The Robomed Healthcare Blockchain: Revolution or Delusion?
Blockchain is hailed as a revolutionary technology that has the potential to disrupt every type of business. Naturally, people have started thinking of ways to use it in healthcare.
Healthcare suffers from lack of transparency. It is a complex domain and not just because of complicated, money-making business models. The complexity is deeply rooted in the nature of care provision, in its science, business, and social dimensions alike.
Blockchains have the potential to solve many of these problems by providing a shared trusted network and smart contracts that are transparent, standardized, and trustworthy.
For example, it is hard to share medical data across organisations, because (a) they cannot easily interoperate because of different practices and technologies, (b) they cannot easily agree or build trust or a viable business model for sharing, and © the data is sensitive and doesn’t (or, well, shouldn’t) belong to them in the first place.
Blockchains have the potential to solve many of these problems by providing a shared trusted network and smart contracts that are transparent, standardized, and trustworthy. Of course, there is some healthy scepticism, but we’ll get to that some other time.
Russian Robomed is a healthcare blockchain startup that claims it will use blockchains in a way that is very close to my own heart and work: care pathways. They have some cool sounding ideas about linking patients to providers at a worldwide scale and moving the healthcare business model to a value-based one. Their ICO launched late last year and they seem to have completed Stage 1.
I am a huge fan of crowdfunding, and ICOs have really revolutionised how both blockchain companies can raise money and how investors can make returns from good projects. As with every hype and many crowdfunding platforms, some ICOs prove to be scams. So what’s the case with Robomed? Is it an opportunity or a scam? A revolution or a delusion?
Robomed talk about linking patients to providers through clinical guidelines encoded in smart contracts. These provide value-based terms to the game: if you get the value you need out of the pathway, the provider gets paid, otherwise you get your money back. This can bring new levels of transparency and trust to the game.
They claim to have 20 clinics on board 1.7M clients (patients?) and 2900 clinical guidelines already digitised.
The Robomed ICO went well, and they are considered one of the biggest success stories in healthcare blockchains in 2017.
They also have Тhomas W. Feeley from the Harvard Business School and University of Texas on board as an advisor. He has interesting work in healthcare business models and is an advocate of value-based payments, i.e. paying for outcomes rather than procedures, particularly in cancer care.
In my effort to chase the evidence of The Good, I have uncovered a few gaps behind the bubble of business marketing, ICO hype, and standard startup pitching. As much as I really want Robomed to be as real as it sounds, as I honestly feel there is an opportunity in this space, I can’t help but pinpoint the holes in their pitch.
The clients and clinics that are signed up are all within Russia. There is no evidence their solutions scale to any other country. There is also no evidence they have used blockchains in any effective way. Their figures seem to be based on their existing Electronic Healthcare Record (EHR) and care management products that their Russian clients are using. They claim to already be using blockchains, but they never demonstrate any of the added value in practice.
We don’t get to see what these guidelines are about. What kind of diseases are they dealing with? What level of abstraction is the digitisation?
This is more evident from their claims about digitising 2900 clinical guidelines. None of them are available online. We don’t get to see what these guidelines are about. What kind of diseases are they dealing with? What level of abstraction is the digitisation?
Are we talking about clinical guidelines in terms of medical decision making, business processes (patient pathways), medical processes (care pathways) or what? The only thing we know is that “ the smart contract lays out milestones for diagnosis and treatment, with various metrics and checkpoints along the way.” What metrics and value-based outcomes have they modelled? Where is an example of a case where the provider gets paid and an example where they don’t?
“In case that the medical service provider does not want to work in the Robomed EHR, it can use any other EHR system and connect to the Robomed Network using the dedicated API under condition of the observation of the communication protocols according to the HL7 standard (or FHIR).”
We don’t even get to see their EHR system and how it currently works. The only tutorial I could find is a YouTube video in Russian. Their Github only contains code for the ICO (which noone cares about and is exposed on the Ethereum blockchain anyway).
Their FAQ is another shot in the foot. We’ll talk about digitising care pathways in The Ugly later. In the question “ what if a hospital doesn’t want to install your EHR?” they give this answer:
So “ they’ll just use an API with HL7 or FHIR and connect their EHR to ours”. In reality this is far more complicated. There are entire multi-year efforts, both in academia and industry, put into interoperation and standardisation of EHRs.
I saw the difficulties of medical data integration first-hand in a recent collaboration with eDRIS, with complex ontology matching techniques being put to practice in order to have any chance of success. Sure HL7 and FHIR are a huge step towards improving this situation, but this is a long-standing problem in healthcare IT and will not be solved overnight.
Here’s their answer to another question about unusual results, a.k.a. “what if something goes wrong?”: “Although 90% of clinical cases are routine, such a thing can rarely happen. In that situation, this patient’s EMR with all his data except identity will become a subject to investigation by Robomed Network medical community with top priority.”
How do you bring safety and security into your contracts? What about insurance or contingency planning?
