Meeting the Challenges of Complexity
David B. Larson
As patients and information flow through the imaging process, value is
added step-by-step when information is acquired, interpreted, and communicated back to the
referring clinician. However, radiology information systems are often plagued with communication
errors and delays.
InBrief talked with David B. Larson, associate chair for performance improvement in the department of radiology at Stanford University, about communication models that help reduce errors as described in his article, "Communication in Diagnostic Radiology: Meeting the Challenges of Complexity," published in the November 2014 issue of AJR.
You note in your article that information is radiology’s stock-in-trade, yet radiology information systems are often plagued by communication errors and delays. What are some of the elements and challenges of complexity that contribute to these difficulties?
Diagnostic radiology is not an isolated operation—our patients are referred to us from other physicians. We acquire imaging information; incorporate any other relevant factors, such as patient history; interpret this information; and return a report to the referring clinician—and, often, to the patient. At each step, information is generated and moved through the system. Moving information is relatively straightforward when it moves linearly, but it becomes much more complex when many players are involved.
For example, large radiology departments consist of many modalities containing numerous scanners, often from several vendors. Tests are performed for many diseases via increasingly specific and sophisticated methods. Radiologists and technologists have varied backgrounds, with varied habits and communication styles. Reports must be provided to many different referring clinicians representing many different types and levels of specialization. A wide variety of machines and electronic systems must communicate with each other. The list goes on. With the explosive advancement in medical technology, knowledge, and computers’ processing power, the complexity problem is rapidly growing worse.
What factors specific to radiology contribute to the dysfunction?
As radiologists, we often don't realize that we are in the information business or appreciate the implications. Often, we view our primary role as providing an interpretation. While it is a critical part of our role, that is not its essence. Our primary role is to enable other medical professionals to make appropriate treatment decisions that are informed by imaging. A diagnostic radiologist’s job is not done until his or her interpretation has been incorporated into the diagnostic process in a meaningful way. As I tell my residents: Moving the football all the way to the 1-yard line is not worth 99% of 6 points—it's worth zero points.
Yet, a reliable communication system is not something that an individual radiologist can implement. It requires that all in the practice value this aspect of their work, make the appropriate investment in developing and implementing effective systems, and employ those systems once implemented.
You and your coauthors outline a regimen—a virtual roadmap—for continuously improving communications in radiology. What are some of the key elements of that strategy?
We need to understand that we are part of a large, complex system. In such systems, there is a natural tension between highly skilled individuals who understand their domains in great detail but cannot appreciate the overall system and the central authorities and system managers who better understand the larger system but cannot understand each individual process. Keeping this incongruence from negatively affecting the practice—and the patients—requires a good-faith relationship between the two groups. For the benefit of the patient, leaders must respect and support individuals’ ability to learn and adapt locally, and provide them with the tools needed to do so effectively; individuals must respect the need to optimize the system as a whole and be willing to compromise accordingly.