Last Friday I posted an xray that showed a long, thin metallic foreign body projecting from the right chest area toward the right side of the heart. What is it?
The patient declined to tell us what he had been doing. And looking at the xray, which is a 2D representation of a 3D object, it’s impossible to tell where the object is in the z axis. It could be in his back, in his clothes, or in his heart.
A complete physical exam showed a pinhole just to the right of the xiphoid process. By gently running a (gloved) finger across the hole, something thin and sharp was palpable.
What should we do next? More images and the final answer tomorrow!
In my previous post (click here to view) I discussed an imaging protocol that we developed and implemented last year. Today, I’ll detail what it has accomplished in our patients.
We looked at 229 patients who had their imaging performed according to the new protocol during a 3 month period and compared them to 215 patients who were imaged the previous year. Each scan administered to each body area (head, chest, abdomen/pelvis, c-spine, t-spine, l-spine, face, neck angio) were tabulated separately.
We found that the overall number of scans performed decreased significantly. We looked at our data and generated numbers per 100 patients. During the control period, we did 298 CT scans per 100 patients. This decreased to 271 during the study period. The number of head scans remained the same (82 per 100 patients during control, 85 per 100 during the study), as did the cervical spine scans (84 vs 86).
The biggest declines were seen in chest CT (53 per 100 control vs 33 per 100 study) and abdominal CT (57 vs 43).
We did see an increase in conventional xrays of the thoracic and lumbar spines to offset the absence of reformatted spine images that would have been generated from the chest and abdominal CT scans. We also noted small increases in CT of the head, cervical spine, and neck angio. This was likely due to better adherence to specific guidelines.
Bottom line: we believe that our work shows that careful adoption of well thought out guidelines can make a difference in practice and significantly decreases radiation exposure in our blunt trauma patients.
Last year, we developed an evidence-based protocol for deciding what radiographic images to order in our blunt trauma patients. For some body regions, there is fairly good literature available for guidance (i.e. Canadian head and cervical spine rules). For other areas, there is not nearly as much.
We convened a small group of people, including trauma surgeons, emergency physicians, radiologists and a radiation physicist to combine the information into a practical tool.
You can view or download the worksheet we use by clicking the link at the bottom of this post. The protocol has been in use for about 9 months, and has significantly decreased the use of higher radiation dose imaging (CT). As a result, there has been a small increase in the use of lower dose conventional imaging (plain spine studies), but no missed injuries.
Tomorrow, I’ll write about the specifics of how this protocol has changed our ordering habits. Click here to view it.
Typical order: “chest CT with and without contrast”
A review of Medicare claims from 2008 showed that 5.4% of patients received double CT scans of the chest. Although the median was about 2% across 3,094 hospitals, 618 hospitals performed double scans on more than 10% of their patients. And 94 did it on more that half! One of the outliers was a small hospital in Michigan that double scanned 89% of Medicare patients! As expected, there was wide variation from hospital to hospital, and from region to region around the US.
Time for some editorial comment.
This practice is very outdated and shows a lack of understanding of the information provided by CT. Furthermore, it demonstrates a lack of concern for radiation exposure by both the ordering physician and the radiologist, who should know better.
Some officials at hospitals that had high scan rates related that radiologists ordered or okayed the extra scan because they believed that “more information was better.” There are two problems with this thinking.
Information for information’s sake is worthless. It is only important if it changes decision making and ultimately makes a difference in outcome.
As with every test we do, there may be false positives. But we don’t know they are false, so we investigate with other tests, most of which have known complications.
The solution is to do only what is clinically necessary and safe. The tests ordered should be based on the best evidence available, which demands familiarity with current literature.
In trauma, there are a few instances where repeat scanning of an area is required. Examples include solid organ lesions which may represent an injury or a hemangioma, and CT cystogram to exclude bladder trauma. In both cases, only a selected area needs to be re-scanned, not the entire torso.
Bottom line: Physicians and hospitals need to take the lead and rapidly adopt or develop guidelines which are literature-based. State or national benchmarking is essential so that we do not continue to jeopardize our patient’s safety and drive up health care costs.
Tomorrow I’ll share the blunt trauma imaging protocol we use which has decreased trauma CT use significantly at Regions Hospital.