Big Data in healthcare – how can it help us?
Originally published in October 2016, last updated in May 2020.
Each day, a massive amount of data (so-called Big Data) is generated by many areas of health-related activity. It is continuously produced by research laboratories, healthcare providers and by the patients themselves. As a quick example, thousands of records are obtained from the medical interviews or during the diagnostic tests in every single minute.
You may now ask, how big are the numbers we’re talking about? It is assumed that by the end of 2025, we will collect a total of 175 zettabytes (ZiB) of data, which written on the DVDs, would create a stack of discs that placed one on top of the other would run about 5,500 times around the world.
Now imagine, that a simple comparison between two pieces of information included in this collection might save millions of human lives, dollars, or hours spent on research. This is the very reason why we need to start thinking seriously about using big data in healthcare right now. To make our world a safer, healthier, and more organised place to live.
What is the essence of big data?
For millennia, people have tried to organise the existing knowledge, as well as to use data analytics to support their decision-making processes. Think about the Alexandrian library – and the concept of the library itself – which gathers the world’s knowledge in one place, assigning every title (i.e., journal or article) to the predefined category.
With the invention of the Internet – followed by an email and mobile communication – people started to exchange massive amounts of digital information and upload them on the servers, which capacity has easily overgrown the capabilities of even the biggest ancient libraries.
Regrettably, the data sets were (and still are) unorganised, thus very chaotic. Moreover, the dynamic technological development of many organisations caused a huge amount of data flooded almost every industry – making it impossible to monitor.
This vast amounts of data – both structured and unstructured – are commonly described as big data. However, don’t be misled by the catchy name. It’s not the amount of data alone that is important, but the unique opportunity of using big data to solve problems related to the selected business, or even to the whole society.
The common elements for big data are summarised by the rule of five V’s, which stands for:
- Volume – the size of the data that is produced every second in the chosen sector. By definition, it is so large and complex that can no longer be downloaded or analysed using traditional data processing methods.
- Velocity – the speed at which the data is generated, analysed and reprocessed. Nowadays, thanks to the implementation of machine learning, it is usually a split of the second.
- Variety – the different types of data and sources. You can categorise i.e., white papers, images, videos, media, or even speech recordings.
- Veracity – the authenticity and credibility of the data.
- Value – the possibility of turning the big data into value.
As you can see, big data refers not only to the data itself but also to a variety of contexts in which it can be used. The key is to analyse the records in order to obtain important conclusions, which leads to making more informed decisions related to the sector. As for medical sciences, a major goal is to use healthcare data analytics to predict epidemics, treat diseases, and improve the quality of life.
What is healthcare data?
Healthcare data (also: medical data or clinical data) is the specific set of information, created in every place and situation, where a patient comes in direct contact with the healthcare.
It refers to the personal data about the physical and mental health of an individual – including the use of healthcare services, pharmacy prescriptions, insurance details, treatment plans, and the course of the therapy – disclosing information about his or her health status. It usually includes both historical and present records.
How is the clinical data used in healthcare?
In the traditional approach, the doctor – who has a certain level of domain knowledge – analyses the patients’ health records. Then, based on deduction, combined with his insights on similar cases and gained field experience, he makes a diagnosis.
To do so, he usually gets information from the individual’s medical history – the primary source of big data in healthcare. Each and every patient comes with a certain volume of their own clinical data (usually written), which describes their health issues and past conditions. The detailed history of previous symptoms helps the doctor to asses the patient’s actual health status, or spot early signs of a serious disease.
Unluckily, some patients tend to hide their symptoms, forget to check their vitals regularly or avoid further appointments on purpose – which may often weigh heavily on the progress of treatment, and the general quality of the data in healthcare. In those cases, some medics decide to support their decision-making by gathering various types of information (i.e., the meals eaten, kilometres walked or patient’s heart rate) with the applications on portable devices.
Even though this data may seem simple, it allows the doctor to ‘get to know’ their patient and their health condition better. This is a perfect substitute of regular check-ups, especially for the older people, which verifies whether the individual follows the health recommendations or not in almost real-time.
Big Data + Healthcare
The analyse of healthcare big data extends the usability of the medical data with the aspect of prediction. As a result, it causes data to be understood as a source of knowledge about the origins of a variety of diseases (even the ones not known yet).
The implementation of big data in medicine automates the process of decision-making by translating the available knowledge into the form of logical rules, establishing an expert diagnostic support system. Consequently, the health data is able to be sorted in a structured manner and examined for further relationships.
To do so, big data algorithms compare the existing data sets with the greatest precision possible, including the results of research, observation of symptoms, and other factors, theoretically unrelated to the diagnosis, i.e., the history of consumed food, medications taken, and many more variables. What’s more, each of these factors can be set as a filter (or alert) to the data sets of the predefined patients, then used to predict (or track) health status and results of conducted therapy.
As a result, a diagnosis of cancer and other severe diseases will be made faster, more effective and at a fraction of the price, being more available to the patients from the less wealthy countries.
The big data analysis potential is being successfully used by the Michael J. Fox Foundation, which equipped several Parkinson patients with a set of wearables in order to collect more data, thus accelerate the development of research on this disease. Hundreds of readings per second are being transferred to a dedicated platform, where doctors are working on developing more effective treatment basically in the real-time.
