Why radiology needs defined image storage guidelines

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon
 - images
Image credit: lifeImage

Cloud data storage is getting really cheap—like storing-a-500-image-CT-study-for-50-cents cheap. The sheer economy of storing images online should make it standard, but a maze of regulation and expensive penalties make it difficult for imaging providers to navigate the issue, according to an article published in the Journal of the American College of Radiology.

Providers might not have to worry about light or moisture degrading stored film, but plenty of questions remain. Chiefly, how long do they have to store image?

Medicare and most states require imaging must be retained for five years, but from there it’s impossible to generalize. Mammography sometimes has longer requirements, toxic exposures can trigger longer storage and the American College of Radiology (ACR) recommends holding images through the statute of limitations for a potential medical malpractice case—two or three years after a malpractice claim would be discovered.

Failure to maintain imaging up to state and federal standards can result in penalties up to $10,000, whether it’s a first offense or a repeat violation. It can also place radiologists at risk of malpractice suits: if a lost or misplaced image results in patient injury, the radiologist personally bears responsibility.

Lead author Jonathan Mezrich, MD, JD, assistant professor of radiology and biomedical imaging at Yale School of Medicine, believes this leads to unnecessary storage.

“It is likely that most facilities do not have a clear sense of how long they need to retain each particular image,” wrote Mezrich et al. “PACS databases do not typically create an image retention flag/date. The typical practice at present is thus likely to err on the side of maintaining everything, forever; thus arguably, a practice that endeavors to delete files at the statutory minimum retention date actually runs the risk of running afoul of the 'standard of care.'”

A single universal storage period would eliminate a myriad of regulations, although that comes with its own problems, according to Mezrich et al.

“As things are today, it would have to comport with the lengthiest possible statutory duty—the life of the patient plus any potential statute of limitations,” they wrote. “That might create logistical issues, as facilities would need to ascertain dates of death. So a focus on strict retention periods is perhaps misplaced.”

Instead, reducing the size of images may be the easiest solution—cloud storage is very inexpensive and will only become cheaper. The debate here is on lossy versus lossless compression: lossless compression is reversible, while lossy is irreversible but much more flexible in application and storage.

The ACR merely requires the amount of compression be “adequately diagnostic,” but the Food and Drug Administration (FDA) does not allow lossy compression for transmission or interpretation of mammograms. They do allow it for prior studies, opening up possibilities, according to Mezrich et al.

“Although lossless may be selected for the first five years after acquisition of an image, when most statutory requirements are in effect, for images older than this initial term, perhaps it would be reasonable for a facility to archive images using lossy compression,” the authors wrote. “Our Canadian counterparts have allowed compression to as much as a 30:1 ratio for certain types of images. Compression at these ratios generally maintains enough image integrity to retain even the least conspicuous findings.”

There is a clear need for a comprehensive image retention guideline, according to the authors, and the ACR is probably the organization to do it.

“We would strongly encourage the development of an ACR image retention guideline with input from expert opinion from the informatics community; we believe this will be useful clinically and from a medicolegal perspective,” wrote Mezrich  et al.