Q&A: How AWS Is Tackling Clinical Workflows with Generative AI

HEALTHTECH: What trends are you seeing related to clinical workflows at healthcare organizations?

SHIPPY: I have the opportunity to talk to customers all across the U.S., and there are a couple of things they tell us are top of mind for them. One is what’s happening with their workforce. They want to make sure they have all of the tools they need right at their fingertips, as well as seamless experiences, meaning that when they’re working, the tools they’re asking for are part of their workflows.

Where we’re seeing great opportunity is in how we make it easier for them to document. How do we make it easier for them to capture what’s being told to them by the patient? How do we make it easier for them to take the information they have and turn that into the clinical insights they need?

A big part of that’s going to come from bringing all of the data together and having it in front of them when they’re seeing the patient. If I have information from the electronic health record, information that the patient just told me, along with outside records, can all of those be summarized? That way, as the clinician is right in front of the patient, he or she has all of the information. Instead of looking down and constantly typing or looking for things, they can have a meaningful dialogue.

Part of how we’re doing that is by providing generative AI tools to summarize all of those clinical notes. Another way to do it is by having organizations bring all of their data together in a data lake. Finally, we’re working with EHR vendors to say, “How do we put these add-on applications into the workflow, so the clinician is never leaving the EHR?”

PREPARE: Expert guidance helps healthcare organizations achieve meaningful transformation with AI.

HEALTHTECH: How is AWS working to improve clinical workflows with its solutions?

SHIPPY: Not only does AWS have a broad and deep set of cloud services but we also have purpose-built services that are just for healthcare. We have a service called AWS HealthScribe, which allows clinicians to focus on having conversations with their patients. Those conversations are then transcribed and made readily available so that they can quickly annotate them, make any corrections needed and get to the next level of care. That’s one way of being able to do it.

Another is with AWS HealthImaging. We can take one image and use it in the cloud, allowing that same image to be used over and over and over. Why is that helpful? The first reason is that it’s readily available to everybody who needs it. The second is that it’s one image being saved, not multiple images being saved. There’s a cost savings in using that particular service.

HEALTHTECH: I’ve heard that AWS HealthScribe was created to be used by all clinicians, not just physicians. How does it work for care teams?

SHIPPY: I’m a physician, so I can say that there are so many times that we talk about the physicians, but physicians know that we are part of a larger care team. There are nurses and allied health professionals, such as occupational therapists, physical therapists, respiratory therapists, etc. Everybody on the team needs better tools to help them do the work that they do and to practice at the top of their licenses, which is so incredibly important for their resilience.

DISCOVER: What does the future of large language models look like in healthcare?

HEALTHTECH: How else can healthcare organizations benefit from AI solutions?

SHIPPY: Healthcare organizations can benefit from AI solutions because they can help them with their operational efficiency. AI can help them anticipate what resources are needed. Predictive analytics can help organizations determine how many patients usually come in during a specific time of year. What areas of the hospital or the clinic are they usually in? What supplies do we need? What does the staffing need to be? It’s going to help them become much more efficient.

Then from a clinical standpoint, AI can help the organization ensure that it has all of the information it needs. How do we ensure that we understand everything about this patient so that when we’re ready to discharge them from the clinic or from the hospital, they have all of the supplies they need and they know exactly when their follow-up appointment is going to be? How can we make that a little bit easier for them?

Then you can start to look at what it means for patients. When they are outside the four walls of the clinic or the hospital, do they have all of the information they need? Can we better educate them on their disease process? Can we help them manage it at home? Can we give them prompts to let them know when it’s time for a medication or to do a particular exercise? Those are all of the things that are going to get better with generative AI and solutions coming from the cloud.

HEALTHTECH: For healthcare organizations that haven’t yet taken significant steps into the AI space, what can they do to prepare to engage with these technologies?

SHIPPY: I’d recommend that they be curious. One of our leadership principles is to learn and be curious. Find out as much information as possible. Talk to other healthcare providers to see what they’re doing. Come to us. We have a lot of information on how to get started in the cloud.  

You should start now. In healthcare, many times when I used to wear my physician and administrative hats, we would say, “We won’t be first, but we won’t be the last.” This is one of those areas where you really want to be first and out there learning, innovating and transforming how patients are taken care of.

HEALTHTECH: What are some common challenges healthcare organizations face as they go to implement these solutions?

SHIPPY: I think everybody’s incredibly excited because there is so much opportunity here. They’re considering whether they want to try things in the clinical setting or help in the back office. I think it’s all of the above. It’s important to determine which applications and use cases are going to be manageable and which ones are going to make it easier to take care of patients. Use cases such as summarizing records and bringing information to the forefront are being looked at and they’re absolutely making a difference, but so are the back-office use cases. You can make it easier to schedule an appointment, or for a patient to come to that appointment prepared. While those aren’t clinically facing, they are very impactful for the actual clinical visit. I’m seeing organizations right now focus on both sides of it.

The other part is how can you make it easier to take all of the research that’s being done and go from bench to bedside even faster? It used to take a long time to look at all of the data from a research study, bring it all together and get the clinical insights. That’s happening faster now because generative AI can help with some of that summarization early. It annotates the information and labels which study it came from, allowing the research process to be much shorter and accelerating the time it takes to understand what the research is showing us and what it means from a clinical standpoint.

EXPLORE: Experts discuss considerations for the responsible use of generative AI in healthcare.

HEALTHTECH: What do healthcare leaders need to know about the future of clinical workflows?

SHIPPY: One of the things about being a physician or a nurse is that we’re trained in an apprentice model. We learn how to do a particular procedure or round on our patients from attending physicians and preceptors, respectively. Going forward, we’re going to have to incorporate technology into the training.

Going further upstream, how are we using technology for our medical students and for our nursing students? How are we utilizing it in the day-to-day? Then, once you get into practice, it’s not like somebody’s teaching you then how to incorporate technology. It’s just been part of your training, so it’s easier.

The second thing is really embracing what cloud technology and cloud innovation open up. If you think about the history of AWS and how we came to be, it was formed because we were helping to power Amazon.com and thought the cloud could help other customers solve a problem. That’s going to be true in healthcare. There’s going to be technology that empowers how we care for patients in the future, and everybody needs to understand it and be able to use it. That’s really exciting. 

HEALTHTECH: Is there anything else you’d like our readers to know?

SHIPPY: We’re going to be supporting pediatric research through the AWS IMAGINE Children’s Health Innovation Award. We’re so excited about the opportunity to amplify children’s causes and allow those researchers to have more resources available to them. We’ve already started with a philanthropic commitment to Children’s National Hospital, Nationwide Children’s Hospital and the Children’s Brain Tumor Network, but we’re also going to be opening up the award to other organizations.

We also want to look at research to improve the well-being of children and those who take care of children. We’re super excited to explore those topics.

We’re also supporting generative AI with the AWS Worldwide Public Sector Generative AI Impact Initiative for Public Sector Organizations. We know there are many organizations out there with ideas and use cases that they want to implement. We want to support them with a grant that’s going to either give them resources or credits to help get that project started. We want to see what they do that’s going to help advance the use of generative AI.