader program to accommodate new developments, such as the advent of computed radiography, through stakeholder workshops, including those held in 1990 (12) and March of 2004. Analogous to a specialty board certification, the B-reader program is one of the central components of a comprehensive approach to assuring quality in obtaining and evaluating chest-radiographic images among dust-exposed workers. However, it does not provide for an ongoing assessment of a reader's performance during the normal course of professional or scientific activities. In contrast, the regulation and enforcement of the competent and ethical practice of medicine is the province of state medical licensing boards.
Lee Petsonk, Surveillance Branch, Division of Respiratory Disease Studies, National Institute for Occupational Safety and Health, Morgantown, WV 26505