Setting a Clearer Standard: Quality Assurance in Diagnostic Imaging

Quality assurance in diagnostic imaging By Jean Plummer Since the release of the Institute of Medicine’s report in 1999 To Err Is Human, concerns regarding medical errors, patient safety and quality care have increased. Consumers are becoming more educated and involved in healthcare decisions, and expect higher levels of quality and service for their dollar. As healthcare delivery systems aim to reduce medical errors and improve overall patient care, there has been a significant amount of effort invested in measuring and enhancing both quality and consistency throughout systems. The radiology market continues to be increasingly competitive, which prompts imaging service providers to deliver a better product to their ordering clinicians. In the end, quality is the leading market differentiator. Moreover, quality is also influencing physician performance as the radiology industry moves toward pay for performance (P4P) and physician quality reporting initiative (PQRI). For physicians to achieve and maintain a competitive market position, they must have the ability to continuously monitor, measure and improve the quality and processes of imaging services. Standardizing quality assurance (QA) principals more quickly enables the identification of problems and trends, so changes can be replicated or avoided, which can improve both current and future radiological outcomes. At Baylor University Medical Center at Dallas (Baylor Dallas), the primary QA goal was to implement an automated method of capturing quality data. The data would support quality improvement initiatives, especially those specific to image capture and report production. To achieve this goal, we had to develop a system that encouraged radiologists on the medical staff to use a speech recognition system to capture statistics related to image quality, image issues and report findings. The new system had to provide data for two specific areas in which to measure and improve: image quality and peer-reviewed report findings. Our mission was to create and enact a mandate to automate the capture of statistics that support QA initiatives. The groundwork for this QA project demanded that we clearly identify our requirements for such a system. The existing process was completely paper-based, which was laborious for the radiologist and not widely adopted. After forms were manually completed, responses were transcribed into a spreadsheet by clerical staff and the results tabulated. The nature of this process made it difficult to get staff to participate and meet both our internal and accreditation commitments. We soon realized that a balance needed to take place between the task of capturing the data and the ongoing productivity and efficiency of the radiologist. As we sought to clarify what worked with the old process and what did not, we continued to use the manual forms to identify the information that still needed to be gathered. Next, we came up with a process that would easily allow the radiologists to perform multiple types of QA activities as a natural outcome of their interpretation and dictation process. After the primary elements were identified, they were tested in a pilot program on a voice recognition system available in the marketplace. Once the proof of concept was complete, and the initial results were shared with the professional standards committee at Baylor Dallas, the new “coding process” was rolled out to other radiologists and started further deployment of the system. During the interpretation and dictation process, the radiologist would capture several pieces of data relating to a prior report, as well as the current case being reviewed. Prior report: concur, minor disagree, disagree Image quality: excellent, acceptable, unacceptable Image issues: positioning, motion, artifacts, no marker, etc. For ease of use, we provided all radiologists on the medical staff with a quick reference card to guide them through the process of capturing the data. One of the keys to success was designing a process to work within the radiologists’ existing interpretation and dictation process, so as not to disrupt workflow. The data is periodically extracted and imported into a spreadsheet for reporting and analysis. Statistics and trends are regularly analyzed, so that quality improvements can be recommended to the professional standards committee. Modifying the “coding process” for the radiologists has significantly increased both the number of radiologists who are willing to assist with these QA activities and the number of cases that are being reviewed monthly for QA. The charts illustrate that we have increased the number of participating radiologists by 1,500 percent, and the number of cases reviewed by 500 percent.  We expect to see those numbers continue to increase over time.   — Jean Plummer is the radiology informatics manager at Baylor University Medical Center at Dallas.