Why patient engagement platforms must reduce staff burden [PODCAST]

Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes

Orthopedic surgeon Kevin J. Campbell discusses the article “Understanding Generation 2 patient engagement platforms.” Kevin discusses the article through a practical distinction between first-generation patient engagement tools that increase access but also increase staff workload, and second-generation platforms designed to answer routine patient questions without pulling every message into an inbox. He explains why portals, apps, messaging systems, and automated content often fail when they turn staff into message processors, and why the next generation of platforms must be built around clinical protocols, physician preferences, clear escalation boundaries, and staff-relieving workflows. This episode explores patient engagement, patient portals, health care technology, surgical communication, staff burden, practice efficiency, and the difference between digital access and true operational leverage. It is a useful conversation about how technology should support patient care without simply making it easier for work to find already overwhelmed teams.

Partner with me on the KevinMD platform. With over three million monthly readers and half a million social media followers, I give you direct access to the doctors and patients who matter most. Whether you need a sponsored article, email campaign, video interview, or a spot right here on the podcast, I offer the trusted space your brand deserves to be heard. Let’s work together to tell your story.

PARTNER WITH KEVINMD → https://kevinmd.com/influencer

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

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

Transcript

Kevin Pho: Hi, and welcome to the show. Subscribe at KevinMD.com/podcast. Today we welcome back Kevin J. Campbell. He’s an orthopedic surgeon. Today’s KevinMD article is “Understanding Generation 2 Patient Engagement Platforms.” Kevin, welcome back to the show.

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

Kevin Pho: All right, so tell us what led you to write this particular article on KevinMD.

Kevin J. Campbell: My interest in writing the article was really related to just the conceptual understanding of patient engagement platforms and the effect that they can have on the clinical office. Over time we’ve sort of been studying these programs over the last 10 years, and we’ve bucketed them into kind of two different versions.

There’s sort of a generation one and a generation two. Patient engagement platforms, I think by definition, are just digital programs that help us keep patients on track as they prepare and recover from surgery. And the distinction that I make in the article is that generation two patient engagement platforms don’t have to be staffed, and that creates a lot of benefits compared to generation one programs, which really, by definition, have to be staffed by clinical team members, and that creates a whole set of burdens for them that are relieved with a generation two platform.

Kevin Pho: So to illustrate what a generation one platform would be like as it relates to, say, an orthopedic procedure, what exactly would that look like from the patient perspective?

Kevin J. Campbell: So, when patients have a procedure, offices will often have an opportunity to support them with a digital program. That may be something like an app that they can download onto their phone or a portal that they can log into or a feature set within the electronic medical record.

And so when a patient is ready for hip or knee replacement, a part of getting them prepared may be getting them enrolled in that digital program. And so, a generation one program from a patient’s perspective would be that they would receive information digitally through either that platform, a message, or an app, but that on the other side of it, if they submitted a question to the team, it would be a sort of a secure inbox that a clinician would respond to.

What patients may notice is that access to their physician’s team is gated by their availability. And so if they submit a question about maybe a fever on the first night after surgery, the response is gated by the availability of the clinician, and so they may not receive a timely response, which ultimately does not necessarily help them in terms of their quest to stay sort of educated and informed as they prepare and recover from surgery.

Kevin Pho: So this sounds very similar to the MyChart messages, of course, with Epic. In my primary care setting, like you said, I’m not in the office every day. I don’t necessarily check Epic every day. So there is going to be a little bit of a time lag when people send me a message through the electronic medical record.

Now, other than the fact that there needs to be an actual clinician or clinical staff member who needs to review the messages, what are other downsides from the generation one models that we’re using?

Kevin J. Campbell: I think what you said is true. If you have a high volume practice where you give patients sort of unlimited access to your team through a digital inbox, it requires a lot of support.

Physicians and their teams may have to hire an entire staff member or multiple members to manage those questions that come in because you’re essentially exchanging messages back and forth. Teams may have to hire an outside firm of nurses to staff that, sort of be a ghost behind the scenes acting on behalf of the clinic, responding back to patients’ questions, and certainly that’s expensive.

And then for patients, there is this inherent delay, especially if you’re outside of the clinician office hours, they may not receive a timely message or response to their question. And so I describe in the article this narrow relevant biological window, which is that when patients are recovering, something that’s concerning to them may no longer be concerning four to six hours later.

