Personalisation, Be Your Own Mouse
Its time to take a brief look at next vertical - health care. What will new technologies and digitalisation mean in that area. What possibilities there are for decentralisation in health and how could that work out.
Personalisation of Health
TL;DR Health provision is becoming available 24*7, enabled by the trend for smaller, cheaper and non-invasive measuring methods combined with automated analysis. This real time feed about how medical interventions really work will be like a continuous, nationwide lab trial on effectiveness of all medical work on real-time. Data also enables patients to do it for themselves.
This set of post on health covers following main areas:
Personalisation of medicine
Data in Health
Decentralisation possibilities
Many other topics could have been discussed as the health sector is under rapid development with topics like vaccines against cancer, longevity research and aids like exoskeletons. But these are likely to evolve rapidly now, making any post outdated in a few years. Topics like personalisation, value of data and decentralisation are more structural and hence unlikely to be as volatile.
Industry structure
Health is highly regulated industry with long and expensive certification processes for approvals before new medical devices or medicines can be accepted into general use. Because of risks involved, the health organisations are conservative and usually large entities that are accustomed to purchase products and services from existing large established suppliers. Market is difficult to penetrate for small innovative firms or they have to have good funding with long-term view.
This is why in general innovations happen on consumer side and especially on customer devices. One of the big issues in health now and in future is to find fun and motivating ways for people to choose slightly healthier paths in life. This is an area where rapid iteration with the help of cheap and wearable devices is of great help.
There still is quite a bit of research and development for the official health care side as market is big, but for small teams the most viable path is to be bought up by an established player and disappear into the behemoth’s product portfolio.
We may also be in a state where most diseases are life-style related and we have a high variety of effective molecules to treat most common diseases. Almost all of them out of license even. The remaining few like cancers, Parkinsons, Alzheimers et. al. have proven very hard to find good medications indicating that new approaches are needed. (and indeed some may be emerging like mRNA vaccines against cancer, ultrasound for Alzheimers and increased understanding for causes of Parkinsons).
Personalisation Trend
The accepted general practice in treatment today is based on large scale trials and available medical records. Exact formulas are known on dosage based on age, weight, and special topics like liver condition etc. Variations in personal metabolism, life styles however mean that they are not totally optimal. Even bigger issue is that tablets and other dosage formats are mass produced in just a few sizes.
Same is true of general recommendations to prevent diseases. By default, such guidelines need to be simplistic: when your blood pressure or BMI is over this set limit (given age, height etc.), it’s a bad thing. The recommendation can be too high for some and too low for others causing long term harm. Everyone would benefit from a more accurate non-linear prediction model tailored exactly to them.
The right level of medication is sought together with the doctor based on available dosage sizes with patient discussion and occasional lab tests but these are infrequent and not everyone complains about side effects.
One trend to overcome these limitations is 2D or 3D printed medicine. It allows creating medicine where dosage is personalised exactly for each patient. In principle it will allow to even create small variations in dosage between days based on the activities in calendar and previous observations. Similarly seasons with large differences in temperature, light conditions may cause need for small changes in dosage. 2D printing creates oral dosage formats that dissolve in the mouth. The medicine can be on the QR code containing patient name and instructions after taking the medicine. Printing locally to users and personalized data on dosages prevents falsifying medicines, which is a global problem.
Several active ingredients can be printed to a single dosage instead of many tablets with personalised size, shape, colour and perhaps tastes like strawberry, raspberry etc.
Printed medicine to exact dosage combined with availability of cheap sensors and ability to upload this information in near real time to cloud enables care at a totally different level of accuracy.
Long term observations joined with other health data reveals what type of medicines, dosages and interventions work well with different types of patients. Treatments become personalised.
Near real time data closes the feedback loop for the first time. Medical profession can start to learn in near real time.
Non-Invasive, Small and Cheap Devices that You May Already Have
This personalisation trend is supported by an avalanche of data coming from a multitude of devices. It should be fairly well-known what type of devices are and can be used to track our health data.
The camera we carry with ourselves is already a more powerful diagnostic tool than what most people realise. For human doctors seeing the patient is an invaluable tool in diagnostics. How the patient looks on overall, how they walk, speak, how the face looks like give a lot of clues to the trained doctor. Certain types of coughs are indicative of heart problems, others of lung inflammations.
