Innovating Medicine: How UpHill Is Transforming Clinical Pathway With AI

Ayush: Hello, Eduardo and Mariana. Glad to see you at HIMSS. And you know, at HIMSS, I’m always super excited to see all the innovation that is happening in health.

You know sometimes I in fact get overwhelmed a bit, to be honest, meeting all the great companies. But glad to know more about Uphill and I thought that it would be super interesting to take this story to our audience and let’s talk about Uphill. So let’s start with you, Eduardo, tell us more about Uphill and your role.

Eduardo: Sure. So we are a care orchestration platform, so essentially we automate clinical pathways for providers, and by doing that, we increase the capacity of that hospital or clinic to serve more patients, which is very aligned with health systems are leading right now in terms of capacity shortages due to essentially workforce struggles that many countries are seeing worldwide.

Mariana: And I mean, by increasing capacity, of course, we are decreasing waiting times, waiting lists, and this impacts greatly also patients outcomes, which is key when we’re talking about health care. So, yeah.

Ayush: Okay. So, would you like to go a bit more deeper into helping us understand the role of uphill in a hospital setting, maybe just give a bit more imagination if you can.

Mariana: Okay. Okay. So as we were talking about care integration, so Uphill’s approach is an approach that is process based care orchestration approach. So what we do is that the software acts like a second layer, as a background layer to the EHR, so to the electronic health record, where we bounded by a clinical pathway that is based on evidence, internal protocols, are automating tasks that must happen within this, perspective.

So what happens is that based on events that already happened on the EHR, there are triggers to our platform that automate actions to the healthcare professional. And that’s where we actually save a lot of time and ensure that the best practices are being implemented. So like automating clinical and administrative tests.

Eduardo: A good example would be, for example diabetic patients. And once we connect through interoperability with the electronic health record system, or the broker underlying that electronic health record system, we’re able to identify patients that are diabetic and are in risk or for example complications of diabetes, we enroll them in a clinical pathway automatically and that clinical pathway can do many things for that healthcare team such as for example routinely prescribing follow up exams and tests that the patient can do.

The results are then imported back to the system and that is a trigger for example anticipating an appointment earlier on or even at the same time adjusting therapy through decision making for the health care teams. And so essentially what we’re doing is the first time bridging the gap between appointments and different levels of care, for example, from primary care to secondary care to home care, as well as speeding up the processes of these clinical pathways inside the hospital.

And it’s typically the benefit results in removal of what we call low value clinical tasks from the heads of Clinicians so that they can focus on high risk patients. They can focus on complex management by complex diagnosis complex workups for patients.

Mariana: Because we have actually nowadays healthcare resources that are highly valuable. All of them are, of course, but even their time costs a lot of money and we have them like spending 30 percent of their time in these low value tasks. So this is time where they could be seeing more patients and that’s how we do increase the capacity.

Ayush: So Eduardo you know, you have been a provider yourself in the past life.
So you know, what was that aha moment that you thought about building something like Uphill? In fact I would also like to hear the story about the name Uphill.

Eduardo: Yeah, sure. They’re tied together, actually, these two, these two questions. Well I was very interested in medicine. That’s why I went to medical school.

Because I wanted to know how the human body worked. Right. So and what I found out, actually, is that there’s huge information asymmetry in the healthcare space. Much like any other sector where Liberal professionals are the ones that essentially define that, that space. And that most of it was tied on how to manage disease, essentially initial workup, diagnosis, stratification, treatment lines, and so on, and follow up.

And in terms of the system itself, if you went to a hospital, it was chaos everywhere. And I’ve been fortunate enough to work in Portugal and other countries as well and have seen the same story over and over again. And so what we started to think was, can we essentially externalize this knowledge to a platform?

That would be able to help physicians, not like EHRs that are typically just there for record taking and information access, but actually knowledge systems that can help the physician decide and help them manage disease over time. And so I had a background in clinical research, which gave me a lot of knowledge in terms of evidence based medicine as well as to assess medical interventions and hence the logo, which is a triangle with a dot climbing the Mountain, so the name actually came from the evidence pyramid.

There’s this symbol in evidence based medicine, which states that High-level clinical evidence, high quality clinical evidence is at the top of this triangle. And those are essentially what we call evidence based decision support systems. And those are built on top of all the evidence that was generated by the medical field from simple case studies to systematic reviews and meet analysis.

