Subscribe to The Podcast by KevinMD. Watch on YouTube. Catch up on old episodes!

Orthopedic surgeon Kevin J. Campbell discusses his article “Health care is having its Yahoo moment.” Kevin explains how the current medical system is stuck in an obsolete pipeline model similar to early internet directories where human gatekeepers create dangerous delays in communication. He introduces the concept of delayed relevance where patient questions about fever or swelling lose value if answered days later and advocates for an AI-driven platform approach to provide immediate physician-vetted responses. The conversation explores how shifting to automated systems can relieve staff burnout while increasing patient trust through consistent and timely guidance. We must embrace this technological evolution to deliver the responsive and effective care that modern medicine demands.

Our presenting sponsor is Microsoft Dragon Copilot.

Microsoft Dragon Copilot, your AI assistant for clinical workflow, is transforming how clinicians work. Now you can streamline and customize documentation, surface information right at the point of care, and automate tasks with just a click.

Part of Microsoft Cloud for Healthcare, Dragon Copilot offers an extensible AI workspace and a single, integrated platform to help unlock new levels of efficiency. Plus, it’s backed by a proven track record and decades of clinical expertise, and it’s built on a foundation of trust.

It’s time to ease your administrative burdens and stay focused on what matters most with Dragon Copilot, your AI assistant for clinical workflow.

VISIT SPONSOR → https://aka.ms/kevinmd

SUBSCRIBE TO THE PODCAST → https://www.kevinmd.com/podcast

RECOMMENDED BY KEVINMD → https://www.kevinmd.com/recommended

Transcript

Kevin Pho: Hi. Welcome to the show. Subscribe at KevinMD.com/podcast. Today, we welcome Kevin J. Campbell. He is an orthopedic surgeon. Today’s KevinMD article is “Health care is having its Yahoo moment.” Kevin, welcome to the show.

Kevin J. Campbell: Thank you very much, Kevin. Thanks for having me.

Kevin Pho: Let’s start by briefly sharing your story. Then we will jump right into your KevinMD article.

Kevin J. Campbell: I am a practicing orthopedic hip and knee replacement surgeon working out of the Orthopedic and Sports Institute in Appleton, Wisconsin. I also work as the CEO and co-founder of STREAMD, which is our AI health care technology company.

Kevin Pho: Your KevinMD article contrasts what is happening now with Yahoo, that legacy internet company. For those who didn’t get a chance to read your article, tell us what it is about.

Kevin J. Campbell: The article is about medicine having its Yahoo moment. It really contrasts the approach that Yahoo and Google took in terms of organizing the internet. Yahoo organized it with human review in their editors, whereas Google took an opposite approach organizing it with their algorithms. So what we are seeing in medicine is a very similar thing.

Right now, a lot of the access to physicians and their teams is gated by humans. When we ask questions, we have to go through a network of people to get the answer. But oftentimes that results in some delayed information and delayed relevance. With the advent of new technologies like AI, we can allow and enable patients to get answers to their questions almost immediately. Those answers come at a time where they are helpful and informative for their questions. It gives us a new opportunity to rethink how we help patients prepare and recover from medical procedures and surgeries.

Kevin Pho: From an orthopedic surgeon standpoint, when you said that medical information or even access to an orthopedic surgeon is gate-kept by humans, give us some examples of what that looks like.

Kevin J. Campbell: Great examples would be a patient recovering from surgery being concerned about bruising and swelling. Oftentimes the best way to get that answer is to review either their instructions that are printed for them or call the office and wait for somebody to give them a call back with the answer. Paper instructions for whatever reason don’t seem to work too well. Patients just have a hard time finding the right information. When they call the office, typically they are met with a delay of anywhere between two hours to a couple of business days.

That results in a lot of anxiety and fear for patients that what is happening to their body as they recover from surgery is abnormal or concerning. By the time they get their answer, oftentimes those symptoms have either improved and resolved themselves or worsened to some degree. Patients are left frustrated by not knowing the right information. With new technology, patients have access to their physician’s instructions immediately so they can ask questions about bruising and swelling and get a response right away that helps them calm their nerves, reassure them that those changes are normal, and then they can continue with their recovery.

