- AI
- Molecular Imaging
- CT
- X-Ray
- Ultrasound
- MRI
- Facility Management
- Mammography
The Maddening Avoidability of Doubled Work on Radiology Reports
Autosave functionality for dictation software and availability of prior imaging are elusive yet very practical essentials that could eliminate a fair amount of redundancy with radiology reports.
I had this week’s blog pretty much done. My usual routine is to get the thing written and then review it a few times, tidying with each pass until things seem ready to go. I was leaving the final run-through for the next day.
For whatever reason, my computer decided to shut down the word processor overnight. Although it helpfully offered to “recover” my file from yesterday, when I agreed to that option, I was rewarded with a blank screen. 1.5 pages of text had vanished into the ether.
I have been here before albeit not with this word processor. When I got this PC, I decided I had enough of Microsoft Office. Once upon a time, MS Office was straightforward and worth my while. Want a word processor? Price X buys you a CD you can install on your machine and use for however long you want. Care for a spreadsheet program? We have got those too. At some point, however, they became a confusing mess of subscription services, having to create a Microsoft account, etc. Plus, of course, their prices just kept on climbing.
I switched to a well-regarded “open source” suite of programs that gave me all the functionality without any of the hassles. Similar to competitors, it has an “autosave” function that periodically makes files of whatever you are working on. If your machine shuts down unexpectedly, you can just load the most recent saved version and lose little to none of your work. Unfortunately, I didn’t know you had to specifically turn on this feature, hence my situation today.
Like I said, I have been here before. The autosave with MS Office didn’t work 100 percent of the time. Once in a while, I would go to work on something from which I had stepped away a few hours or a day earlier, and the expensive software suite had nothing for me. Much like today, I would then go searching online for ways to manually recover lost files — including digging in the word processor’s menus for the “path” to see what folders might hold my lost work — and that worked 0.0000 percent of the time.
Word processors are not alone in this deficiency. Dictation software, including that which we use for making radiology reports, sometimes makes us redo everything we have done. It seems unfair to point out the specific product that has most recently vexed me since others have in the past. I don’t remember their details but none of them stand out as particularly bad (or good) actors.
The most recent entity had a software upgrade which was followed by periodic hiccups. The dictation window would vanish and, shortly afterward, the login screen for the software would appear. I would log back in, and of course my entire dictation would be gone without any options for recovery. It could just be my bad luck, but three out of four times, I lost entire CT reports (each a page plus of verbiage, including measurements).
In my ensuing wrangling with support people, some of whom more interested/helpful than others, I kept coming back to the same point: Yes, whatever is causing the software to crash probably came from your recent upgrade, but aside from that, why don’t you guys have some sort of autosave-function in your apparatus? There are plenty of circumstances that can cause dictations to get lost. Swatting each of those scenarios individually would be nice, but how about a global safety net that ensures rads don’t wind up having to redo entire cases?
I added that it wouldn’t even be a hefty task to add an autosave feature since a bunch of other software has the function. You don’t even need to program the subroutine from scratch, It is easily lifted from elsewhere, such as the open-source word processor I use.
When I could actually find support people engaged enough to hear my suggestion, it baffled me to hear answers like “Yeah, that would be a useful function. Maybe a future version of the product will have that.”
In my mind, this was such a low-hanging fruit that it would be a no-brainer to prioritize for implementation. “Hey, we have got something we could easily fix that would prevent legions of rads from possibly infinite future episodes of losing their work and having to do it all over again, cursing our names in the process. Let’s make this happen ASAP!”
The lackadaisical sort of response that I got, instead, is a prime example of the sort of thing that drives me (and I suspect most other rads) crazy. We see something that needlessly multiplies the time and/or effort we spend on our job, we point out easy fixes that would require little/no resources beyond compliance from others in the radiological system ... and somehow it motivates nobody to take action.
I have written in this blog about other needless work duplication scenarios that bedevil us, for instance the unavailable but actually available prior studies routine. I will get a full-body scan, for instance, with no priors on the system. I read it out. Maybe I find no significant pathology, or I see some stuff that would ideally be compared against previous imaging if any existed.
Presto! After I have done my work, I am notified that priors have been uploaded. Sometimes, this happens later the same day and now have to make a comparison addendum. This at least doubles my work. In most practices, there is no mechanism to credit me for the added time I spend on it.
Suppose when I receive the same scan, I figure out that the patient probably had prior imaging. Perhaps there has been surgery and maybe some radiation therapy. I make a point of asking whether there is previous imaging that might be retrieved before I read the thing. In my experience, there is at least a 50 percent chance that, when I am told “no” and proceed to read the thing as is, priors magically turn up (again, sometimes later the same day), and I get my addendum request.
Of course, it is extra frustrating to those of us in high-pressure roles, for instance, with high expectations of RVU productivity. If you are going to expect us to turn out more work than, say, 75 percent of other rads out there, and not sacrifice quality, you should pretty much do anything we think will help us out that sounds remotely reasonable.
The Reading Room: Artificial Intelligence: What RSNA 2020 Offered, and What 2021 Could Bring
December 5th 2020Nina Kottler, M.D., chief medical officer of AI at Radiology Partners, discusses, during RSNA 2020, what new developments the annual meeting provided about these technologies, sessions to access, and what to expect in the coming year.