And so the idea of the logo in the name Uphill is to bring the teams through this, which we know is a strenuous challenge of moving from the status where they have fragmented gear. They don’t have specific externalization of knowledge they can rely on, into a best practice approach that streamline and embedded within the systems.

It’s also because we were found that at the hillside of the biggest mountain in Portugal. So that also was an interesting one.

Ayush: Was the idea incepted while you were going uphill? Yeah.

Eduardo: Yeah, exactly. Okay,

Ayush: Great. And I understand like the name is the best whatever gets you meetings. So if it stays, then it’s good.

So Mariana,since you are you know, managing more the marketing and go to market side of Uphill, now tell us more about you know, your experiences of dealing with hospitals and you know, like, is this a problem that the hospitals recognize and understand or they are still ignorant and they are happy in their own, shell that, you know, this is how we do it. We have been doing it. How has been the market?

Mariana: Okay. Okay. I mean, it’s different from hospital to hospital for, if we’re talking about the public sector or the private care sector. When we’re talking about the public sector, the pains that we see the most are we just, what are Eduardo say a couple minutes ago, like medical shortages, the lack of capacity, this struggling to meet the needs that are increasing. So they do recognize the pain of the lack of capacity and we are not increasing exponentially the number of resources we have available so that they do recognize they need technology to help them with that, increase the capacity, leverage their knowledge through technology.

The private sector is a little bit different because the focus might be more related to quality, experience, nonetheless, cost optimization, exactly, cost efficiency. So this also impacts because if they increase the efficiency of their teams, they decrease costs and they can see more patients. So that’s definitely something they do feel like they need.

Eduardo: Yes. I think that you cannot solve an exponential problem with linear solution. And I think that most governments and provider groups are trying to keep up with demands by bringing more doctors and nurses in. But that just won’t cut it because we have an exponential problem that is growing on top of aging populations, chronic diseases. Chronic diseases are now diagnosed early on chronic diseases that have complex therapies to be managed.

You have the wellness and health care being blended together. So there’s no differentiation, No more and on top of all of these you have emergent diseases like Covid. And so it’s just completely unable for providers to keep up with this. And so The thing is they don’t feel the pain points this way.

Let me tell you a quick example For example, heart failure patients, they now should have the dosage of the drugs curated, to the maximum dosages in just a few weeks time. If not, symptoms are going to persist and they will end up in the ER. Every hospital that we work with, has struggle with keeping up with the demand in the ER, but they cannot afford to have multiple cardiology or medicine appointments with these patients like every three days just to curate the dosage of a drug.

But if you build a clinical pathway that assesses the symptoms automatically, built on top of healthcare data coming from platforms and suggest healthcare teams to do rated doses even remotely. Even nurses can do it remotely based on a protocol that has been concentralized with medical doctors. And so you’re able to get those patients to maximum dose on time, if they need it, and then prevent re-hospitalization and ER admissions. Okay. It’s very good example that, you touches and touch points that they have and the need that they have in a very simple way in trivial way and that has direct impact in their day to day life.

Ayush: And you know, when you talk about you know, being able to find those touch points. So for Uphill, like is this data that resides in the hospital system, like an EHR, or you’re also capturing data that resides say on a patient’s personal device or outside the hospital.

Mariana: I mean, we’re doing both. There are several ways to input data in this care journey. So of course the EHR is fundamental. That’s why we are such integrated, integrated with them. So integrated with them. But there’s also the channel between the hospital and the patient. That’s also fundamental because if the patient is home, we need data regarding the patient when the patient is home changes in signs and symptoms, for example.

So yeah, we do also keep, we give the hospital the possibility to have that direct channel with the patient to input data in their journey.

Eduardo: Yeah, we do it that way and think about everything we do at Uphill, essentially what we advocate as being journey based care is this idea that you should start with the clinical pathway and then the data requirements will come from the clinical pathway.

So, you know, to treat and diagnose colorectal cancer, you need data from colonoscopies, you need data from lab results, you need data from CT scans and MRIs. So we don’t have a one size fits all in terms of interoperability approach because the data is minimized depending on the pathways that are run in the hospital.

And so there isn’t things in terms of interoperability and in terms of change management and data privacy for patients.

