Expert: Bi-Enabled Technology in Radiation Oncology Is Shortening Time From Diagnosis to Treatment

Video

Through automated tools related to scheduling, prior authorization, and radiation planning tasks, there are efforts to shorten the time from diagnosis to treatment to improve patient outcomes.

Pharmacy Times® interviewed Matthew Manning, MD, FASTRO, a radiation oncologist at Cone Health Cancer Center, on the ACCC Annual Meeting & Cancer Center Business Summit facilitated workshop he will be participating in titled “Bi-Enabled Technology Solutions.” The workshop looks to address predictive modeling and data analytics that help reduce costs and improve revenue cycle management, technology platforms and AI-enabled algorithms that can support scheduling and resource utilization, and data collection and reporting that can help increase participation in alternative payment models, meet payer-mandated requirements, and improve issues pertaining to social determinants of health.

Pharmacy Times: What are some of the new opportunities made available by technology platforms and AI-enabled algorithms in oncology?

Matthew Manning, MD, FASTRO: There are a lot of opportunities in oncology to improve precision medicine, increase early detection, and improve the workflow of oncology care. And so business intelligence tools that are empowered by [AI] or even just simple automation algorithms can help to improve the quality of care for individual patients and improve the efficiency of clinical operations as patients go through their oncology journey.

Specifically, there are tools in radiation oncology, which is my field, that are beginning to shorten the time from the initial diagnosis of the cancer to the actual initiation of treatment. There's plenty of data out there that delays in treatment of cancer lead to worse cancer outcomes. And the time from planning radiation or being consulted for a treatment with radiation to the actual treatment sometimes lasts between 5 and 20 days. And through automated tools related to scheduling, prior authorization, radiation planning tasks, there are efforts to shorten that time to improve patient outcomes.

This session with [ACCC], a deep dive into some of these newer tools, allows us to share some of the existing state tools and moderate a session, but also interact with members of ACCC to see what members are actually using and doing. And, in the future, we hope that the content generated from this annual meeting can be disseminated to other members of ACCC to raise the bar and raise awareness to move AI forward in oncology.

Pharmacy Times: How can technology focused on data collection and reporting support participation in alternative payment models and meet payer-mandated requirements?

Manning: AI tools and Bi-tools can help health systems participate in accountable care organizations through things like automated data collection, finding multiple sources of varied information and data, and pulling records in electronic medical records from hospitals and clinics and doctor offices into data lakes that can then be shared to provide point of care tools for providers. It also—the intelligence here can be used to risk-adjust patients, and better pay for ongoing services.

The analytics tools and reporting also provide you with the ability to provide quality metrics that are not only lagging at the end of the month but can be real time dashboards related to patient care. And they can drive providers at the elbow with clinical decision support tools. So, if you've got patients who have missed their screening mammogram or they're eligible for a screening, lung cancer screening CT, the electronic medical record may prompt providers at the point of care to provide those services that increase the quality provided by the ACO and the health system.

Pharmacy Times: How about the potential application of AI-enabled algorithms for patient data analysis to address social determinants of health (SDOH)?

Manning: So [AI], the power of it is the potential to aggregate an enormous amount of data from a variety of sources, both demographic data, clinical data, patient characteristics, zip codes, and aggregate that data in such a way that patterns can emerge and machines can learn patterns that can help categorize personalized care for patients and allocate resource and stratify patients by their risk and if the attention is on Social Determinants of Health, then the AI can assist providers and health systems to allocate the resources properly. The downside of AI with social determinants of health and racial inequities in health care, though, has been established that if the data that's put into the model for learning, if it contains racial disparities or health inequities in the existing data, sometimes that's what's produced by the AI algorithm. And furthermore, AI algorithms and AI tools in healthcare are being implemented in well-resourced communities and well-resourced countries, to the point where if AI does provide better quality of care either now or in the future, it may be doing so for more privileged communities. So there's a potential risk with AI of, in fact, worsening disparities in care. But if the AI is designed to address the disparities and address and identify the root causes of some of these differences that we see in patient outcomes, it can be a potent way to reverse inequities.

There's some data that using real time registries to track patients through the continuum of care in oncology can allow you to eliminate racial disparities in care there. If you take a new diagnosis of lung cancer, and you plug a patient into a registry saying that they should have treatment for their lung cancer within 6 weeks, and this treatment does not occur and the registry tracks that and notifies the provider in an automated way, that we can find when patients have fallen off the track and get patients to complete the courses of therapy that have been recommended. And there’s some evidence that those types of registries, those types of real time patient-monitoring algorithms, can eliminate disparities in treatment completion, which can subsequently eliminate disparities in outcomes.

Pharmacy Times: What do you think may be the rollout time for the adoption of some of these technologies in oncology care?

Manning: There are already tools available to clinics that offer [AI] and business insights. And they include things like overreads of screening CTs for lung cancer, where computer-assisted detection helps radiologists to find small abnormalities. They don't replace physicians, but they do help assist physicians in their current state. There are also tools like there are tools that allow chemotherapy infusion centers to level-set their schedules and eliminate spikes in volume through the course of the day to help optimize the efficiency of chemotherapy infusions. And in doing so, they can increase the throughput of patients who need chemotherapy and avoid delays in care. And they can also allocate the resources throughout the course of the day in a more measured way to help the workforce. And many of those tools exist in current state. However, I do think in the future, we're going to see disruptive technologies, where not only clinical decision support for physicians but also patient interactions, such as chat bots, remote monitoring through AI will help augment the quality of medical care that we offer now. And I think that the implementation of that is going to be rapid over the next 5 years.

Pharmacy Times: What do you see as being the future of the application of these technologies in oncology?

Manning: Well, I think that AI is helping us in drug development now, so helping to identify potential targets for new drug development. I think that in individual patient care, AI is able to aggregate a patient's genomic and genetic information along with their environmental information and their disease characteristics to help design a tailored plan of care for each individual patient. So, I think that in the future, we may see that the clinical decision support tools that currently just help remind providers about screening and things like that, that those clinical decision support tools will actually lay out plans of care that are based on a volume of information that an individual human being couldn't really ingest and process. And so, I think in the future, we will see more and more clinical support and clinical guidance on treatment regimens based on AI.

Related Videos
Image Credit: © javier - stock.adobe.com
Image Credit: Adobe Stock - nataliaderiabina
© 2024 MJH Life Sciences

All rights reserved.