Update: The HIMSS Global Conference in Orlando has been canceled due to the coronavirus. Read more here.

As healthcare systems around the world come to the realisation that they need to take action in order to become sustainable and provide the services that citizens now expect, many have turned to the use of digital, data and technology.

But it’s no secret that it's tough to crack the challenges of putting that transformation from theory to practice at scale.

Ahead of HIMSS20, MobiHealthNews caught up with Professor Ran Balicer [pictured on the right], chief innovation officer of Clalit Health Services and founding director of the Clalit Research Institute in Israel, to talk about their approach.

This interview has been edited for length and clarity.

MobiHealthNews: First of all, could you give our readers an overview of your work at Clalit Health Services and the Clalit Research Institute, of which you are a founding director?

Balicer: Clalit is a unique payer-provider system that is responsible for the care of half of the Israeli population. What's unique about it, beyond its reach on a population level, is that, generally speaking, people are members of Clalit from cradle to grave. And we provide all types of care under one roof, primary care, specialty care, hospital care, as well as being both the payer and the provider. This gives us a unique opportunity to give integrated care and the unique incentive to care for the patients and keep them healthy rather than to provide them, sell them services.

This is point number one. The other one is that our organisation has been digital in terms of electronic health records and central data warehouses for over 20 years now. The availability of a central integrated data repository and all of our patients and all various types of care is something that is, I think, quite impactful and holds great promise in order to improve and change patient care and move from the paradigm of reactive care towards predictive, proactive and preventive care, which is what we are trying to do and have been doing at scale for the last at least a decade and a half.

Now, one of the things we are really trying to do is to put predictive medicine at scale into the practice of our general practitioners, as well as our hospital. To give an example to understand what predictive medicine means for us, if we know that within our population there is a segment of patients that look well and feel well, but the entirety of their laboratory tests and comorbidities and several other features hint that they are about to be suffering from an illness or an event, we are trying to identify them to bring this information directly to the physician or directly to the patient and change the course of their care so that event or disease would be prevented.

And if we take one example, for instance, if we talk about chronic kidney disease, this is a disease that if you identified through symptoms, you’re too late, so you cannot change too much the course of events whilst the patient is already symptomatic and suffering from pain and discomfort. You need to come earlier, and this is what we do we, we address the needs of these patients that are destined to become renal failure patients five years before this happens and, at that point in time, we are able to prevent this from happening to them at a much lower cost, much lower impact on their lives, and much less adverse events associated with the care. That's what we provide.

We’ve been able to halt the increase in new dialysis patients in the last few years because of this approach that has been ongoing for tens of thousands of people per year over the last eight years or so. So this is one example. But I could go on.

MobiHealthNews: A couple of questions from what you said. Obviously, Israel's push for the use of digital in healthcare is well known. As you mentioned, you've been paperless for around two decades. And we also know that you've got a vibrant startup scene. I think there was a report that said there were over 500 digital health companies in the country, and that was a couple of years ago. What do you think are some key factors that have set the foundation for this to happen?

Balicer: Sure, sure thing, so there are several [inaudible] that make this possible. One, we call them the four Is, and I will refer you to a Lancet paper [which can be accessed here] that I've written that has those four Is, of the Israeli ecosystem that make it available for digital health innovation. One being the integrated data, the next one is the disruption, where the institutions, the hospitals, and the HMOs [health maintenance organisations] are really interested in trying new things, it’s a whole culture in Israel that people always like to change the status quo and challenge it. That mentality is also true for the institutions themselves, so they are ready. Then there is the entire innovation ecosystem that supports new startups with both funding and other types of support. And the fourth one is [incentives aligned with prevention and care improvement].

MobiHealthNews: You mention this in your TEDx talk from a couple of years ago, where you also talk about a collaboration through which you built a model to predict who will suffer from an osteoporotic bone fracture, and found that the algorithm indicated that eye diseases were at fault. And you said that you initially thought that was a glitch, before realising that vision issues led to patients’ falls. I thought this was a strong example of how the use of data and technology could help improve healthcare. However, we know that wide-scale deployment of this type of transformation is a significant challenge for systems around the world. Why is that, and how have you approached this?

Balicer: Good. Before we go into that, just to mention that this bone fracture work has now been augmented, not only with what you have described and not only with our BMJ paper from three years ago, in which, for the first time, we have compared three different predictive models on the same common grounds, which is really critical to understand which one of them in the past work and works better.

What's happened this year is that we have been able in January to publish in Nature Medicine our work with one of the Israeli startup companies called Zebra Medical Vision a paper that shows that through CT-driven AI biomarkers, we are able to predict osteoporosis better than the best of breed models that currently exist that are based on classic data. So from occasional CTs, one can find biomarkers that are very good in predicting osteoporosis, so these have improved our model beyond the best of practice. So this is something that I think you'll find interesting, that Nature publication from January.

Now, back to your question, there are several impediments to the large-scale implementation of digital health. And they require some attention, one of them being that some of these innovations are not driven by the true needs of the medical community, but rather from technology that looks for an implementation, and therefore, it's not really solving a critical problem.