So if they’re having nausea, they’re going to want an answer to help them relieve that symptom immediately, right? They’re not going to want to wait four to six hours feeling sick before you’re back in and staffing the inbox. And so, that 24/7, 365 access in a generation two platform where maybe the engine powering the response system is like AI or a digital version of the physician, I think that that’s really powerful.

And the example that you shared I think is really relevant, right? So certainly you as a primary care physician have your treatment preferences related to things that patients are going to have questions about. Maybe that’s vaccinations or being treated with Tamiflu if they’re feeling sick, or if they need to receive treatment if they tested positive for COVID.

And they may submit that question to you on Epic or in a MyChart message. And just because you’re not physically available to respond to their question, I think doesn’t necessarily mean that the patient should not have access to your treatment program. So you could imagine a world in which if you deployed a generation two system and the patient tested positive for flu at home, they could text into your sort of digital version and say, “Hey, doc, I just tested positive for the flu. Do you recommend I get treatment, or do you recommend I get started on Tamiflu?”

And your recommendation could come back to them as a digital response that says, “Hey, if you’re an otherwise healthy person without diabetes and other medical comorbidities, I typically recommend just supportive care. But if you’re a patient of mine who has diabetes or heart disease or lung disease, typically I do send in Tamiflu. And so, if you fall into that bucket, give us a call.”

I imagine a digital world in which in the future, that messaging program is even more intelligent. It already has the patient’s medical context. It already has your medical context. They’re interacting in this messaging program, and maybe the human or clinician in the loop only comes in at the finish line where you e-prescribe or you hit authorize on that Tamiflu prescription. So it’s really evolving, but I think that’s ultimately the vision.

Kevin Pho: So tell us how that’s different from, say, a patient if they get a delay with a generation one platform and they just go on a publicly available model like a GPT, Claude, or Gemini and simply ask their question. So draw the distinction between the solution that you’re describing versus a patient just using any publicly available model to ask that same question.

Kevin J. Campbell: Yeah. A publicly available model does not have your physician’s saved treatment preferences in the model, and so the model’s response is going to be generated based on its training data. And so it’s not going to be able to deliver your treatment preference, and oftentimes those models will default to maybe a more aggressive treatment program that may not be exactly what your physician wants.

And I think when I say may not, I would say it probably is not. There’s a lot of published articles that are encouraging people not to go to a model to answer their medical questions. But I think why people like models is because it gives them instant information and instant answers. They may not be the best answers, but it’s about the time relevance of it.

And so if you could marry those two concepts, which is immediately available, plus it’s actually your physician’s recommendations and it’s staff relieving that it did not have to come from a human, I think that that could potentially be a sweet spot that people are going to be very interested as they explore these sort of generation two offerings.

Kevin Pho: So from your experience, especially in orthopedic surgery and knowing how quickly the evolution of these publicly available models is, even though they’re not giving the best answers, are the answers they’re giving good enough answers?

Kevin J. Campbell: I would say no. Certainly they can provide enough context for the patient to sort of maybe better understand what question to ask their doctor or how to ask it. The answers in our world tend to be very specific. Bandage changes at three days after surgery or seven. There’s just a lot of little idiosyncrasies that are not captured in a generic model.

Do I think that there are benefits to generic publicly available models? Absolutely. I’m a huge proponent of it for a number of different reasons, but I think when it comes to especially something so serious as a knee or hip replacement and needing that specific guidance, people are always going to be looking for their physician’s recommendations.

Kevin Pho: So let’s go to that example, hip and knee replacements. Just in that orthopedic surgery world, is there variation between orthopedic centers, between orthopedic surgeons, especially for post-op discharge instructions for a relatively common procedure like a hip or knee replacement?

Kevin J. Campbell: Yeah. We support many physicians on the program. Being an orthopedic surgeon myself, I can kind of compare to my colleagues, but there’s not one recipe. There’s a recipe in terms of the general things to be thinking about, showering instructions, bandage care, therapy recommendations, medications, but everyone’s protocol, right or wrong, is different, and surgeons are sort of very narrow-minded in terms of what they’re willing to accept outside of their protocol, so it’s very difficult to get everyone to come to the same pathway.

