I Love Lucy
Dr. Mirvis is the Editor-in-Chief of this journal and a Professor of Radiology, Diagnostic Imaging Department, University of Maryland School of Medicine, Baltimore, MD.
Many of you are old enough to remember the I Love Lucy show or have seen it in reruns on TV Land. It’s amazing how many episodes of that show I watched as a child are etched into my cerebrum; certainly more so than the Krebs cycle. I think about one of the Lucy episodes more today than others and for a good reason. Whenever I feel overwhelmed at work, trying to keep up with reading an avalanche of studies, I recall the episode where Lucy is working at a conveyor belt trying to wrap chocolates as they go by her station. As she begins her task, the candies go in a slow, predictable way, allowing her to carefully put each wrapper neatly on. She is quite comfortable at this. Then the belt begins to speed up steadily and she has no way to slow it down. Before long, she is wrapping only some of the chocolates. Some are falling on the floor. She is eating some and putting them in her hat and dress. She finally becomes totally flustered, gives up, and a huge mess results. None of the candies get the attention they need. This is the kind of bit where Lucille Ball was at her comedic best.
Well, what’s going on here is that as imaging cases go by faster and faster, with no ability to change the speed, I wonder how many candies/cases are getting less than their proper share of attention. Just about every trend and change we have seen in imaging has been geared toward faster study acquisition, transfer, interpretation, and communication of results. A film alternator could hold only so many films at a time before a mandatory reload. CT scanners were also a rate-limiting step in a patient work-up. A total body scan, as commonly done in trauma today, took perhaps an hour in the single-slice computed tomography (CT) era before it was ready to be interpreted. In the hard copy era, lots of films were taken and maybe were not officially reviewed until days later, if they were ever returned. Everybody had time for lunch and conference. Half of the residents in my department went to lunch at the public market up the street for good, cheap food. From 11:30 AM to 12:30 PM, you could fire a rocket through the heart of the department and not hit anyone. Faculty usually disappeared into their offices to reappear sometime in the late afternoon. Perhaps things were too relaxed and we needed more work to keep us in the department working. Well, those times are certainly long gone.
Today, radiologists live in a different world and perhaps a much scarier one. CT scans of multiple body parts are done in minutes and ready to be seen about the time the anode is spinning down. Instead of 12 or 24 images of a body part, there are hundreds. Reformatted and even 3D displays are routine in many cases. Image sets are typically sent in several standard windows/levels. In the ED, oral contrast has been all but eliminated because it takes too long for the patient to be scanned if you use it. Most ED patients come to CT for some reason, related or not to their presentation. Moving patients rapidly through the ED is a priority. The CT work-up usually does supply a highly reliable positive or negative result allowing patient disposition to be made more rapidly. The speed at which cases reach the workstation constantly accelerates as the years move along. To the typical interpretation time, add the time to respond to frequent phone calls, technologist questions, tailoring of protocols, teaching (rare), reviewing comparison studies, entering outside studies to PACS and interpreting them, finding and calling other physicians about vital or unexpected findings, and engaging in inevitable debates with referring staff who are quite certain they know what they want, if not what they need, and how to get it.
The general response to this unregulated increase in workload is to work harder and faster rather than add radiologists or try to regulate the volume of studies performed. In an era where quality control is the focus of so much attention and documentation, there is a rather large gap when it comes to studying the effect that a constantly increasing workload has on physicians caring for patients, including their imaging studies. Imaging has evolved into a great funnel, through which so many patients flow because of all it offers. In the quality process currently utilized in my department, one of every 10 or 20 cases is reviewed by a radiologist blinded to the original reader for evidence omission of significant findings. For whatever reason, my sense is that this method is very insensitive and unreliable in truly measuring diagnostic accuracy. One often gets feedback on missed diagnoses from clinicians, other radiologists, your own later encounters with the same studies, and, unfortunately, lawyers in identifying our mistakes, but most of this feedback is off the record.
Some people would say that you are trained well enough and compensated well enough financially to cope no matter how great the workload. Well, of course, that is nonsense. Increasing the workload is a simple inverse relationship permitting less time available for each case review. A single complex case could easily require an hour to analyze, requiring a speed up in interpreting studies that follow to keep up with clinical time demands. There is less time for consultation with online references or other radiologists or to get more clinical information (almost always deficient initially). In this circumstance, short-cuts are inevitable and short-cuts force you to miss some of the scenery.
Just as in most assembly line work, making products of the highest quality requires adjustment of the speed of the line to ensure near “perfect” performance at each manufacturing step. Imagine cars racing along the line so fast that on average a third of bolts holding the engine blocks are too loose. Dropping an engine on the highway is inconvenient. This was actually the case with a Corvair Monza my parents owned a long time ago. Fortunately, the motor dropped out when the car was parked in front of the house.
How do we know when a radiologist is reading studies too quickly or an ER doc is seeing too many patients per hour, or a nurse covering the floor has too many patients to care for? Should we have parameters to follow such things? Should it require a significant medical error to raise the issue? At least doctors in training have some limitations on work hours, but what they safely can accomplish in that time is another matter. The potential for physical and psychological problems this stress can cause cannot be dismissed, either. I am confident many readers have been in this situation illustrated so well in that “I Love Lucy” episode. There is real risk here and medical workloads need to be measured, standardized, and controlled. There is busy and then there is overwhelmed. The difference needs to be recognized. Perhaps one day a plaintiff in a malpractice case can add the risk management team as a defendant for not monitoring physician workload and its impact on patient care.
While Dr. Leonard Berlin has expressed his opinion on radiologists’ overburdened workload as of July 2000, there was no study at that time that was able to, and would not in the future likely be able to, determine a set workload number or interpretation time as to constitute a standard of care since there are so many individual factors influencing what is “appropriate” in a given setting. In the same article, he went on to suggest that if radiologists could not give 100% of their knowledge and expertise in interpreting studies, they should delay the interpretation until adequately rested or ask a radiology colleague to interpret the study.1 It is sound advice, except it is a luxury many of us seldom have (like at 3 AM) or would feel comfortable actually using. This is an issue that needs a lot more consideration.
References
- Berlin L. Liabilities of interpreting too many radiographs. AJR Am J Roentgenol. 2000;175:17-22.