Mariana: And also in the patient’s perspective, we are the least. Disruptive we, we, we can be so there’s no patient app developed by us. We use channels that are known to the patient like phone calls, text messages, emails, etc.

Eduardo: Today, many providers, I think most providers have already, set up call centers and apps and web portals. And so they have all the stack They just don’t have the intelligence to use that stack in favor of the patient. Providing that knowledge that is about the management.

Ayush: Building tech is easy, building intelligence is rather tough.

Eduardo: Exactly, and when you merge both the value is just way more relevant for physicians. And I think that in everything we do, we see it from the lens of clinicians. So they always ask, and I think it’s a fundamental question, is the stack going to improve outcomes? Or is it just going to be incremental? And if we think about ambient AI, newer technologies we get excited about it, but in the end, if you’re just recognizing speech and putting that into a note, that will be incremental for the patient journey.

That will not change the, the outcomes, the health outcomes for that patient in particular.

Ayush: So, Eduardo, a question to you since you know, you had your journey like from being a physician to you know, becoming an entrepreneur and, you know, a lot of our audience. Has this question, like you know, especially for a health startup, like, you know, finding your first customer or getting your thesis right early on, for example, convincing a health system, a larger health system to use the product, like, you know it’s always hard.

So how was that experience for you and how did you crack the early code you know, to be able to get yourself in, into those early customers, those early wins?

Eduardo: It’s a strenuous journey and the courage. Yeah, no, for sure. I think it’s, it’s, it’s all about courage and some persistence, but I think you just have to assume ideas don’t matter and it’s all about iteration.

So for example, we started aiming to this knowledge or this asymmetry of knowledge and information. In terms of the best practices and we started with the product that was directed to training physicians. Okay, but we found out earlier on that they didn’t want to be trained that frequently and that training didn’t have the retention that we expected.

Ayush: So training about care.

Eduardo: Yeah. Okay, and so we developed a decision support system essentially, we’re expecting physicians to go there and follow the best practices. And we understood very rapidly that they don’t want to do that. They don’t have time to do that in the point, at the point of care. So we essentially discovered the thesis that in order to increase compliance with best practices, you need to provide productivity value for the physician.

And that comes through automation. At the same time, you’re also complying or assuming that the pathway is going to be complied with. And so I think that the, the two lessons I can take from this journey at this point is the need for iterating the idea into product. And I think the faster the iteration is, the more successful the product will be.

And the second is, which essentially challenges the first one, but is the note that solutions in healthcare cannot be half picked. You need to have a full fledged solution that ticks all the boxes in order for you to be successful.

Ayush: As you’ve lost your prospect for life.

Eduardo: And I think that there are many companies in the digital health space that have this approach, this agile approach about testing something that has a core feature but has nothing around it.

It has no interoperability, it has no safety safeguards and it has no compliance with the standards. And that is a risk.

Ayush: So Mariana, since you know, you are someone who has the pulse of the market, like, you know, what’s happening. So over the years you know, like, as an outsider, I hear a lot of you know, renaissance and revolution going in the European healthcare.

But as a provider of a software, do you see that on the ground? Like, have you seen the attitude changing towards technology by the hospitals.

Mariana: Yes, yes. You know, we see that nationally. It’s a European level. There are initiatives, digitalization, interoperability, etc. I think our approach as always or almost always been like a step further.

So we’ve been told quite often, even here at HIMSS that we are in the future. But the, the, the thing is that we do have the results in the present and we are,

Ayush: Has it become easier to crack a deal?

Mariana: I mean, yes, yes. Once you start having results, once you start showing that this impacts not only health teams efficiency, but also patient outcomes in a major way. Of course, they do want that. And even patients, they feel more accompanied by the hospital. So yes, it is.

Eduardo: And I think that, There has been historically an approach from providers to develop their own solutions. The pressure that clinicians themselves and patients are putting in IT teams from healthcare groups major systems and simple providers and clinics is completely impossible to manage internally.

And the systems require compliance as a medical device cyber security compliance and so on and so forth. And so I think that providers have been more willing to experiment and to buy solutions from providers are actually showing that they are able to move the needle in terms of outcomes and cost efficiency.

Mariana: Then there’s also the different, the difference when it comes to technological providers that we see a lot with or without AI that they specialize in a specific part of a journey, for example, like only the post discharge part of the journey or only on oncological clinical pathways, something like this.

