A Suggestion for Decreasing Unnecessary Radiology Studies

In a recent note, I raised the issue of how the rapidly increasing number of medical imaging studies was raising the cost of care. I must admit that I was very pessimistic about most efforts to "bend the curve downward" and listed a number of reasons for my position (see: CTs, MRIs, and PET Scans Drive Cost of Healthcare Upward). A reader, Jackie Walsh, responded with an interesting and practical comment:

Bruce, I agree that it will be hard to reverse this trend for the reasons you state. However, there is a category of imaging exams that can be reduced to offset the overall growing volume. That is the percentage of exams that are ordered simply because a similar or recent exam for the same patient isn't accessible (perhaps because it lives within the firewall of another hospital, or was burned to a CD and from some technical reason is inaccessible). In the US it has been estimated that over 10% of all diagnostic exams may be wasteful duplications because physicians do not have access to their patients’ previous records. This waste is costing billions of dollars annually and some patients are exposed to excessive radiation as a direct result of redundant exams. This is part of the overall cost of diagnostic imaging that we could really tackle.

Jackie's idea deserves serious discussion. Most hospital-based personal health information (PHI) is not particularly portable and cannot easily be reviewed if a patient is admitted to a different health system than the one in which the electronic records were initially generated and stored. She refers to such information as "living within the firewall of another hospital." I will include within this PHR category the radiology reports, residing on the first hospital radiology system (RIS), and the images themselves that reside in the first hospital PACS. When a patient is admitted to a new hospital, it is both possible and recommended that a CD be obtained from the previous radiology department and carried to the new one so that they can be reviewed. It may be possible to avoid redundant radiology studies if this latter step is taken. One would think that the burden of responsibility should fall on the original hospital to transfer a patient's records and images when notified of a status change but I fear that it's the patient who usually needs to perform this task.

In the future, a solution could be devised to correct this medical imaging access problem to avoid rework. If Jackie's assertion above that some 10% of all imaging exams are redundant, such a solution could result in substantial cost savings. Recall that I have posted a number of previous notes about cloud storage, which is particularly relevant for radiology images. A suitable solution for this problem would be a national central image repository for radiology reports and images using cloud storage. An accompanying metafile could contain patient demographic data plus the date and type of all radiology studies performed in the U.S. for, say, the last decade. When a patient is admitted to a U.S. hospital, a query could be automatically generated to this metafile to quickly determine whether there was a record of any recent imaging studies for that patient. If the records show the existence of such studies and with proper permission, the images themselves could be downloaded to the requesting facility. This approach would not solve the need for overreads, where the new hospital policy would require that their radiologists review the newly downloaded studies to determine whether there is agreement with the previous reports.

Comments

[Mark Cox]

In NZ this is a key focus and justification for a number of projects.

Have a look at http://www.testsafe.co.nz/ & http://www.testsafe.co.nz/images/testsafe_diagram_webready.png

This is a system set up that contains all pathology (hospital and community), and radiology for a region. In this case Auckland with a population of 1M. Soon to be added is community Pharmacy dispensing which will have all drugs ordered by a GP and dispensed at the local pharmacy.

One of the key justifications for this project was retesting when patients moved from one hospital to another or from secondary to primary care.

In suppise this is one of the big advantages of living in a socialised environment where the government can invest in regional clinical data respositories.

Mark Cox
cox.mark@sysmex.co.nz