The Real-World Impact of Human Data Science
The application of data science to human data is already happening in healthcare and its impact is positive.
Let’s look at the impact of Human Data Science on three levels:
The first is at the health system level.
The second is at a disease prevention and treatment level,
And the third level focuses on delivering human health services.
At the health system level, unnecessary services can drive up costs and expose patients to stress and risks.
Many ACOs function under global payment models that were created to meet the needs of high-risk patients.
But are they really effective?
If we only look at high-risk patients, we are missing the boat on achieving the most effective use of health system resources.
Many strategies are incentivized by global payment models that focus only on reducing waste and lowering spending in high-risk patients. But what if we want to know the best way to reduce waste and spending overall?
An analysis of Part A and B Medicare claims shows that 17% of Medicare beneficiaries were considered at high-risk of incurring future costs.
However, those patients only accounted for 27% of low-value service use.
So, targeting patients with high predicted spending misses opportunities for waste reduction.
The application of Human Data Science to look at overall patient wellness actually reduced unnecessary, low-value tests and treatments for all patients.
Now let’s look at Human Data Science in the context of disease prevention, interception, and treatment.
Patient engagement and the collection and use of anonymized patient data used with advanced analytics can bring life-altering therapies to humans in ways we didn’t think were feasible.
Let’s explore two current cancer therapies: Roche’s Rozlytrek and Pfizer’s Imbrance.
In both cases, the drug development approval pipeline needed patient subgroups that are very challenging to recruit for in a traditional clinical trial.
Instead of going a traditional route that would delay the delivery of life-altering treatments, a different approach, an approach drive by Human Data Science was needed.
In both cases the companies developed clinical programs driven by real world data and evidence to identify patients for clinical trials.
The use of Real World evidence was multidimensional—used as a comparator and as a virtual control arm in the clinical trials.
Furthermore, this human data science approach using real world evidence, positively impacted the regulatory filing and approval process.
In our last example, we look at healthcare delivery as it relates to maternal care.
Gaps in longitudinal patient care, and particularly maternal care where there is a general lack of robust data collection, are missed opportunities to support patients and mitigate health risk.
The United States currently lacks uniform care in the postpartum period, when more than 60% of global maternal deaths occur.
There is a need for national data collection on postpartum maternal morbidity to address gaps in this care.
By examining human health through a human data science lens, The Arkansas Health Care Payment Improvement Initiative showed that a bundled payment model actually led to a decrease in episode spending while simultaneously improving the maternal care services.
This type of program incentivizes better care coordination and quality.
Beyond the real-time impact on care for the mother, the data collected from these programs can be analyzed to determine if quality goals are being met and their impact on cost----ultimately to continually improve care and inform future decisions.
We can see where challenges and gaps in healthcare worldwide are daunting.
But a change in approach can make a difference in outcome. This new approach known as Human Data Science is happening and offers reason to be optimistic.
The powerful forces of human ingenuity, breakthrough science, and disruptive technology that Human Data Science has unleashed promise to power future healthcare advances and improve health outcomes for individuals and populations globally.