And so, yes, there is a tremendous amount of variability and it affects patients significantly. And so when you think about a generation one program that needs to be staffed, that even brings up more idiosyncrasies of the staff having to remember the physician’s treatment preferences. So they’ll sit there with an Excel file, and they’ll have the surgeons’ names at the top, the topics of patients’ questions on the side, and they have answers to all of the physicians’ preferences.

So, for me, I may say you can have dental visits three months after surgery and antibiotics for the first year and then nothing in the future, but my partner, who does the exact same procedures as I do, may recommend dental prophylaxis for the patient’s lifetime, right? And those recommendations are constantly changing as new recommendations and research come out. So the short answer is everyone is so different.

Kevin Pho: Now I know that you offer a solution that provides this customized generation two model. So I’m interested in how you train that model to meet the specific requirements of whomever surgeon, medical institution, or practice that it’s tailored for. How’s your model trained?

Kevin J. Campbell: Models aren’t necessarily trained on the physician’s treatment preferences, but they have access to the physician’s treatment preferences when they’re responding to patients’ questions. So a part of onboarding a physician and getting them enrolled in the program is for them to sort of document their treatment preferences, and then that information is available to the model to answer the patient’s questions.

Really emulating the type of thing that we do with our clinician colleagues, right? So if you onboard a new medical assistant or a physician assistant, you sort of give them access to your treatment preferences so that when they staff a phone call, they’ll have your answers to provide to patients. So the model isn’t necessarily having to be trained on the physician’s data, but it definitely needs access to their treatment preferences.

Kevin Pho: How comprehensive is that onboarding? Because even within the niche of orthopedic surgery, I could only imagine that there are almost endless variations and scenarios that your platform needs to consider. So just give us an example of the scope that we’re talking about.

Kevin J. Campbell: The onboarding process has been refined over time. It’s actually quite turnkey in the sense that physicians themselves have already accumulated a sort of a mass amount of information that they give to patients. So those are in the form of instructions, FAQs, things like that. And so we’ll build the physician’s data repository off of the information that they’ve already created, leveraging technology. So while I think the repository that’s built for the physician’s treatment preferences is very robust, the onboarding process is not arduous, and it’s really leveraging technology.

It’s not something like your Epic example, right? Where a hospital goes live on Epic, and it’s a several-year process, and it’s championed, and they’ll sort of deploy an implementation team. That’s not at all what’s needed with Streamed. You can actually log online, set up an account yourself, upload your instructions, select the procedure, and then that builds the information automatically.

So, we’re doing that from a business perspective. We want it to be a scalable solution, leveraging network effects, a platform, will you say. And so, that’s something that we’ve put a lot of time and energy because no clinician has time to curate and build this stuff out by themselves. So we’ve been very mindful with that because, of course, we’re clinicians ourselves.

Kevin Pho: So tell us a success story where your particular generation two model really moved the needle and made a difference in terms of patient outcomes. So tell us what success would look like.

Kevin J. Campbell: Yeah. We study our programs. We’ve had seven published clinical trials with what it looks like to deploy our program even in a system that’s already a very well-oiled machine. And what we’re seeing is that when patients have access to their physician’s treatment preferences, there’s less ER visits, there’s less hospital readmissions, there’s less phone calls to the office, patients are taking their narcotics for a shorter period of time, and they’re more satisfied.

And I think that just kind of speaks to the quality of the program and the fact that I think patients are comfortable receiving information in an automated way, as long as it’s relevant, as long as it’s accurate, and as long as it comes to them as their physician’s prescribed preferences rather than having to determine if the general information is something that actually their physician would support or not.

Kevin Pho: We’re talking to Kevin J. Campbell. He’s an orthopedic surgeon. Today’s KevinMD article is “Understanding Generation 2 Patient Engagement Platforms.” Kevin, as always, we’ll end with take-home messages that you want to leave with the KevinMD audience.

Kevin J. Campbell: Yeah. I would just say, you can engage and educate your patients in a way that does not have to overburden the office staff. So there are programs available that leverage technology to keep patients informed, inspired, get their questions answered, but not have to build out a team of clinicians to support those patients. So, technology should be your next clinical staff member, and I think that’s really the purpose of the article.

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

Kevin J. Campbell: My pleasure. Thanks for having me.

Listen and followApple PodcastsSpotifyYouTubeMore apps · Browse all episodes