Flux Insights

Leverages public and private data to enable firms to capitalise, on proprietary and non-proprietary information. Using open sourced technologies and statistical models.

Healthcare Data & Precision Medicine

Healthcare Data & Precision Medicine

The British Medical Journal (BMJ) estimated the Healthcare industry to be worth US$10tn, in 2018. It is expected to reach a value of US$11.8 tn in 2021 according to Market Watch.  An increase in growth of 18%.

The healthcare industry is comprised of, 

  • Health care services and facilities;

  • Medical devices, equipment, and hospital supplies manufacturers;

  • Medical insurance, medical services and

  • Managed care and Pharmaceuticals & Related Segments

Other industries, including nutrition, sports and fitness, telecommunications, and banking. Are broadening the reach and impact of global healthcare markets.

Healthcare Data

The past decade has seen a tsunami of data produced within and outside the healthcare apparatus. Modern inquiries into determining healthcare outcomes at an individual level and across populations. Requires significant trans-disciplinary expertise to extract valuable information, and gain actionable knowledge to deliver positive healthcare outcomes.

To achieve this necessitates the analysis of a vast array of healthcare data. Via the collection, processing and interpretation of heterogeneous and complementary data.  Transformed into quantitative and qualitative information leading to new knowledge and insights.

The size and complexity of healthcare, biomedical and social research data collected by healthcare professionals in government, academia, insurance agencies and scientist. And by consumers from their fitness and healthcare devices. Doubles every 12-14 months.       

The revolution in healthcare is related to the development of targeted healthcare treatment specifically for an individual. There are two attributes that are necessary for the successful application of machine learning in healthcare. Intelligent algorithms and rich datasets. 

However, when comparing data sets and the algorithms deployed for modelling the data.  It is the quality of the data that is critical. As it confers the greatest value to a machine learning model.

To illustrate when IBM’s Deep Blue defeated chess grandmaster Garry Kasparov, the algorithm was 14 years old, but the dataset was only six years old.  Google learnt to translate Chinese to English using a 17-year old algorithm and data they collected in 2015. On average the time for a breakthrough to be made is 18 years for algorithms, but only three years for datasets.

Types of Healthcare Data

One of the challenges start-ups encounter when selecting a market to disrupt is identifying the most relevant data.  In this blog post, some of the types of healthcare data used by organisations incumbent in the healthcare industry are covered.

Various forms of data have emerged in modern healthcare, that are categorised across three broad areas, consumer level genetic data, EMR data and Wearables data. Healthcare data is classified as follows;

Factors Impacting Health at a Community Level

At a community level the World Health Organisation (WHO), provides data that may be deployed to rank the health of communities. Across three broad areas; Policies and Programs, Health Factors, and Health Outcomes. 

As you can see in the diagram below, social and economic factors at (40%) and health behaviours at (30%) are the biggest predictors of the quality and length of life of an individual.

Hence communities where there is high unemployment, low incomes and entrenched negative health behaviours. May experience negative health outcome when compared to other communities a couple of miles away. That may present a different community health profile.

Community Health Ranking Data

Community Health Ranking Data

However, with the substantial amount of healthcare information available on the internet. And the establishment of the wellness movement.

Consumers and communities empowered by the internet are taking matters into their own hands through sharing information about how to adopt healthy lifestyles.

Including how to tackle issues around sustainable food production and the environment. Including growing their own food in urban as well as rural areas. And engaging in the delivery of credible user-generated content.

Examples of websites used by consumers to obtain information about health and wellbeing include; ParsleyHealth(Medical), PremiseHealth (Healthcare), RacetoRecover(Physical Therapy), and Intense Health (Fitness).

Wellness Healthcare Factors

There is a combination of wellness factors that interact with one another to drive healthcare outcomes at an individual level. This wellness data includes;

As consumers demand more personalised healthcare and governments adopt new measures to reduce costs associated with a one-size-fits-all approach to medicine.  Healthcare professionals, start-ups and tech firms are all converging on precision medicine. 

Medicine has migrated from a one-size-fits-all paradigm to stratified medicine in which clusters of individuals with a similar disease, risk profiles, demographics, socio-economic, clinical features, biomarkers and molecular sub-populations. Are assumed will respond in a similar manner to a given set of medical programmes.  

Precision medicine is the next step in this shift. Insofar as each patient is treated as an individual. Hence receives personalised healthcare.

Data Useful for the Practice of Precision Medicine

These troves of information become the foundation for biomedical research… We are beginning to reconstruct the relationships between genes and life and health in ways that are likely to be transformative.”
— Stanford University

Omics Data

Omics represents the study of information within an individual’s genome and the biological derivatives of the genes. Omics data is applied to precision medicine.  

There are 2,000 genetic tests available to aid in the diagnosis and therapy of more than 1,000 different diseases. Nutrigenomics is another area that omics data is applied. To aid the investigation of genome-wide influences on nutrition.   

Wellness Data

Other forms of healthcare data include data prominently associated with wearable fitness tracking. Individuals deploy fitness trackers to monitor their overall health.

However, wellness data also includes; 

  • Wireless Scales

  • Digital Pill Boxes

    - Medication Adherence  

Personal Medical Devices

- Digital Glucometers

- Personal Blood Pressure Cuffs

-  Pulse Oximeters

Pervasive Monitoring Tools

- Home Monitoring

Wellness data is primarily used for real-time monitoring of the health of users. They also enable GP’s to maintain a connection with patients between visits.

And provide individuals with a means to monitor and quantify their health.

At a more higher level, this data is used in research studies. To determine the efficiency of clinical trials. The evaluation of novel therapeutics and measuring the recovery of patients.      

Demographic Data

This type of data is used to track large scale clinical trials, population-wide trends in different types of disease and positive healthcare outcomes.

Administrative Data

This takes the form of financial data, logistical data and quality assessment data. It is normally used to analyse the costs associated with the delivery of healthcare services. It also includes but is not limited to unnecessary service, admin costs, and costs associated with the delivery of care and fraud.

One of the challenges healthcare networks are grappling with is the delivery of high-quality healthcare at a reasonable price. Listed below is the cost of healthcare around the world.

The Graphs Represent: 1) Global Market for Personalised Medicine from 2015 to 2022, by Product (in billion U.S. dollars); 2) World Health Organisation (WHO) Total health expenditure per capita in U.S. dollars. 189 countries; 3) World Health Organisation (WHO) Total health expenditure per capita in U.S. dollars. 189 countries.

Female Founders in Healthcare

Forecasting Customer Churn with Machine Learning

Forecasting Customer Churn with Machine Learning

Nutrigenomics,  & Personalised Health

Nutrigenomics, & Personalised Health