So they do not have, as much of an integrated perspective of the entire journey of a patient within a system, within a hospital. I mean, a patient that has heart failure can also go to surgery for another reason. So this should be a perspective, a hospital should have a single perspective of the patient across different pathways.

Eduardo: And it costs a mountain just for drugs. Right. If you have the same perspective for technology, it’s just impossible to keep the sustainability of healthcare systems as well.

Mariana: And also limits a lot the scalability of the solutions, especially if we’re talking about the public sector. If you develop something specific for a hospital, but you want to have it nationwide, it probably won’t fit the same way in all hospitals.
So the scalability, the interoperability is, is key. Yeah.

Ayush: And what about AI? So has that made life easy or difficult or both? Maybe, you know, there could be different contexts like where it has made life easy and where it has made life difficult.

Eduardo: I think that AI has actually broadened the horizon of decision making.
I think that was the first advantage. Not even a technical one for us and for them. They just, they were confronted with the speed that technology is taking. And that they cannot hide it under the rug. And that has actually contributed to an easier go to market. The thing about AI in practice, we use it a lot internally in terms of communication channels, building pathways, iterating the pathways.

But I think that hospitals are also falling into another extreme, which is now let us put Generative AI everywhere, and if you think about it, if you have Generative AI in each step of the way, this has a compounding effect on the risk, and that’s why we go back to basics and we think about clinical pathways.

Define your journeys first. And understand where AI can be leveraged and the perimeter of safety for specific models, that being clinical models or operational models. And then you are taking a deliberate approach to AI, a hospital and AI is coming from the windows and all the doors, and you’re not controlling it, and you’re not controlling the effects for patients.

Ayush: Anything you want to add?

Mariana: No, I mean, I think it’s, it’s, it’s fundamental to have this clinical pathway and evidence, of course. It’s bounding all the recommendations and suggestions Gen AI may give to healthcare professionals, for example. Yeah. So back to basic.

Ayush: So when you say easy, like it makes an easy go to market do you feel that it also increases the competition? Because now probably like more people can claim, you know, building that solution.

Mariana: Yes. Yes, definitely.

Ayush: Like the intelligence piece, like you mentioned, you know, more people can say that they have that intelligence. Yeah. Yeah.

Eduardo: I think that I think the overlap and saturation of the market is happening, but I think more and more providers have been willing to take a second look and to separate what is a claim with generative AI than what is an actual model and an actual approach that has results proven and scientifically tested.

And we need to think about medicine, like, medical decision making has been always based on evidence or epidemiological evidence. And epidemiological evidence, either you want it or not, is mostly based on models that are classic models. And so, this idea that we’ll have personalized completely black boxed models to decide every single step, It’s just not something that you can translate into day-to-day practice.

And so we need to really understand how it fits the workflow. And we are very well positioned in doing so, rather than just a point of care solution that does a specific stratification for a specific disease, or an incremental approach in terms of note taking, or some other operational admin strategy.

Mariana: But I mean, the bigger you get on the market, of course, the more competition you’ll see when it comes to, to solutions. But I think there are some key characteristics where, where Uphill actually has a lot of advantage, whether it comes with the interoperability, with the clinical knowledge that we do provide and it saves like two to three years to health teams on care mapping, whether it’s regarding certifications when it comes to medical device compliance.

Because. With AI or not, as I said before, there are some sort of certifications and regulations that you must comply with to be able to, for example, automate clinical tasks. So this is, these are some of the characteristics we see that we are a step further.

Eduardo: I think it’s all about execution, right? It’s all about having a clear success story out of the claim that you’re, that you’re making.

And what we’ve seen with our customers that they are exposed to many solutions that we have this AI model that does this specific task but then you have to implement it. You have to train the professionals you have to include in the workflow you have to certify it from a security standpoint from a medical device standpoint and so all of this burden is being shifted towards the provider and so they are leaning more towards solutions that are a wall to wall rather than just this point solutions where they have to manage everything.

Ayush: So what are the next trends that each of you are excited about in patient care? So, you know, what is something that makes you feel excited that, okay, wow, this is going to.

Eduardo: Care integration and self effectiveness for the patient and efficacy for the patient. So I think in, throughout the world we’re seeing a movement towards care integration and some countries have been able to execute it and by care integration I mean having a unified journey that is uninterrupted throughout the continuum of care.

