Structured Reporting? Why not Interactive Reporting?

June 3, 2011

My vision of idealized reporting - just like Google maps

This past weekend, during some superb weather, I unexpectedly found myself with a few hours of free time, and decided to go for a hike. I wanted to get out the door and on the move ASAP. The less time spent at home planning my route, the better.

My favorite tool for situations like this is “Google Earth.” In case you haven’t heard of it, this marvelous software produces a virtual globe you can spin, magnify, tilt, and otherwise manipulate to get a local view of almost anywhere you’re likely to go (and a lot of places you’d otherwise never see at all). Satellite-acquired imagery shows a bird’s-eye view of the area of interest. Not only can you find your house, you could well see your car parked in its driveway.

One of the features I find most useful is the presence of points of interest on the interactive map. These range from Wikipedia-links for municipalities, to Google-search results for businesses, parks, and monuments, to photos posted by other users. For instance, someone taking a snapshot of a beautiful sunrise at the beach can post the image as an icon on that very beach…and now other users viewing that area can select that icon to see the photographer’s shot.

Looking at such on-the-spot pictures and their descriptions helps me plan out my hiking-route, since I can know in advance the sorts of sights I’ll be seeing.

It occurred to me that such an interface could be of interest in reporting imaging-studies. Perhaps Google could develop it, since they did so well with Earth.

You’d fire up a patient’s CT, MR, whatever, and you’d get a rotating MIP of the study (or the sagittal and coronal scouts, if you prefer). Each area of abnormality could be represented on the anatomic structure by an icon, placed by the interpreting radiologist; move the mouse over the icon, and you’d get a small window of text containing the report’s statements pertaining to that organ. Click on it, and you’d get the relevant image-slices showing the actual lesion. Meanwhile, on the linked text-report for the study, all abnormalities would be clickable links, such that selecting the word “angiomyolipoma” would prompt the images to zoom in on the renal lesion.

There could even be links to CME articles, in case the user didn’t know what an AML was but would like to learn.

We’re not quite there yet, of course. For starters, radiology reports and patient demographics often reside in a software entity separate from the actual PACS, and you’re lucky if those two systems interact harmoniously. Some of the nicer systems out there do combine them, and it’s then possible to link key images to portions of the radiology report. Even if you have such a fancy setup, however, you’re probably aware of just how few of your referring clinicians share your bells and whistles…and thus cannot make use of them.

Needless to say, the increased cost of such cutting-edge tech (not to mention the extra time required of the radiologist who’s placing all of those helpful annotations) won’t result in any more reimbursement from insurers or the federal government; in their eyes, you’re not providing any more of a service than the guy who’s still annotating hard-copy films with a china-marker.

Maybe that should change.