User Story: Radiation Oncology
Ordering Provider
Erqi Pollom, MD, Assistant Professor of Radiation Oncology, and by courtesy, Neurosurgery at Stanford University
How did you hear about the informatics consult program?
I saw the announcement in the “What’s new in SHC Epic?” email and then ordered my first consult after the Atropos presentation to the radiation oncology team.
What is your primary area of interest?
I primarily treat central nervous system and GI tumours, plus I have a background in epidemiology so I have worked with claims databases and cancer registries in the past doing outcomes research on patterns of care and cost-effectiveness analysis.
For example, using SEER-Medicare, we investigated the differences in emergency department utilization and cancer care costs for patients with mental health diagnoses vs. those without [Atropos note: publications here and here]. We also studied the impact of health parity laws on the out-of-pocket costs of oral anti-cancer medications.
What question did you ask the Atropos informatics consult service?
Is the use of immunosuppressive/immunomodulatory agents associated with the subsequent development of glioma?
What would you have used to answer this type of question in the past (before the informatics consult service was available)?
Many of our studies require manual data collection via charts or accessing cancer registry data. If we need a large claims data set we can go to the population health sciences team to get access to Optum/MarketScan data.
But generally, access to data is a bottleneck that can take a long time, and the even bigger issue is often finding people with the skill set to clean up the data to get it into the format we need to analyze it. Running the statistics is actually pretty straightforward, but getting the data ready is hard.
In your own words, please describe the Atropos informatics consult service.
It’s a service that leverages data in the EMR to answer clinical or research questions. It’s especially useful when brainstorming projects among research teams.
If it’s a question that can be answered using Stanford’s EMR or your national data set, you can get answers very quickly.
What was your overall impression of the service?
My first impression was that I was very impressed with the speed. It was also great to have some back and forth with an informed clinical analyst to further refine the study.
The report is actually very readable. I know that because one of the folks on my team is a pre-med student and they had a pretty good grasp of the material right away.
I’m still exploring what’s possible with the service and it would be nice to see a roadmap that lays out in detail what data elements you have access to today vs. the future. Also, I understand there are also additional efforts underway to continue to improve the transparency of methods and minimize bias which will be immensely helpful.
Any concluding thoughts?
When I forwarded my first Prognostogram to some of my colleagues, I was like, “Look how quick this was! This is awesome!”.
Sometimes when we treat patients with rare tumors, we reference old notes or spreadsheets that many of us keep to look up how we treated similar patients in the past. That’s obviously not an ideal way to recall medical knowledge.
When you want to know things like is there increased toxicity when some of these newer drugs are paired with radiation, your service can definitely help out.
Erqi Pollom, MD, Assistant Professor of Radiation Oncology, and by courtesy, Neurosurgery at Stanford University
