Defining and Managing Standardized Imaging Protocols with Appropriateness Criteria
 
Authors:
Janice C. Honeyman-Buck, PhD, FSIIM, Medical Imaging Consultants, Inc; Mary Shaw, MA; Meryll M. Frost
 
Background:
In recent years imaging technology has enjoyed rapid evolution, resulting in increasingly complex imaging potential, which should improve diagnostic accuracy for the patient and efficiency for imaging personnel. However, since most facilities are already operating at maximum imaging capacity, taking the time to optimize the use of the new technology is difficult, if not impossible. In addition, there is a national recognition that it is important to perform the appropriate imaging procedure for a patient. This requires that correct indications for imaging procedures are identified and that everyone in the healthcare system understands about them. Failure to perform the correct procedure can result in multiple procedures, delayed diagnosis, and inefficient operations or, in a more extreme case, medication errors and missed diagnosis. Indications for imaging procedures are part of a complete protocol.

Each imaging procedure requires a precise protocol. In some cases, an imaging procedure may have a protocol with optional supplemental protocols for special indications; however, these should all be well defined. Imaging procedures consist of one or more series or sequences of images. In the case of Computed Tomography (CT) or Magnetic Resonance (MRI), these are usually stacks of images. In the case of Ultrasound (US), these are sets of individual images, a “run” of dynamic images, Doppler imaging, and measurement information. In the case of Digital Radiography (DR), these may be individual images with one image per series or sequence. In all cases, the naming of the series or sequence must be consistent throughout an institution for all similar imaging modalities. Radiologists depend on those series or sequence names, also known as descriptions, to define how the imaging procedure is displayed for interpretation. Without consistent naming, the radiologist cannot define his or her display preferences or “hanging protocols.” This combination of appropriate indications, well defined protocols, and consistent naming of sequences for series for an imaging procedure are needed to guarantee the best possible diagnostic image set, the most efficient use of high-end imaging technology, the most consistent reading environment for the radiologist and, ultimately, the best patient care.

We were assigned the task of standardizing imaging protocols across a Veteran’s Integrated Service Network that consisted of ten VA Medical Centers, three of which had University affiliations. The group had enjoyed a rapid deployment of new acquisition modalities and, although the various radiologists and technologists communicated with each other, each institution used different acquisition protocols. Patients often transferred from one medical center to another and, at times, the new facility required different imaging protocols, which led to either duplicating studies or radiologists being forced to deal with images that could not fit into their hanging protocols and that may not have contained all the series they desired. In addition, the radiologists often over-read studies from the more remote facilities, leading to more frustration.

 
Evaluation:
After gathering information from all the facilities on their imaging protocols, we focused on each group, found commonalities and differences, and proposed sets of standardized protocols for the entire system. The proposed protocols were based on the ones being used, along with research into the best practices for imaging. We also proposed a set of appropriateness criteria for each protocol. The proposed protocols were then presented to the radiologists, technologists, and vendors and, through an iterative process, a final set of protocols was agreed upon.
 
Discussion:
Of course, not all the protocols were appropriate to each institution. For example, some of the smaller facilities did not do CT Cardiac work, and their acquisition equipment and post processing was not even in place to perform these studies. In addition, we developed some “alternative protocols” for sites where research was being performed with affiliate institutions.

In general, we ended up with protocols that would standardize 80% or more of the studies performed. There will always be outliers or add-on series or sequences and the radiologists are not always forced to follow the standard protocols. The radiologist always has the final say on a protocol, but standardized protocols are used for the most common procedures. The appropriateness criteria help the radiologist communicate with the ordering physicians on the best study to order for a patient’s condition.
 
Conclusion:
We believe the problem of non-standardized protocols is common in large institutions and multi-facility environments. Our experience shows that standardized protocols are needed when patients move among facilities, when radiologists are performing interpretation for a sister facility, when facilities implement personalized hanging protocols, when radiology departments want to assist ordering physicians in selecting the appropriate study for their patient, and when institutions are implementing high end image processing, CAD, and image guided procedures.