Amid concern from U.S. health organizations over the last few years, the Connecticut Hospital Association is taking steps to implement a statewide data repository of how much radiation a patient is being exposed to during such radiological procedures as CT scans. The repository could lead toward benchmarking appropriate levels of radiation dosage nationwide.

The CHA claims its dose management repository is the first of its kind in the United States. Regions of Canada and Europe, where the government controls health care, have systems in place, according to Mary Reich Cooper, M.D., chief quality officer of the CHA and senior vice president of clinical services.

She says awareness of radiation dosage became prominent about four years ago, when organizations such as the American College of Radiology and the American Board of Internal Medicine and its Choosing Wisely initiative were presenting radiation dose guidelines to help patients navigate potential overuses of care. Risks of overexposure include a potential of cancer as well as injury to the epidermis, such as skin erythema. The ECRI Institute, which guides its hospital members on how to use medical devices safely, labeled radiation dosage a “Top 10 Health Technology Hazard” in 2013 and 2014.

The CHA partnered with Bayer and its Radimetrics software to collect radiation dosage from machines at the 27 hospitals in the CHA, pumping the data into a repository that can be further analyzed. Bayer, which was picked during a one-year vendor selection process, installed the repository for free. 

Cooper says that the CHA will begin looking at the aggregated data this summer and report to the hospitals “a range of distribution of performance.” About half of the hospitals are reporting data, and the CHA currently has about 150,000 images.

Jason Launders, director of operations in the Health Devices Group at the ECRI Institute, says tracking software is a step in the right direction, but there are issues to address. For one, there aren’t standard names for the protocols. If one hospital reports dosage as a  “head scan” and the other a “brain scan,” it reads differently in the databases even though it’s the same thing.

Also, it is hard to measure what are recommended levels of appropriate radiation dosage. A heavier patient, for example, may require more radiation to penetrate their skeletal system or any fat surrounding their organs. “If I give less radiation and am unable to visualize the tumor, that’s much more of a consequence than my giving more radiation and, over time, increasing the risk of radiation-related cancer,” Cooper says. 

Launders calls the reporting “a data-mining research project of gigantic proportions,” so there’s an issue of manpower. And lastly, there’s variability between new and old radiology machines.

Mani Adib, senior project officer in the Health Devices Group at ECRI Institute, says newer machines produce images with lower doses of radiation but doctors have become accustomed to the high-quality images on older machines, asking to increase dosage.

Overall, Cooper is excited about the potential of the data and the ability to give patients more knowledge. “They go into an emergency room and take into account the tests they’ve had before and have awareness of cumulative doses of radiation,” she says, adding, “It’s like sunshine. Lots and lots of sunshine isn’t great for you, but sunshine is great for you.”