I think that that is one point to bear in mind. So it's very important that the innovators work with clinicians early on in order to make sure that they are solving a problem. The second thing is that when you're implementing, you need to take into account the problem or the challenge of integrating your solution with the workflows of the clinical staff during their normal day of work.

And if you expect that your gadget or solution would be integrated when it increases the workload of the clinician, then you might just as well forget it, this will not fly. You need to have a solution that doesn't only bring the technology, but also the workflow engineering that allows it to be implemented in the normal working routines of physicians, nurses and allied staff. So this can actually work.

I think the third point that I would mention why this is not implemented at scale is that in many cases people are doing the attempts, various pilots, and those pilots are by design so unrealistic in what they put in place that you know that once this will go by beyond the pilots and into scale, this will never hold.

So if one wants to succeed, they need to have the pilot process not as a separate standalone pilot, but as the first installation of something that can actually work on large scale, and plan it from point A, and then there's a chance that this will work.

MobiHealthNews: When you say ‘unrealistic’, what exactly do you mean by that?

Balicer: For instance, they have a lot of support staff that is unique for the pilots, like research assistant, etc., that creates an environment for the pilots that would never work when you try to scale it by 100 times or a thousand times. So the pilot might succeed, but it will never be implemented at scale.

MobiHealthNews: Yes. And we see a lot of the time pilots that work well but don’t go further than that.

Balicer: Yeah, agreed.

MobiHealthNews: If we go back to your TEDx presentation, there, you talk a lot about clinical intuition. As more and more digital tools are deployed across healthcare, how do you see the role of the clinician changing?

MobiHealthNews: Okay, so, definitely, the roles of the clinicians are changing and will continue to evolve and adapt and it will be [different for] different specialties, but some will be impacted earlier and some would be impacted later. I think the first ones to be impacted are those that are working with images, hence the radiologists, pathologists, and eye physicians, ophthalmologists.

These will be first and I would say that the key thing to remember is that doctors will not be replaced, they will be augmented, and AI will not to take the role of the physician, but it will allow them to do more precise work [with] less burden on them. And therefore, in many instances, these should be embraced, rather than feared or rejected.

And we need to make sure that we understand where the limits of these new tools lie, and what is the point in the human-machine interaction where the machine can provide you with inferences that are useful, and what is the point in which you need to drive the decision back to the hands of the physician.

And again, I think that, generally speaking, AI would allow physicians to take away from them some of the mundane, repetitive tasks that they are doing and allow them to focus their scarce time and energy on things that make a bigger impact on patient care and being, for instance, having more time to speak to the patient and showing the empathy that is such a critical component of healthcare, as it used to be and as it hopefully will return to be also thanks to some of these new advances.

MobiHealthNews: Now, in addition to your current role at Clalit, you are also an advisor for the Israeli Ministry of Health on infectious disease epidemiology, health policy and emergency preparedness. At the moment, everyone is talking about the coronavirus. I saw an article this week stating that the Israeli Ministry of Health urged citizens to consider not travelling abroad if they don't need to. What is the situation in Israel like, and how have you seen technology being used to limit the spread of the outbreak and help patients?

Balicer: The point at which we are right now is that every country that does not suffer widespread infections has a strong vested interest to delay as much as possible the introduction of uncontrolled dissemination from person to person within that country. Israel has a higher likelihood of succeeding in this, compared to other countries, because of its limited points of connection with the rest of the world, one central airport, and few land pathways.

With proper public health measures, Israel has a good shot at significantly delaying the introduction and dissemination of the disease from person to person in Israel. And this is the key thing that the system is now focusing on with, I think, major success, not without the cost, because Israel has been very proactive in trying to delay or very proactive in its recommendations on restrictions [to] travel and the isolation of patients that have returned [from travelling].

That being said, I think that we will see in the coming weeks, like in many other countries, a change of strategy from containment to mitigation, at which point the key aim would be to, I would say, create slower dissemination patterns of what is already an inevitable dissemination process. And this will be done mainly by social distancing, and this will also include healthcare provision.

And so I think this will be a great time for new innovation to put into practice in terms of digital health and telemedicine, and we have several such technologies that [are at] the forefront of that. And I think that we will see a lot of implementation happening in the coming future.

MobiHealthNews: Lastly, HIMSS20 is quickly approaching, and you will be giving a presentation there on the work that you just touched on. What should attendees expect?

Balicer: In this session, I will be sharing a whole set of case studies for the implementation of digital health in practice and at scale. I will give examples of success stories and not as successful attempts to try to change healthcare quality, improve access and reduce waste in healthcare systems using AI and put in practice.

I will also have a fireside chat with the chief medical officer of HIMSS, Dr Charles Alessi. As part of this, we will discuss the unique challenges that we've been able to overcome and what the future might hold in terms of the introduction, of the impact of digital health on the transformation of healthcare systems worldwide.

MobiHealthNews is a HIMSS Media publication.