I like to think about it as navigating to a new place using GPS or Google Maps. Certainly, we know that the robot powering Google Maps is automated. But the information that it gives us is so relevant, timely, and accurate that we don’t care. I think patients are starting to feel that way about some of the aspects of their medical care. They are just really looking for a timely answer to their question that helps either resolve their fears or direct them into a better care pathway. That is really the concept that we are highlighting in the article.

Kevin Pho: Exactly what you said. As a primary care doctor, sometimes it is very difficult to see a clinician, especially in primary care, so they go online. Whereas back then they Googled their information, now we have AI and large language models (publicly available models like ChatGPT and Gemini) and patients are getting quite detailed answers to their questions.

Kevin J. Campbell: I think for patients it just allows them to have access to the best information as quickly as possible. It is better than the alternative of calling the office and not getting an answer. So I think that these new technologies, if used in the right way, can be very powerful to help patients feel more supported, experience quicker recoveries, and be more satisfied as they interact with the health care system.

Kevin Pho: So you said used in the right way. As an orthopedic surgeon, as patients are going online or to AI to research their problems, what are some of the pitfalls that you see in your practice?

Kevin J. Campbell: Great question. I think as patients use “Dr. Google” to answer their questions, the trouble with that is that sometimes Dr. Google is not giving the same recommendations as their own physician, or the instructions that Google is providing is not as specific as they may need for the instructions to be helpful. With our company STREAMD, we use the physician’s instructions as the knowledge base that the models reference when answering patients’ questions. So unlike talking with Gemini or GPT about any health-related concern, when they ask questions into the physician’s messaging program, they get answers that are prescribed by their physician.

There are very specific answers that match the care culture and treatment preferences of their doctor. So patients can feel that extension of the physician when they are interacting with our messaging programs rather than the generic instructions that are provided by Google.

Kevin Pho: When it comes again in orthopedic surgery (that is your field of specialty), what would be an example of some of the variation between general information on GPT and something that is more personalized from yourself?

Kevin J. Campbell: Just a simple question like: “When can I take a shower after knee or hip replacement?” or “When can I change my bandage?” Asking those types of questions online, they are going to get a generic response that says: “Oftentimes your surgeon will put a bandage on that can be left on for anywhere between a couple of days to a couple of weeks after surgery. You should just follow your surgeon’s prescribed instructions.”

That is so generic that it is actually not that helpful for patients and can create more questions and more calls to the office. If those questions are texted into the STREAMD messaging program, they are going to get an answer that is very specific and prescribed by their physician. That response might say something like: “Your bandage is waterproof. You can leave it on for the first seven days of your recovery. When you shower, you can let water wash over the bandage. Once a week has passed, you can take off the bandage and leave the incision uncovered at that point.”

That is such a comprehensive response that it answers the patient’s questions immediately. It doesn’t require any staff involvement, and patients are able to go on with their day. When they have a positive interaction like that and it provides them the information that they need, then they will keep coming back to the program with additional questions because they are trusting that process as they go through it. Just like when we are driving down the road and somehow Google knows about construction that we are not aware of and it reroutes us, it enacts additional trust in that program for us. That is really what we are trying to harness and build on. Patients know their docs can’t text them one or two times each day. So they are comfortable with it being an automated program, but the information matches their treatment preferences. Thus, they become very accustomed to interacting with that channel.

Kevin Pho: How common is it that there is such a variation, for instance, with discharge instructions between orthopedic surgery practices?

Kevin J. Campbell: It is very common. Each surgeon has their own specific training background or little nuances. While you could imagine: “I had a knee or hip replacement; shouldn’t it be the same?” we all have very different treatment instructions. It makes two things very difficult for patients. They are not able to find the information online. They are having to search through their physician’s instructions.

For the office staff, it creates a lot of friction because as we work with our medical assistants or nurses, they may come to learn my treatment preferences, but if they are cross-covering another physician, they may not know theirs. So you are having to get this cross-coverage and multiple people training. It is just a lot to remember. I will even see in our practices that the nurses will keep little cheat sheets, Excel files, or Word documents housing physicians’ treatment preferences, and they go to those as patients have questions. We are really trying to block and tackle those common challenges in the clinic and deploy these digital messaging programs that allow patients to get that instant information.

Kevin Pho: Give us a sense from the patient experience standpoint when using your AI solution. Tell us a story or scenario about what it is like for them.

