Small data, big possibilities. Could predictive analytics become the gateway to personalized medicine?
In the health care industry, big data receives a lot of attention, promising to improve everything from readmissions to medication costs. But there’s another type of data that has equal promise. “Small Data,” a term coined by author Martin Lindstrom in the book of the same name, represents a data set that applies to one specific item – or in the case of health care, one person. Small data is occurring at the transactional level of an individual patient in real-time and is relevant, predictive and everywhere -- across the care continuum.
For more than two thousand years, dating back to Hippocrates, medicine has aspired to be personalized. But delivering that quality, personalized care has become increasing more challenging requiring the knowledge and experience to read and interpret a growing set of health data and compare that to individual patient profiles to determine the best possible treatment. This is why predictive analytics and algorithms applied to small-personalized data is so promising and is the next big thing in patient-centered care. Predictive analytics is the process of learning from historical data in order to make predictions about the future and for health care, predictive analytics will enable the best decisions to be made, allowing for care to be personalized to each individual.
Small data analytics uses statistical methods to scour and analyze tremendous amounts of medical data, then compare it with a patient’s own biometric data to predict numerous possible outcomes – medication responses, recovery time, the likely cause of disease and ultimately targeted interventions.
Using techniques such as machine learning and artificial intelligence, predictive modeling creates a unique, individualized algorithm (known as a prediction profile) for each patient. As more data becomes available, the profile becomes more accurate; making associations the human brain may never arrive upon. Unlike evidence-based medicine, prediction modeling does not rely on a bell curve, nor does it make assumptions based on a “typical” patient. In the world of small data, treatment guidelines become much clearer because the evidence is collected and applied to a data set of one.
Predictive analytics can greatly increase the accuracy of diagnoses, allowing physicians to see patterns more easily and quickly. It can also be used to identify at-risk populations, almost down to an individual, so physicians and payors can target them with disease prevention or early interventions. It can also help create cost transparency around individual patients and similar cohorts.
Patients will benefit most of all from this new form of hyper-personalized health care, and data shows they are ready and willing to embrace this new era of health care. In fact, 90% of consumers said they would be willing to share wearable or app data with medical providers. Physicians will be able to offer more definitive diagnoses and more effective treatments for all sorts of conditions, and they’ll be better able to prevent or tamp down complication associated with many common diseases like diabetes and COPD. Patients will become more informed and seeing the “writing on the wall” could activate patients to take more responsibility for their health.
Where do you see the future of small data and predictive analytics going? Will it deliver on the promise of truly personalized medicine? Share your thoughts in the comments section below.
 Accenture Research. Patients Want a Heavy Dose of Digital. 2016.