Guest Editorial

Dr. Banerian is an independent Radiologist/Neuroradiologist providing diagnostic services to hospitals and outpatient clinics in Michigan, Florida, and California.

Not so many years ago, you could diligently review a contrast-enhanced brain MRI, including all standard sequences, by perusing roughly 150 images. Typically, you had the referring physician's request with pertinent history/information, as well as the patient's prior films and reports. Today, because of higher image resolution and additional scanning sequences added in recent years, you may need to review more than 350 images when interpreting a contrast-enhanced brain MRI. Similar technologic advances mean hundreds more images to review for a combined chest, abdomen, and pelvis CT study. In many institutions, there is still incomplete integration of the radiology information system (RIS), the hospital information system (HIS), and the picture archiving and communication system (PACS), which adds to the time and complexity of ascertaining patient history and retrieving prior relevant examinations and associated reports.

While a standard "normal" report can still be quickly dispatched, the greater anatomic coverage and detail now available means that more abnormalities are discovered with a corresponding rare opportunity to use the quick "normal" report. I can not only see that 7-mm adrenal nodule on CT, but, in most instances, I can also characterize it. When scans were collimated at 8 to 10 mm, that 7-mm adrenal nodule usually wasn't discernable: Normal report, on to the next case.

Today it requires more work to interpret each case. Given the large annual national increases in imaging utilization along with the diminishing reimbursement for each case, our case load keeps expanding. There are more studies to interpret, more images per case, more findings to report, and more need to make comparisons to prior examinations. More and longer transcribed reports require greater time for proofreading and correction.

Where can we find some relief? Shortcuts. There are a number of acceptable and very helpful shortcuts for handling the massive amounts of information we process. Prepared standard normal reports and computer-aided diagnosis (CAD) can help. It is very helpful to have a properly configured PACS that is fully integrated with the RIS and the HIS, so patient history and all prior examinations and reports are readily available. The capacity of PACS to link several scan sequences from different prior studies or from different imaging planes or sequences in the same study saves time over trying to do the same with film. Having well-trained ancillary personnel do some of the leg work of contacting physicians or trouble-shooting computer glitches is helpful. These are some of the appropriate shortcuts.

We can, perhaps, rationalize a number of less desirable, possibly perilous, but nonetheless tempting work shortcuts. It is possible to find a reason not to review every image in a study. Perhaps, we may rationalize that there are simply too many images in a chest CT for "possible pulmonary embolism" to take the time to also carefully assess the included lung parenchyma for a pulmonary nodule or for interstitial lung disease. After all, if the study was requested to assess for pulmonary embolism, shouldn't that be where we put all our effort? Can't we just skip or skim through the lung windows? A CT angiographic study of the carotid arteries usually requires 500 images and covers essentially all of the soft tissues of the neck, the cervical spine, and portions of the chest and head. Should we spend time worrying about all this other stuff? If the referring physicians need a study of the chest, neck soft tissues, cervical spine, or head, they can order a CT of the chest, neck soft tissues, cervical spine, or head, right?

Rationalizations like these are most often clustered near the end of the day. Hey, it's 4:30 and we've got to "clean up" these cases before we leave. Memo from the top: "You must sign all of your dictated reports before leaving work each day." Perhaps, I can just proofread the report's impression for accuracy before signing. That's good enough, isn't it? That will save lots of time. An error in the history, comparison, procedure, or findings portions of the report isn't really that important, is it? They probably won't even look at that part. It's been a long day, and it's now past 6:00; I'll just batch sign/approve these last 37 reports. I'm too burned out to proofread them now, and our transcriptionists don't make that many mistakes anyway.

Another dangerous shortcut is taking insufficient time to correlate current findings with prior relevant studies. In the long run, this does not save time and can cause harm to the patient. For example, consider the chest radiographic interpretation of a 1-cm pulmonary nodule within the right lower lobe with a recommendation for a follow-up chest CT. As it turns out, the patient had an abdomen CT 4 years earlier that documented that the pulmonary nodule was a benign granuloma. The radiologist saved time-that day-by not looking for and correlating with prior relevant examinations. Now a problem is created that doesn't actually exist. The patient receives the bad news that there is something wrong on the chest X-ray that "could be cancer." An unnecessary chest CT is scheduled a week later, and the anxious patient must take the afternoon off from work. The test requires an IV, and the contrast injection offers the potential for the rare extravasation, an allergic reaction, an idiosyncratic reaction, or acute renal failure. Also, don't forget to add exposure to a sizable dose of ionizing radiation. So the radiologist who may have saved some time interpreting the original chest film has now generated an unnecessary chest CT that requires time to interpret and unnecessarily stresses a patient.

Lastly, some radiologists have decided to change the well-accepted American College of Radiology recommended format for radiology reports and forgo the conclusion/impression portion. Findings are dictated, sometimes mixed with differential diagnosis and/or follow-up recommendations. While forgoing the formulation, dictation, and review of a final impression will certainly save some time for the radiologist, it will waste the referring physicians' time as they try to figure out which findings are important. Also, any radiologist who reads this report to correlate the findings with later studies will need to hunt through the entire report to determine the relevant information.

Some shortcuts used in doing our work can lead us down the wrong path. It behooves us to strive for the highest quality to remain the preferred specialty for medical imaging interpretation. Radiologists must maximize their use of legitimate shortcuts:

• Have all available history in a single reference location on PACS.
• Engage PACS information technology personnel to help tailor hot keys and hanging protocols.
• Confer with your radiologist colleagues who may have developed efficient hanging protocols and hot keys for PACS.
• Have a radiologist assistant readily available to phone reports or to track down a technologist or other colleague to review a case.
• Review reports carefully for correct content and spelling. Consider that an incorrect or missing word in any portion of your report could, at a minimum, reflect poorly on the quality of your service and could result in harm to the patient.
• Refrain from rushing through any case for any reason.