Using artificial intelligence to leverage healthcare data and create efficiencies
Using artificial intelligence to leverage healthcare data and create efficiencies20 May 2020 | 7min
Keeping on top of the ever-growing wealth of evidence-based medical information is not humanly possible
Artificial intelligence algorithms have enormous potential to make this information accessible and provide clinical decision support
Resistance to change within healthcare systems is a major barrier in adapting new tools or technologies, but ultimately, if the solution improves patient care and reduces healthcare costs, then success is possible
Making use of the vast amount of fragmented healthcare data has enormous potential to add value and transform many, if not all, facets of healthcare.
The executives of The Medical Algorithms Co. Ltd. sat down with us to talk about the promising use of artificial intelligence (AI) algorithms and medical analytics to improve patient care, reduce costs and enhance the patient and clinician experience.
Improving the accessibility of information for decision-making support
HT: What unmet need did you set out to overcome with your startup in the medical space?
Johannes: We offer a wide range of quantitative decision-support analytics for healthcare practitioners, so they can use evidence-based medical tools to improve the quality of care, increase their productivity and save valuable time.
John: Our primary goal was to make a comprehensive resource for clinicians that would support clinical decision-making. A second was to identify tasks that could be automated, thereby reducing the cognitive burden that providers face.
One task to achieve these goals is to identify and codify scientific knowledge and data. A second task is to combine them in meaningful ways for problem-solving. The final task is to present the right information at the right time without the clinician needing to ask for it
HT: So the idea is to take all the evidence from scientific data, which is very rapidly growing, and make it available to whomever is making the treatment decisions?
John: Yes. What we see is that healthcare professionals just can’t keep track of the large amount of information. Doctors are aware that new information is out there, but they don’t have the time to look it up – which means that vast quantities of knowledge aren’t used. It can be difficult to distinguish knowledge that is useful from clutter. The hope is that if you have enough resources, you can use algorithms to extract what is relevant and make the knowledge accessible.
Johannes: As you say, the quantity of data is increasing by a factor of more than two every year. This means that the more data is available, the more data will be analyzed, and the more data analyzed, the more algorithms will be developed. It is about taking these different algorithms and putting them on one platform so that people in the healthcare sector can use them to improve quality and compliance.
The potential of AI to leverage clinically validated healthcare data
HT: You have a collection of about 38,000 algorithms in the solution that you offer – how do you ensure the clinical validity of the algorithms that are in your system.
Johannes: In a scientific medical study, usually the team consists of researchers, practitioners, university professors, experienced scientists, and so forth. In a first step they develop the study design, perform the research and provide the results. In a second step, the study gets reviewed by a team of peers who reviews the quality and suggests improvements. In a third step, a well known reputable Journal decides whether the study is beneficial for current medical practice.
So, essentially there are three steps which evaluate the validity and usefulness of the results of the study. On top of that, our own validation is that we have a thirty-three-step approach to test whether the results of the study are exactly mirrored on our platform.
John: One question that is asked when automation in healthcare is discussed is, “Are you trying to replace doctors?” The better philosophy is how do you enhance practitioner performance. It is impossible to use algorithms to replace physicians because the physician is the ultimate judge.
When driving with your car navigation system hopefully you do not blindly follow every instruction or you will drive into a pond. What you need is for the provider to see the big picture and then offer solutions that she or he can pick and choose from to help them solve the problems that they face.
Ideally, the care will be done faster, better and cheaper than they can do it without analytic support. If you only save 5% of healthcare costs with this type of automation, then you’re saving billions.
Drivers of resistance and change
HT: What are the main barriers for startups in healthcare and how do you see the collaboration between startups and large industry players contributing to healthcare transformation?
John: Currently, we are limited by a lack of resources. Partnering with someone who has deeper pockets would allow us to take the system to the next level.
Johannes: That’s one of the reasons why we think collaboration or a strategic deal is the way forward for us, just like it is for many other small startups. Additionally, market penetration and scalability is easier and faster when you have a bigger parent.
John: One of the barriers in healthcare is the hesitancy around change if it’s not required. When knowledge is found, it may take ten years before it’s implemented. Another barrier is actually getting it into the hospital. It’s one thing to get users comfortable in using the knowledge, but it may also need to be approved by the hospital administration which can take quite some time.
HT: Would you say this resistance to change in the healthcare industry is the greatest challenge you face?
John: There are many different players in healthcare and each player resists change in a different way. Government regulations restrict what hospitals can do in terms of innovation. Hospital administrators pursue their own agendas and may not support changes that do not further these. Healthcare providers may resist change, particularly those that are embedded in their own ways. Legal liability also restricts change since providers are judged against the standard of care in the community, which changes slowly. Healthcare is rarely agile.
Cost is another barrier, but if the physicians love a tool because it is up to date and it adds value and they want to use it, they will get it.
Claus: If they’re saving money, time, and costs, and have an advantage of high quality, then there are big arguments for the doctors and the administrators for the system. The trend towards value-based care is a big driver towards automated analytic support.
HT: What do you think the answer to resisting change would be?
John: Right now, no one is required to use these new solutions or tools. If there was a requirement then people should start using them. It is also important to have the tools in a format which people can actually use. If it requires effort then people will not use them.
3 keys to success: find a paying client, get back up after failure, and look for synergistic partnerships
HT: What advice would you give other startups wanting to enter the healthcare industry?
Johannes: My advice to startups is that when you have an idea, the first thing you want to look for is a client. Look for someone who is using a solution or tool to solve their problem and see if they would pay for it. Arianna Huffington says something like, “Failure is not the opposite of success; it is a stepping stone to success”. The difference between success and failure is that you have to get up one more time than falling down.
John: Expect to fail. But get up again.
Johannes: A lot of startups come to the point where a larger company desires to integrate such external innovations; and it is often highly beneficial. This approach is internationally considered a good way to accelerate innovation and sustainability in our healthcare technology and pharma companies.
The Medical Algorithms Co., Ltd. successfully participated in Startup Creasphere, a leading digital health accelerator that strives to transform healthcare together with startups.
Claus Puhlmann, PhD is a former research scientist with Max-Planck-Institute, has extensive experience in cancer biology and immunology, and over 25 years experience in publishing scientific and medical information for professionals, content management, and patient-focused, internet-based medical information and health record systems. Claus is also author of more than 200 medical review articles for patients and an award winning photographer. His primary focus is on media structure, quality control, and online content.
Johannes Harl, PhD, MBA is the CEO of The Medical Algorithms Co. and has functional responsibility for strategy, project management, finance and legal issues. He also has extensive experience in model development and data analytics. Previously he served at JP Morgan in New York and as a professor at New York University, Leonard N. Stern School of Business. He has a PhD in Operations Research.
John Svirbely, MD was a practicing pathologist for 30 years prior to coming to work full-time at the Medical Algorithms Company. He is the author of 6 books on medical algorithms. His current interest is healthcare automation using standards-based business process models.