Health Analytics

There’s a tsunami of information coming our way, the world’s digital data is expected to double every two years for the next decade.1 This data is of a new form, what we call big data, which is unstructured data from a wide variety of sources. The International Data Corporation has estimated that unstructured data accounts for more than 80 percent of currently available data. We generate terabytes of data from CT scans, MRI scans, patient genomics, patients themselves, activity levels, sleeping, diets, information about our environment, air quality, weather, pollution, traffic etc. Given advances in computing capacity and new forms of algorithms, we are now able to use these complex data structures to be able to do analytics on whole environments, not just of individual phenomenon. This holds out the possibility of moving us from a world of isolated snapshots, reacting to ill health based upon generic one size fits all models and procedures to a world where health care is pervasive and predictive and thus can be preemptive, preventive and personalized.

In a recent article Shomit Ghose illustrates this well when he writes “once upon a time, brands brought add inventory on actual print media months in advance and targeted their consumers at the population level. Big data and the Internet revolutionized the media industry through 40-millisecond-latency real-time bidding and micro-targeting at the individual level. Big data and the Internet will enable a similar transformation to real-time and micro-targeted healthcare.”2

Big Data

Big data advanced analytics is a powerful new tool at our disposal but the question is how can we apply it in a holistic fashion. Too often data analytics are applied in a very narrow focused way – analytics technology is currently being rolled out within many health systems for such applications as diagnosis processing, treatment protocol development, personalized medicine, drug development, and patient monitoring – although valuable in their own right these use-cases are in general simply extending the existing paradigm. What is really needed is to use these technologies towards enabling a new paradigm where we are able to look at the whole context or environment within which the individual finds themselves and how that affects their health outcomes.

Big data is pervasive data which can provide situational awareness within space and over time, allowing us to see the context within which health issues emerge and not just the issues themselves. The capacities of advanced analytics to cross-correlate millions of parameters can help us to move into a nonlinear paradigm for understanding health issues – no longer a simple cause and effect analysis but a much richer and dynamic complex view of people’s health which is required for dealing with the chronic, behavioral and environment-related conditions of today.

Going forward, our cell phones and wearables will be by far our biggest source of health-related data – mobile health can look at your diet, your nutrition, physical activity, your heart rate, the level of hemoglobin in your blood, body temperature, rhythm of breath, physical activity level, blood alcohol level  – providing a continuous stream of information that promises to enable the practice of continuous health care, for the first time. This data is collected non-invasively and will specifically help us understand how user behaviors affect user outcomes.

Continuous data, when supplemented with other data sources, promises to revolutionize the current methods of population health, including the monitoring of disease and disease vectors in real-time. Analytics and visualizations are key to responding to the growing threat of pandemics in an interconnected world. With more population health data it will be possible to use advanced algorithms to discover widespread, though potentially unseen, population health risks. Predictive modeling, simulations and heat maps for crisis prediction are just some examples. The combination of big data and better algorithms can help healthcare providers to find and prioritize at-risk and underserved patient pools to enhance care and reduce societal risks.

Continuous & Pervasive Health Care

A software application may now have thousands of algorithms performing analytics that replicate the guidance and direction that a physician or nurse would give you if they were with you at that specific time and location to help you understand what you should eat, your exercise behaviors, how active you should be, your sleeping patterns and how they should be managed, the medicine that you should be taking, when you should be taking them, all those sorts of behaviors that are going to really lead to better outcomes. With analytics, this can be done for the marginal cost of a few cents, while requiring less professional staff.

To meet the current challenges we need healthcare networks that are real-time; feedback has to be continuous not legacy. User data and environmental data needs to be processed by algorithms in real-time to provide the imminent feedback mechanisms for people to adapt and adjust their behavior towards those that are beneficial to sustainable health. This again requires the movement of information onto blockchain networks so that people can amass and store their own data, and make the data available on-demand in real-time for the networks in their environment that may access it to provide them with personalized health information. In a world of big data, health systems have a 360-degree view of the individual and instead of a one size fits all, micro-complex data can be amassed and analyzed to find information about people who are similar to the patient in question and thus customize solutions to their specific needs can be realized.

For the cost of only a few cents we can do complex data analytics to deliver solutions that are personalized to the individual – targeting therapies, treatments, health recommendations etc. specifically to the individual, how they live, their physiology, their particular physical environment, and how they respond. This can work to take a large amount of unnecessary cost, time and energy out of our current system by simply prescribing the right solution for the right person.

Security & Privacy

Secure distributed and anonymous storage of medical data could revolutionize the field of public health. The value gained from this data to maintain public health could be converted into health tokens used to incentivize individuals to make their health data accessible for public health research and analytics. Real-time population-wide data could be analyzed and visualized to present a real-time picture to people of current health issues or threats. Likewise with data from a multiplicity of sources and advanced analytics information about any given environment – such as a park, school, or whole city – could be amassed, analyzed and delivered back to the users as an accessible visualization in public spaces, thus displaying the “healthiness” of that particular environment, in terms of pollutants, conduciveness to physical activity etc.  

However, the use of such technology must come with appropriate usage and an awareness of its risks. For the information revolution in health to be sustainable major structural changes to the existing model will be required. The growing centralized silos of data that are honeypots for hackers is unsustainable, data needs to be moved to secure blockchain networks over time. Likewise, black box algorithms the remove decisions from patients and clinicians without explaining what is going on is an unsustainable model. Moving towards user-centric health networks means keeping users engaged and informed about what is going on, dealing with all this health data may require automated decision making by algorithms but the workings of that system should be exposed, explained, and even visualized so that the end user has at least a high-level overview of what is happening while specific technical details can be abstracted away for the sake of usability and engagement. Technocratic health systems are not a sustainable solution, the user has to be at the center, empowered and informed by a flexible network that they understand and can influence towards realizing their own health aspirations.