Kevin J. Campbell: As patients sign up for surgery, certainly they are going to have a lot of questions and concerns. Our office time with patients is limited. It may even be limited to just 20 or 30 minutes. What I always tell my patients as they say, “There is a lot to remember,” is: “Would you like to receive some text messages to help you prepare and recover?” “Of course I would.” “OK, great. Have you ever received a text message?” And they will say, “Yeah, I have.”

So we sign them up for the messaging program, and those messages will start trickling into their phones for a couple of weeks before surgery and for the first six to eight weeks afterwards. Those messages, of course, are informational, but they are also supportive. Patients feel a connection to my office because I am highlighting all the specific challenges that they could have emotionally as they have anxiety as they prepare and recover.

All of my patients say that they love my messages. It helps keep them on track as they prepare and recover from surgery. They comment that it feels like there was a camera watching over them. It just seemed like the right information came at the right time. I think that is built out of the fact that I am an orthopedic surgeon. I know what questions they are going to have and when. Helping to give them that information just makes that recovery process so much smoother. It actually feels a bit like a concierge experience for them without having to increase the headcount and the overhead of my practice. To provide this level of support to patients in a manual way (just like what Yahoo was having to overcome), we would never be able to do it. We just couldn’t do that with a human team. That is why we have leveraged the technology to act as an extension of our team.

Kevin Pho: How do you train your model with all your different preferences and customizations that you want to give the patient?

Kevin J. Campbell: When we onboard physicians into our program, we collect their treatment preferences, and then those treatment preferences serve as the knowledge base that the models reference when replying to patients’ questions. So we are not having to spin up our own models and create our own training environments. But those models are held within guardrails based on the treatment preferences that are stored in our system.

Kevin Pho: Do you have any clinicians or even your colleagues where they may be a little bit technology averse and maybe want to give more personal recommendations rather than something that is AI-generated?

Kevin J. Campbell: I think that there is always that initial apprehension that says: “This technology is not going to be as good as talking to my staff or talking to me.” But what we have found is we have an extremely high customer retention rate. Anyone who has started to use our program has continued to stay with us. We are not trying to replace the physician or team. We are just trying to augment it and be an extension of it.

We have always had our focus on simplicity and ease of use. That is why our messages come through traditional text messages (SMS). Patients are not having to download an app, create a username, or remember a password, and that just makes it technologically appropriate for patients to interact with. On the physician side, there is nothing for them to manage. It really feels invisible for them, and that is purposeful. We knew that if we had to create additional workflow or have the physicians monitor another clinician inbox, it would never work because we are actually trying to cut down on staff responsibilities rather than add to them. I know if I brought it to your clinic and I said: “This has to be staffed. You are going to have to hire one to two full-time employees to manage this software, but it is great for the patients,” well, that is great, but it is not going to work in our clinic.

Kevin Pho: Any issues with things like hallucinations or if it misinterprets the syntax of a patient question, they’ll give the wrong information? Any concerns about that?

Kevin J. Campbell: It is always a fear. Certainly it is something that we monitor. Hallucinations are a problem when context is not included. In all of the interactions that patients have with the program, we embed their patient-specific context within that conversation as well as paired to the physician’s treatment preferences. Keeping the model very focused and aligned with that reference source is very helpful to prevent those hallucinations or the models kind of veering off transversely.

Kevin Pho: We are talking to Kevin J. Campbell. He is an orthopedic surgeon. Today’s KevinMD article is “Health care is having its Yahoo moment.” Kevin, let’s end with some take-home messages that you want to leave with the KevinMD audience.

Kevin J. Campbell: Firstly, I would just like to say thank you for having me and keeping your mind open to this new technology. I think anytime you are bringing technology into your medical practice, there are some things that you have to overcome because of the fear that it is going to be like the technology that was brought into your practice in the past. We really try to take an opposite approach here, helping the patients but also the physicians and their teams to create a more efficient workflow while making patients feel more supported and cared for. I would just like to say thanks again for having me. It has been a pleasure to share our story with your audience, and I am always available for any questions or follow-up commentary.

Kevin Pho: Kevin, thank you so much for sharing your perspective and insight. Thanks again for coming on the show.

Kevin J. Campbell: Thank you.