DALLAS—Suboptimal workflows plague radiology, Luciano M. Prevedello, MD, MPH, from the department of radiology at Brigham & Women’s Hospital in Boston, said during the Society for Imaging Informatics in Medicine (SIIM) annual meeting. Various forms of decision support offer a way forward, he contended.
Before decision support
Prevedello offered an all-too-familiar scenario to illustrate the spectrum of suboptimal workflow. A 53-year-old woman presents to the emergency department with an acute headache and VI nerve palsy. The physician orders a head CT to rule out a mass.
The somewhat vague nature of the order generates questions and a phone call from the radiologist to the referring physician. After a discussion of the patient’s symptoms, the radiologist recommends MR rather than CT to answer the clinical question, which prompts the ordering physician to change the requisition.
At the next stage of the process, image acquisition, staff realizes the patient has renal function issues, forcing the radiologist and technologist to determine if the patient can tolerate the exam.
Finally, when the study is completed the radiologist needs to know how to best communicate the findings to the referring physician. The process is compromised because by the time the results are in hand, the patient has been passed to another physician.
Finally, at the end of the suboptimal workflow cycle, the billing department connects with radiology to remind the department that the indication was not billable. Thus, the study represents lost income. In addition, it also lacked an attestation statement as required by the Centers for Medicare & Medicaid Services.
The decision support difference
The way to improve what is broken and realize improved safety, quality and efficiency is through informatics, said Prevedello. Clinical decision support provides evidence, knowledge and patient-specific information intelligently filtered or presented at appropriate times to enhance healthcare, he continued.
He redrew the use case with decision support.
At the time of order, embedded decision support would have alerted the referring physician that a head CT required more information and ultimately that MR offered a better exam. Decision support also could allow the physician to revise the order with a click.
Meanwhile, with decision support the radiologist could more easily determine the appropriate exam, selecting the protocol from a dropdown menu. In addition, Brigham & Women’s Hospital has built a Google-like engine that includes relevant policies and protocols to assist residents and other radiologists who may not be as familiar with local protocols.
Similarly, service-oriented architecture also could enable the practice to determine that it has included all variables needed to attest for Meaningful Use and establish a framework for improved diagnostic imaging by including links to studies focused on the patient’s diagnosis.
At the reporting level, integrated decision support enables critical results communication stratified by urgency and acuity.
Finally, decision support can close the loop by ensuring that physicians adhere to follow-up recommendations by issuing alerts to communicate the time frame for studies such as follow-up chest CTs for a pulmonary nodule.
Prevedello concluded by emphasizing that IT offers a means to improve efficiency, quality and safety.