Some countries have been integrating vertically hospitals and primary care centers, some others have been integrating them horizontally but what we’ve seen is that the benefits of having. The right patient at the right level of care in terms of sustainability are huge. We’re having trivial stuff being treated in high complexity hospitals with very much most differentiated resources and personnel there is in the world.

And this is completely unbearable. And so this idea that we can place, we can have a radar that places the patient at the right time in the right level of care is essential, for us to optimize the sustainability of healthcare. AI is going to have a huge role in this in terms of risk stratification.

Clinical passwords are going to have a major role in this in terms of connecting all the touchpoints across different providers, and I’m super excited with it. We’ve seen, for example, in the U.S., the multiplication of virtual care applications and solutions, and that itself is just fragmanting more and more and more care that is provided.

It’s increasing access, but at the same time it’s, it’s fragmenting the healthcare system. So we do also need to have the glue for all those applications in terms of the single clinical pathway. It can be shared by the patient, known by the patient, accessible with the tools that physicians already have and that is super important and also this will allow the patient itself to have more power and to be more empowered to execute its own journey.

Mariana: Yeah. I think from the, from the patient’s perspective, as younger generations get older, I think they’re more capable of also distinguishing what are the interactions, hybridization of care, digitalization, etc, that actually contribute for their progress on their care journey and what’s not really that relevant and not take control.

Yes. That doesn’t really provide a resolution to their need for care. So they, I think we’re being more and more capable of distinguishing when to be engaged.

Ayush: And one last question for both of you. Like what is something that surprised you in this journey? By surprise, I mean that something that you assumed that this is how it works or this is how people think or whatever.
And then, you know during your own learnings, you figured that, okay, I was wrong.

Eduardo: I think one thing or probably the thing that surprised me the most was, in healthcare IT, doctors are the enemies. They’re always the enemies, like, they’re seen as the blockers, they’re seen as the ones who can pause an enormous project of digitization.

But what I’ve come to learn is that actually that’s not true. Doctors, when they are Doctors are protecting their patients from useless stuff. Yeah, cut the noise. Yeah. And once they are convinced, you know you have a good product. And they will defend that product through the last minute. And so that’s one thing.

And I’m a doctor and I always seen my colleagues and I as, as some sort of blockers to, to new technologies. And in the end I understood why it’s just neat fruit to cut all the noise and just let pass what’s really useful for the patient.

Mariana: Yeah. And also when it comes to project implementations, we know the doctor just don’t really have much time, not even for their patients rather for technological implementations, they wouldn’t even have less time, so it’s important to shift this weight of project implementation, these clinical processes, all the information we should be able to provide them, like, ready in the beginning of the project and not take more of their time when it comes to it.

Eduardo: We’ve seen with our customers projects that have been going on for years and haven’t reached a go live. It’s, very strenuous for them. And and I think that it’s I think in the end when the product is good they will be the champions moving that product towards the hospital from simple implementations to multiple implementations.

Ayush: Okay. That’s fantastic. Thank you so much. Mariana and Eduardo for these thoughts and I believe that that goes a long way in helping our audience learn about building a product around patient experience and better delivery of care.

Eduardo & Mariana: Okay. Thank you so much. Thank you so much.

In this video of Healthtech With Purpose, Eduardo Freire Rodrigues, MD, CEO & Co-founder of UpHill Health, and Mariana Bandeira Senior Account Executive at UpHill discuss how UpHill’s care orchestration platform helps hospitals improve patient outcomes and clinician efficiency.

Here’s what you’ll learn:
✅ How UpHill automates clinical pathways to increase capacity and improve patient outcomes
✅ The challenges of healthcare data integration and how UpHill addresses them
✅ The importance of clinician buy-in for successful healthcare technology adoption
✅ Why AI needs to be grounded in clinical evidence to be truly valuable
✅ The future of care integration and patient self-empowerment

UpHill’s software acts like a second layer to the electronic health record (EHR) and automates tasks that must happen within a clinical pathway. For example, UpHill can automatically prescribe follow-up exams and tests for diabetic patients. This frees up clinicians’ time so they can focus on more complex patients.

UpHill is finding success because it is a full-fledged solution that ticks all the boxes. It is interoperable, has safety safeguards, and complies with standards. AI is another area of interest. Both Mariana and Eduardo are excited about the trend toward care integration and self-effectiveness for patients.