From an image or video feed a machine learning model or a real doctor can detect huge amount of information and diagnose for example visible traumas based on swellings or distortions, observe skin colour and bruises, rash, cuts in skin, eye movements, pupil reactions in eye, facial expressions in general, speech, breathing, images or tongue or mouth in general. The patient may be instructed to press parts of their body like abdomen and tell what they feel, follow limb movements, check arm-leg coordination and balance when patient is sitting, standing or walking.
With additional cheap devices connected to smart phones it becomes possible to take more detailed images. For example, to take pictures of ear drum and other parts of ear (otoscope) or retinal base and send them to a doctor or automated analysis program. Other add-on devices can collect sounds from heart, lung, bowel and artery or taking more detailed images of skin (dermatoscope).
Wearables like smart watches, head sets and rings being the most common other categories , but also increasing diversity with smart scales, indoor air quality monitors etc.
With these inexpensive tools it is possible to monitor both vitals and ambient environment like sleep, physical activity, stress, recovery, weight, body temperature, heart (blood pressure, pulse, electrocardiogram), blood oxidation level, air temperature, noise etc.
And final device slowly becoming more common is the speaker with speech recognition capable virtual assistant. This assistant can from the voice detect lung inflammations and characterise them, some heart conditions and possibly more diseases. Sound can also be used to detect someone falling, starting fires, bigger accidents outside of home and in general be a contact point for asking health related questions.
All these capabilities come naturally with privacy problems.
On top completely new types of non-invasive technologies are being developed to measure the function of the body from air we breathe out, sweat, spit or tears. For example, exhaled air can be used to detect certain types of cancers and sweat to measure glucose, lactate and cortisol levels.
Printed electronics is one of the technologies used to enable this. You dip a slip in your saliva and get results within a few minutes. For example, the Promiless alcohol test works like this and costs about 3 euros retail in Finland.
Ingestible pills that power from the acidity in the stomach and report measurements. These can simply be used to monitor whether the patient is taking their medicine (or is accidentally taking too much). Pills can also detect the movement of patients. Patients with memory problems may wake up and move at odd hours of the day and the movement data can reveal the progression.
Ultrasound scanning is a diagnostic tool used to examine many different parts of the body including the liver, pelvic organs, kidneys and gallbladder. It’s also very useful in identifying musculoskeletal injuries in joints such as the shoulder, knee and ankle and it can be used to examine blood flow and to check for any thin or blocked blood vessels. For more: https://www.butterflynetwork.com/
Consumer devices also mean that many people have more and better data at disposal than average doctors and have followed their personal vital signs over years and know in detail how their health is developing and specific responses to various factors like alcohol, stress, weight gain or loss and various diets etc.
The data and above-mentioned devices are helping movements like biohacking where individuals use the latest medical research and test it on themselves to see how their own individual body reacts to nutritional or other environmental changes. And then they share the results with peers.
Be Your Own Mouse
The ease and low-cost of measuring own response to various stimuluses has created the biohacking movement, where people try the latest findings of science on themselves and adjust diet and habits to optimise own health and minimise ill effects of natural reactions like jet lag etc.
This will in future be combined with another upcoming capability - ability to synthesise almost any molecule at home or nearby. We’ll treat this in more length in future post
New laboratory-scale solutions for manufacturing chemicals are available that allow to synthesise molecules in small-scale. Molecules that used to be manufactured in huge chemical factories can also be made in micro-reactors using flow chemistry.
This capability to make medicine allows people to do various experiments at home. Already now some patient associations have found ways to treat or slow down the progress of a few rare diseases. These experiments may become more common when seriously ill people are willing to try new treatments on themselves when everything else has failed.
Doctors cannot do them due to risks and moral considerations but no one can prevent patients from trying them out on their own. Patient associations (new or existing) could start organising this type of activities. New treatments will emerge.
The fact that people can on their own make diagnosis and themselves or via patient associations discover new treatments means that medicine will be increasingly crowd-sourced and decentralised.
Risks about – young people not understanding the fragility of life experimenting guided by people who lack proper medical education is one scenario that comes mind. Still, it is medicine by patients for patients for new directions and diseases where some of the patients have deep medical experience themselves. Patients can become their own mice.
In the next post we’ll take a look at what type of data affects health and where it is generated.