Our articles

Painful injection: what mistakes medical startups make

As is well known, only 1% of startups survive, the rest die at various stages and for a million reasons. In the segment of medical startups, there are no separate well-known statistics. But the reasons why they cease to exist are subject to a completely separate logic and principles, which are quite specific. Svetlana Pervykh, a medical business consultant, co-founder of several start-ups in this field and a practising doctor, discusses the mistakes made by entrepreneurs wishing to launch a project in the MedTech sector

First, let's understand what kind of startups we consider to be medical. This definition includes those startups that are related to the human patient and the human doctor, and the products of such projects can be various drugs, instruments and devices, including those that measure such data as blood sugar, heart rate, temperature, etc. - anything that helps a doctor in assessing a person's vital signs, health and illness.

All other startups are paramedical (related to medicine but not based on scientific knowledge. - Forbes) and they are not our focus. Today - 10 "harmful" tips that will definitely kill your medical startup.


"When building a startup's financial model, be sure to take the market average as a reference"

A medical startup has parameters that no other startup has. These are professional consultations, use of expensive medical equipment, authorisation documentation and much more.

If we compare in absolute figures, a medical "endeavour" will cost an order of magnitude more, even compared to biotech. For example, the average entry price around the world ranges from $2.5 million.

And it is virtually impossible to calculate costs in advance, because there are a whole range of variables. The cost of testing always varies, as does the cost of, for example, the state duty (in Russia this ranges from 50,000 to 350,000 roubles for just one element of a future product). And these changes can take place either once a fortnight or once a month.


"Expect to start earning money from the project in a year, two years at the most."

Everything that concerns work in medicine has compulsory registration - from an alcohol wipe to bulky machines and applications in which a doctor keeps records. Such permits in Russia are issued by Roszdravnadzor.

Obtaining authorisation is a complex and lengthy process. The first stage is a preliminary check of all documentation, which may include disclosure of patent codes and a full description of the development. Then the stages of technical, toxic and clinical trials begin. These take place in special accreditation centres, the list of these centres is also formed by Roszdravnadzor.

For example, you have created a computer programme for gymnastics, and at the stage of clinical trials it will be checked to make sure that it will not cause an epilepsy attack in people suffering from this disease.

Toxicological and technical trials last an average of six months, the most difficult and unpredictable stage is the clinical trial. Simply put, human trials, which can go on average a year and a half. In oncology, on average five years.

Further on, even after passing the stage of accreditation centres, the project can be rolled out by special subordinate organisations of the regulatory body, which review the results from the accreditation centres. That is, even if you have received a positive conclusion from such a centre, it is not a fact that the regulatory body will agree with it. You may leave for revision, clarification of additional parameters, or there may be doubt that the studies conducted contain the completeness of the declared safety or effectiveness.


"Never hire a doctor on the team."

Not having a medical practitioner on the team or the model of hiring one for ad hoc consultations is one of the most common mistakes in a medical startup.

For example, you've designed a syringe that administers medicine in 0.5 seconds instead of five seconds, but didn't consider that the diameter of the girth is such that it's awkward to hold it between your index finger and thumb and push the piston at the same time. Only a practitioner would know for sure that this would lead to a series of actions that would, on the contrary, increase the insertion time to, for example, 15 seconds. Although in the technical testing phase, the model looked just fine.

Only a practicing (!) doctor can accurately evaluate the development for practical application.


"Plan to go pitch investors right after MVP."

In a medtech startup, MVP is not a commercialisation point, but a comma. A minimally viable product must pass human clinical trials or go into piloting its development.

For example, in Moscow today there are special piloting centres where you can "bring" your experiment. With their support, it is possible to go to implementation even without a registration certificate.

And only after that comes the moment to look for investment.


"Rely on marketing and PR."

Medicine is very conservative due to the degree of responsibility of healthcare providers for people's lives and health. They always perceive new things with caution, and this seriously reduces the speed of innovation in this field.

Medics trust only their own - and that is normal. That is why there should be a medical advisor in the team, who can combine both the role of a practising doctor and the role of an evangelist in the medical community. And this person should be inside the product and the team, and not brought in as an external expert - both because a serious expert should know the product from the inside, he will never risk his name on the market, and because he will protect the team from a huge number of mistakes.


"Engineers will solve any problem."

The laws of mechanics or mathematical algorithm are straightforward and logical, but biological laws are multi-component, multi-step and not fully explored. Otherwise medical discoveries would not have happened until now.

In engineering, the pattern "it worked a thousand times - it will work a thousand times for the first time" works, but in biology and medicine it is not so.

In addition to racial biological peculiarities (for example, the well-known phenomenon of the absence of the alcohol dehydrogenase enzyme in northern or Asian peoples), there are individual human reactions, allergies, side effects from the use and application of various medical devices.


"Be sure that everyone needs your product."

Again, only a practitioner can calculate the real size of the market, and he is constantly in the flow of advanced information - exhibitions, conferences, special literature, current research.

For example, in my practice there was a case when some Russian devices were left without Western consumables, which are now unavailable. There was a project that wanted to produce one of these consumables in the field of ophthalmology, the cost per unit being a few roubles.

Besides, it turned out that to create a meaningful quantity of these consumables would require an expensive mould, it was not obvious that, except in Russia, this product would not be in demand, because even the markets of the nearest neighbours from the CIS countries continue to use the former "sources".

Also, we should not forget that the tasks that help to fight socially significant diseases (diabetes, oncological and cardiovascular diseases) are of advanced interest. These are diseases that cause a large percentage of the population to become disabled or die. The rest of the projects address narrower problems and have a smaller percentage of population involvement.


"Entrust the maximum work to artificial intelligence"

Today, the possibilities of AI seem limitless. Take conducting research: let's break down why AI in a medical startup is not the main helper (for the same reasons why AI can't write a medical thesis).

Medicine uses completely different sources for analysis and research. Some of these sources are closed, some of even medical journals a human doctor will evaluate as "rubbish", even though they are popular. And, for example, it is quite difficult for a machine to understand that the publication of a study in an eminent and authoritative medical journal cannot be used because the study was conducted on six patients and the publication should be neglected.

The cost of such incorrect analytics is higher than in other industries, because here you have to roll back not just one stage, but the very beginning.


"Don't think about the ethical aspect at all."

Any research not only on human beings but also on animals requires ethical committee approval. This is also a limitation for many medical projects. For example, we must remember that work with embryos is prohibited, and work with cells requires special authorisation because of the high risk of producing a chimera.

Not so long ago, China conducted an experiment: they wanted to grow kidneys for transplantation in the body of a pig, and at first everything went well. But the human cells they implanted into the pigs to grow organs partially migrated to the brain and other organs. The experiment was urgently stopped and the animals were killed.


"Take on the most global task."

The very last item on our list, but also one of the most important. The more global the task you want to solve (for example, a cure for all types of cancer), the more money and time you will need. Because you have to constantly check the feasibility of each stage of your fantasy with the available opportunities.

At some stage it may turn out that there are simply no tools to solve one or more subtasks.

A good example is the well-known Theranos project, which wanted to determine all human diseases by a drop of blood, while today even to examine biochemistry and basic indicators separately, two tubes of blood are taken. That is, it is still possible to believe that you can find out everything about diabetes from a drop of blood, but it is a very long way from being able to make all analyses global.