It’s worth to mention, that the results of big data analytics are not only fully accurate, complete and current but also numerous and representative enough to be able to translate them into conclusions for the entire population. It is also important for its credibility that the original source of the data is fully trackable, making it possible to identify the so-called “patient zero”.
What’s more, big data analytics is useful in the field of search for new drugs by supporting the researchers’ decision about choosing the safest substances, gathering even more data on the diseases from the DNA and cell testing or predict drug production cost. It is proven that the use of IT solutions reduces the costs of research, increase its efficiency and significantly eliminate human errors.
Safety of big data in healthcare
No data is more personal than the medical records, thus it’s crucial to secure data collecting systems from the leakages and data theft. Amongst other things, this is the reason, why complete computerisation of health centres is a difficult challenge.
Cybercriminals prefer to hack into healthcare databases since they can earn more money from the blackmail than even from the theft of credit card data. Such systems are usually poorly secured, outdated or fragile – continuing to be vulnerable to new attacks. Some healthcare organisations use software already available on the market (so-called box software), which is, unluckily, even easier to break in, because of its public accessibility.
The number of data collected is growing rapidly from day to day, which forces medical facilities to invest in proper IT infrastructure. Due to the specific needs of each one, we highly recommend opting for bespoke software development, a unique solution, fully tailored to the one’s needs. It is prepared by the team of software developers and supported 24 hours per week, in order to maximise both patient and user data safety.
Despite the dangers of introducing new technologies into healthcare, the benefits that big data can bring for the sector are much greater. The computerisation of medical data may raise concerns but – as long as security is maintained at a high level – it strongly contributes to the development of new treatment methods.
Big data in the time of Coronavirus
By gathering big data in medicine, we are able to predict epidemics. The strategy already worked in Africa, where the analysis of phone location data proved to be very valuable in tracking population movements, which helped predict the spread of the Ebola virus.
The analysis of big data helps in the fight against the spread of the epidemic. Having access to global health-related databases increases the detectability of many infectious diseases, making it possible to spot epidemic symptoms at an early stage – thus, separate an individual from society.
As a rule of thumb: the more data we gather, the more accurate will be the results of the analytics. Along with the growing information base, the algorithms will be more effective in prompting treatment methods and predicting its effectiveness based on similar treatments. Thus, the observations can be quicker classified to a specific disease, supporting the prevention and therapy of patients’ diseases.
It’s worth to mention that big data can also minimize the costs associated with late diagnosis of epidemics, help cure diseases and prevent further deaths. All due to analysing recurrence patterns of particular cases and learning on them.
How is big data affecting healthcare?
Even though the quality of care has already improved and developed to cater to our needs, it can still do better and, thanks to the use of big data in healthcare, these changes will appear faster.
Considered as the future of healthcare, personalised medicine must go a long way before it fully replaces traditional trial and error practices. It is predicted that – mainly due to the Coronavirus outbreak – the way the doctor will work with you may change in the nearest future.
Medical care wants to keep the patient away from the hospital, opting for telemedicine and trusted healthcare tools. As a result, the patient could receive a diagnosis by conducting a self-diagnosis online or Skype with a qualified specialist.
This communication will, of course, leave behind a large amount of data that can be analysed to provide valuable information on the general state of public health and patient care, upcoming disease outbreaks or the new ways of treatment. As technology advances, especially in healthcare, big data sets will grow as a domain and size, perhaps to levels that we can’t even imagine. The enormity of data necessary to process for each individual patient precludes analysis by a team of experts.
Even if we are healthy, access to an extensive and ever-growing database of information, and the state of health of the public will allow us to predict problems before they occur in real life and thus prepare remedies (therapeutic or educational) in advance.
This very data is used to combine further, more general, diagnostics about people’s health and habits, to draw a comprehensive picture of a typical Joe. All to even better adjust the devices inside the hospitals and hospitals themselves.
The bright future of data analytics in the healthcare industry
Over half of the healthcare organisations in the world are already using big data analytics. Large data sets are becoming our everyday life and it is not surprising that more and more companies want to benefit from the potential offered by the analysis of such amount of knowledge.
By working with big data we provide ourselves with a high-performance tool for advanced analytics of various medical data, which increases the scope and effectiveness of scientific research, helps in the development of new diagnostic methods and new therapies, ensures improvement of quality of care and supports and standardises research processes.
It’s worth to note that today’s data market is growing six times faster than the entire IT industry. Devices connected to the internet or various ICT systems generate massive amounts of data. Portable devices connected to central databases are a great way to collect information about the patient’s disease. An example of such an application is the special cover for asthmatics, where the use of the inhaler is being tracked through the smartphone app.
Such large amounts of data combined with advanced analytical methods allow obtaining much better insight into the patient’s behaviour. However, the clue is to analyse them carefully and get practical conclusions. It should be remembered that the actual value of the data is the accuracy of its processing and drawing conclusions from the analysis.
And for this, you need qualified specialists and a fully customised software. The task of big data analysts is to holistically adopt the analytics to the business model, organizational strategy, and the context of business operations.
The influence of big data analytics cannot be underestimated since it can play a vital role in helping us overcome hurdles that were difficult to overcome years ago. It has a visible impact on operational activities, resulting in a noticeable improvement in patient care and the acceleration of the decision-making process. And let us ask you: is there anything more important than our health?