Issue Tracking for Image Quality Improvement

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
Paul Nagy, PhD, is director of quality and informatics research and associate professor of radiology at the University of Maryland School of Medicine, Baltimore. On November 27, 2007, he presented Developing the Infrastructure (Quality Control in Radiology) at the annual RSNA meeting in Chicago, with the stated goal of helping his audience understand how lack of communication in radiology can be decreased through the use of the appropriate information technology. Most of the time, Nagy says, the improvement of quality in radiology lies within the radiologist’s sphere of influence, even though this activity has often fallen to others. Operational quality is frequently the responsibility of administrative personnel, and image quality can easily become the technologist’s domain. Radiologists, however, are uniquely qualified to help improve quality in both realms by bringing all concerned together as a team through enhanced communication. Technological Challenges Despite its many advantages, information technology has had an inadvertent but detrimental effect on image quality in many settings. Nagy explains that this has taken place because the technology is so good at improving efficiency and productivity that the waiting times that were once common in radiology have largely disappeared. Unfortunately, it was during these intervals that many of the steps needed for quality control and problem correction were taken, typically in the form of casual conversations between staff members and radiologists. For example, Nagy says, there are four primary areas in which digital technologies have had adverse effects on image quality. First, many information-handling technologies in radiology, including PACS, were initially introduced with assertions that they required no designated quality-control changes or mechanisms, possibly with the assumption that image quality would somehow regulate itself once image distribution and access were simplified. Second, technologists have been under increasing levels of pressure to be more productive since their enforced waiting times largely disappeared, so they have had little time available to review image quality. Third, the interaction between radiologists and technologists has been streamlined, so the opportunity for radiologists to give technologists feedback concerning image quality has been squeezed; at times, this has convinced some technologists either that they are doing perfect work or, if not, that the radiologists no longer care. Fourth, the senior technologist, in many practices, was once physically positioned near the film processor and was checking image quality throughout the day as a matter of course, acting on problems as they arose. Where there is no longer a film processor, this ongoing quality-control mechanism has also disappeared. Turning the Tables Nagy points out that information technology can also solve the quality-control problems that it created. Of course, there have always been mechanisms for reporting quality problems in digital radiology departments, but they have often been too cumbersome for radiologists to use. Frequently, they have also lacked feedback mechanisms and resolution follow-up procedures, making radiologists feel that they were not worth the time needed to pursue them. Now, however, an open-source, Web-based tool for issue tracking is available to run on computers from servers to laptops. Radtracker, written in 2001, has been downloaded more than 1,900 times and is available at no cost at Working within PACS, an issue-tracking tool such as Radtracker adds each problem to its database, notifies the parties involved (including sending messages to technologists’ text pagers), tracks the action taken to solve the problem, notifies the submitter when the issue has been resolved, and performs basic reporting functions that let users analyze the root causes of problems. Experience Nagy’s department implemented a Radtracker issue-reporting system incorporated into its PACS in September 2006. Before that time, about 10 image-quality problems per month were being reported; now, about 300 per month are handled. The implication is not that there are now more quality problems, but that problems are no longer being ignored. While an image is being viewed, the radiologist can click on a resident button that automatically launches a Web page for quality-control reporting. Later, the same button can be clicked to learn what action has been taken to resolve the problem. The ability to follow the improvement process makes it worth the radiologist’s time to report problems; in turn, this feedback helps technologists and other staff improve positioning and procedures on a continuing basis. During a year’s experience with Radtracker, Nagy reports, 2,472 quality-control problems were reported for 292,000 procedures. Within an hour, 40% of all reported problems had been resolved. Of total problems, 75% were assigned a root cause that could be addressed on a systematic basis. All the problems that were reported accounted for single-digit percentages of the total, so the department did not face any overwhelming challenges. Poor patient positioning generated the most reports, followed by order–examination mismatching, incorrect protocols, incomplete examinations, errors in patient demographics, missing images, images placed with the wrong examination records, incorrect markers, or mistaken accession numbers. Nagy describes the department’s changes in quality improvement as a data-driven metamorphosis. It took a month to convince radiologists to report problems, another month to prompt supervisors to resolve problems, and a third month to get them to do so promptly. Assignment by supervisors of a root cause (and, if applicable, a technologist) for each problem took three months. The final step, in which supervisors use issue resolution to manage customer relations, was achieved at the six-month point. Nagy notes that the system is also helpful for giving positive recognition and is used for employee performance appraisals. In fact, providing recognition for an excellent study is the seventh most frequent reason for submitting an exception report. The system functions, as well, as a knowledge base of common imaging errors that can be useful for training. Exception reporting will not take place unless it is embedded in clinical workflow, Nagy says. Whether a practice uses Radtracker or another tool, the improvement of quality in a digital department must emphasize six steps. The effort must first focus on data, not staff emotions, to determine where problems originate. Next, it must analyze the overlapping spheres of influence involved so that the right people can address the identified problems. It must find the best practice, wherever it exists, to replace the inadequate one. A task that causes problems must be split into its constituent elements for evaluation; then, components that add no value should be eliminated. Last, Nagy calls sunlight the best disinfectant: unless a problem is seen clearly by all involved, it cannot be eradicated.