Big data, with the associated technologies of smartphones, wearables and artificial intelligence are disrupting healthcare in the drive to digital health, but progress is slow.
Healthcare is undergoing a major digital disruption, which affects all the stakeholders, from pharmaceutical companies and medical professionals to the patients themselves. Emerging technologies have made it possible for the first time to connect all the dots in a patient's health management. However, there is some resistance to this change, which is evidenced by multiple names for this phenomenon, such as Digital Health, Health 2.0, Medicine 2.0, mHealth and Connected Health. Although all of these schools of thoughts have differing definitions, the overriding objective is simple:
to transform disease-centred healthcare to patient-centred healthcare
The current paradigm of healthcare consists of a single patient surrounded by many data silos, belonging to every provider involved in managing his health. This data is disparate, ranging from doctors' handwritten notes to MRI and Doppler scans and physical indicators, such as bone or tissue biopsies. In order to have a holistic view of the patient, all this data must be aggregated to give a single patient view. This is where the initial need for Big Data arises, to cope with the variety of data. Currently the volumes of one patient's data is small, according to Doug Anderson of HealthCatalyst, there is no need for Big Data databases to store patient data, because the volume and speed of retrieval (velocity) do not warrant it, but this too will change, as will the source of the data.
Monitoring the Patient on a Continuous Basis
For any patient with a chronic or life-threatening illness, only a small proportion of their time is spent with their doctor or doctors. During that time, they will discuss their condition with the doctor, compressing all the time since their last appointment into half-an-hour. This is where technology can step in; wearables, such as a smartwatch, can monitor a number of key indicators, from heart rate to insomnia, on a minute-by-minute basis, giving a true indication of the patient's health, and transmit it back to be stored. Other data that cannot be picked up by a device can be captured by the patient via a mobile application. Monitoring of medication use can be recorded and transmitted by smart pills and smart pill containers. This is especially useful where very addictive or dangerous drugs have been prescribed, or where patient behavior such as neglecting to take medication or administration of painkillers is being monitored.
All this data is invaluable in improving patient care, but will radically increase the volumes of data that needs to be retained. During DWHS16 (the Digital Health and Wellness Summit) held in Barcelona during February 2016, Michael Nova, CIO of Pathway Genomics predicted that:
"By 2020, global healthcare data will double every three days and to truly personalise medicine, this healthcare data needs to translate into accurate and actionable recommendations for both patients and physician".
Location-Based Analysis of Health Data
While the focus of gathering this data is to improve the treatment of the patient on a personal basis – where this data can be aggregated and mined – it can contribute to our knowledge on a global, regional and demographic basis. Artificial intelligence techniques can be used to identify past and present patterns and to predict future scenarios. Future threats of pandemics and the emergence of new health risks probably cannot be avoided in the near future, but their management can be improved. Three examples that come to mind are the outbreaks of the Ebola virus in West Africa, the drastic modification in behaviour of the Zika virus and the evolution of TB into multi-drug-resistant (MDR) TB and extreme-drug-resistant (XDR) TB.
Disease outbreaks are frequently initiated by extreme weather conditions. Right now, people in Somalia are coping with cholera, one of the consequences of the extreme drought being experienced in the Horn of Africa. Very cold and hot weather is a threat to the very young and the elderly; this is not a phenomenon restricted to developing countries, the first world has experienced these problems too. A connected health system will be proactive in counteracting these threats, but relies on comprehensive and recent data to find the patterns.
Using Data Science to advance Scientific Data
The advances in genomics mean that indications of a person's potential health problems can be predicted well in advance. Many health issues are genetic in origin. Pooling population data from genetic aberrations to illnesses, treatments and reactions to treatments will enable doctors to select the best therapy for the individual, based on his or her genetic inheritance. All that is required is that the huge base of data required for this is consolidated and shared, but this is taking far too long for people who have been given only a few years of life expectancy.
Patients are Doing it for Themselves
Imagine finding out, at 29 years old, that you have amyotrophic lateral sclerosis (ALS). You may not recognise its scientific name, as it is most commonly called Lou Gehrig's disease, named after the unfortunate baseball player who suffered from it. In 1998, Stephen Heywood was diagnosed with ALS. He was fortunate in that his brothers stepped up to the plate to do everything they could to find a cure, from drug regimes to stem cell transplants. Stephen died in 2006, but his legacy remains. Jamie and Ben Heywood and friend Jeff Cole set up a website devoted to ALS, where they loaded all of Stephen's patient data and welcomed other ALS sufferers to contribute via the website "PatientsLikeMe". This then expanded to include multiple sclerosis and Parkinson's Disease.
Today, the site has 500,000 members, with a variety of life-threatening illnesses. It is not a chat group where sufferers look for sympathy, it is a knowledge base of treatments and their repercussions and an inside view of what it is like to have a particular cancer or chronic disease. On enrolment, patients fill in a questionnaire appropriate to their condition, which is anonymised and added to the database. PatientsLikeMe is not a non-profit, the anonymous data is sold to interested parties to keep the business funded.
We Need a Dash of Speed
The healthcare industry seems to be dragging its feet in getting to where we need to be for a variety of reasons, mainly centred around financial gain and patient confidentiality. Keeping patient information private is a very valid problem and beyond the scope of this article. Perhaps the solution to the problem does not lie in the hands of the practitioners, but in the communities of patients and patients' friends and families. There are other sites similar to PatientsLikeMe which are valuable and impartial conveyors of information about specific diseases. While medical professionals are generally negative about such sites, other players, such as pharmaceutical companies and researchers welcome the valuable data gleaned from these sites. They may even invite patients to participate in studies of new drugs and treatments, creating a win-win for all.