Q&A: A perspective on digital radiation oncology transformation
Radiation oncology department workflow is extremely complex due to the dynamic, interactive peer-review based nature of the workflow involved in imaging patients and treatment planning. Health Imaging & IT spoke with the head of one department that has made the shift to all-digital, John E. Bayouth, Ph.D., director, Radiation Oncology Medical Physics Division, Department of Radiation Oncology, University of Iowa Hospitals and Clinics, to get his view on the challenges and benefits of going filmless and paperless in such an environment.What systems did you adopt to make the shift to a paperless/filmless environment?
Two years ago we began the process of preparing to receive a new set of treatment machines, a new set of imaging machines, and a new electronic patient treatment record. In receiving all of that new information, we put a lot of thought into how we were going to make the transition from a paper chart and physical copy films to a purely digital environment.
What do you think is unique about this change for a radiation oncology department as opposed to, for instance, a general radiology department?
Fundamentally this process is different in radiation oncology compared to radiology. In radiation oncology we take a set of information that is viewed in parallel by multiple people, then one individual will review the resulting information and use it to formulate the treatment plan for that patient. That information is passed from one person to the next each adding and modifying as it works its way through. So the same DICOM dataset is reviewed and information is built onto that dataset from one person to the next, maybe a half dozen times before we are finally finished with it.
It is important for us to account for the progression of the information within the process. We can’t simply have six different copies of that information living out amongst these six people, because if the sixth person in the process were to try to make modifications on the dataset that came from the first person and not from the fifth person, errors would occur. So workflow in radiation oncology is dramatically different than the workflow in radiology.
We also have a lot of DICOM objects that are unique to radiation oncology – for example, the radiation dose distribution. The distribution of dose has to be computed and registered to the DICOM imaging datasets, and from that information we generate a DICOM object called a DICOM RT plan which has in it all of the parameters that are necessary at the treatment machine in order to generate the patient’s treatment. That information is transferred to the treatment machine, followed by multiple checks that are in place to validate accuracy.
It’s also true that we use that same information 30 or 40 times for the same patient on a daily basis. Each time we deliver a treatment the information is recorded and built onto itself so that we are able to verify exactly what we’ve done for that patient as they’ve progressed through their treatment.
Are the adjustments to the patient information all logged in a timetable?
There are parts of it that are inherent in the software as you have described. In addition there have to be administrative controls that we place on how the information is used to impede the possibility of multiple people working on the same piece of information at the same time, or the likelihood of using the wrong version of information.
The way we do business has changed dramatically. We spent a year defining what our new workflow would be in our new department. For example, in the old department when you needed a physician to review something in a patient’s chart you could put a sticky note on a patient’s chart and place it in their chair. Odds are when they sat down they’d realize that something needed to be done. But when there is no physical object you have to have electronic ways to indicate what the workflow needs to be. So, the way in which patients move throughout the clinical process has dramatically changed.
What do you think the core benefits have been to your department?
First and foremost clinically, it takes away the time that we used to spend searching for patient charts. We would spend hours every day with people running around looking for charts because the therapists needed to have them laid out in front of them before they treated a patient. You never knew where the chart was located. Was it with the attending? Was it with nursing? Was it back in physics? It could have been anywhere. There was really no way to track the location of the chart. So, a lot of time was spent doing that and it drove everyone crazy.
More fundamentally, I can tell you that the amount of information we have in the patient’s charts is far more detailed than what we had in the paper chart before. And I believe that’s because it’s easier and more efficient to document the information electronically than it used to be with a paper chart.
How do you think patient care has improved?
On a weekly basis we review patients’ treatment plans and how patients are progressing through their treatments. It’s called ‘chart rounds’ and the process contains significant peer review. This peer review now relies upon a greater breadth and depth of information, an order of magnitude more information, which greatly enhances the peer review.
That includes information on the imaging side used in diagnosis, generating the treatment plan, and how the patient is doing during the course of treatment. We not only have digital planar images that show us how the patient is being aligned, today we also have the capability of acquiring cone beam CT of the patient in treatment position while they are laying underneath the treatment machine. So we can see on a daily basis how the patient is setting up, and how their anatomy may be changing throughout the course of treatment. We have a level of peer review through the entire process that is fundamentally better.
What have been some of the other core benefits to your department?
Having the patient record and all associated information stored electronically in a single database also enables us to audit so many things about what we are doing. It’s easy for us to review patient treatment records and look for holes in information using internally-developed computer programs. We can write code that helps us mine statistics on patient care including clinical response to therapy. We are also able to audit what is being billed and reconcile this information with what is well documented in the patient charts.