So if someone dies because of an unforeseen outcome in a smart contract (that is the real risk you are working with in healthcare after all), the medical community will investigate what happened with “top priority”? What does that even mean? How do you bring safety and security into your contracts? What about insurance or contingency planning? How can any patient read this and be confident of the level of care they will receive in Robomed? “Oh you’ll probably be in the 90% of routine cases so nothing to worry about.”
The Bad is bad, but not that bad. These things are details in the grand scheme of things and they can (probably) be ironed out through practice, experience, and scrutiny, provided the principles and goals are in the right place. There is, however, one fundamental thing that I consider the biggest roadblock in Robomed’s mission: digitising pathways.
Anybody who has worked on digitising clinical guidelines, particularly as a way to standardize them across different providers, knows how complex, difficult, and time consuming it is.
Anybody who has worked on digitising clinical guidelines, particularly as a way to standardize them across different providers, knows how complex, difficult, and time consuming it is. Look at OpenClinical, for example, and the hurdles they have been trying to overcome for years.
When standardising across providers, people focus on clinical guidelines, i.e. the decision making process. This helps clinical staff make better decisions, perform complex procedures more easily, and keep up with changing care standards. NICE in the UK and SIGN in Scotland are good exemplars of this type of modelling activity.
Implementing care pathways for a given clinical guideline, i.e. the exact steps and procedures that need to be followed from beginning to end, is a whole other story. It depends on the resources (including human, medical, infrastructure, etc.), capabilities, scope/coverage, team dynamics, etc.
This is also compounded by the fact that each patient is different, with a different set of conditions, different needs, and different variances in their outcomes. This makes care pathway modelling extremely complex and hard to scale, and standard BPM practices are much harder to implement in this domain.
A typical example is one of the major challenges of modern medicine, both in terms of medical science and technological support:comorbidity (or generally when a patient has more than one conditions at the same time, which is very often the case). Pathways that reflect the best practices for the two conditions individually may cause harm if applied simultaneously, and instead alternative methods may be required. A trivial example is when a patient has an allergy to the drug that the clinical guidelines otherwise suggest as the best option.
Same goes for outcomes. It is hard to standardise outcome measurements for every single patient-provider pair. It is like building a Business Process Model and asking every single company in a sector to use the same model. It is also hard to determine quantifiable measures from the medical perspective. Feeley, Robomed’s respected advisor, partly admits this in his own recent article: “In some respects, it is not surprising that identifying and measuring meaningful outcomes in oncology is difficult.”
The output of the International Consortium for Health Outcomes Measurement (ICHOM), which Robomed claim to build upon, is a valiant and very important effort. ICHOM urges providers “to start measuring these outcomes to better understand how to improve the lives of their patients”. Their measures are supposed to be used to drive change in healthcare provision and not to quantify the monetary value of provided services!
ICHOM measures are supposed to be used to drive change and improvement in healthcare provision and not to quantify the monetary value of provided services.
Finally, maintenance of care pathways is tedious as well, especially with the enormous amount of new medical advances, procedures, drugs, technologies being constantly developed. Robomed claim that “ the active involvement of the professional medical community will be motivated by RBM payments for each vote, and the competence of holders of RBM Tokens admitted to professional medical voting will be checked throughout the authorization process” and believe that the “ combined knowledge of professional medical community is there to help”. That sounds naive.
People have supported initiatives like OpenClinical to the extent that they have because, as its name suggests, it is open! You really expect the medical community to undertake such an enormous effort and responsibility simply to make more money in the form of digital, volatile, and partially controlled RBM tokens? And all this while contributing to a proprietary effort in a company that, to date, hasn’t published a single of their 2900 digitised clinical guidelines (as far as I can tell anyway)? You literally expect to crowdsource care pathway modelling and maintenance; be like the Uber of healthcare knowledge!
You literally expect to crowdsource care pathway modelling and maintenance like the Uber of healthcare knowledge! Monetary incentives promote quantity rather than quality.
And what about validation? Who decides what is best practice internationally and how do you get everyone to agree? Monetary incentives promote quantity rather than quality.
To those outside, the Robomed solution appears as if out of a fairytale; the heaven of global, transparent, trustworthy care provision. Even with little experience in the healthcare IT industry, you can already see through some of the hype arguments. Most of these technologies and ideas are not new anyway. There are real reasons they have not scaled up and that adoption and evolution in healthcare is so very slow. These reasons will, sadly, not go away overnight just because someone mentioned healthcare, blockchain, and ICO in the same sentence.
The vision is important, however, and I think Robomed got that part right. This is indeed a difficult, but worthy goal to strive for, and I hope more initiatives follow in these steps, to at the very least raise awareness and expectations from the public.
I really hope Robomed succeeds in its mission. It probably will up to a point and within its capabilities to convince providers (and patients) to join its cause and adhere to its smart contract rules. The challenges are real and should not be underestimated, so let’s not celebrate just yet.
Originally published at https://www.linkedin.com on January 22, 2018.