Any large health organization that seeks to care for large numbers of individuals faces a daunting problem: how to best allocate scarce resources and be more efficient. Lack of efficiency comes in several forms: failure to deliver care according to best practices, uncoordinated or fragmented care, over-treatment, and fraud.

Predictive analytic models have been used to improve the understanding of healthcare delivery for decades.  Lack of ability to predict health outcomes was a challenging task until the advent of big data. Now predictive analytics is used to help clinicians and care managers anticipate problems before they develop, and mitigate health issues before they worsen.

With the use of big data analytics, predictive models can be designed to:

  • Incorporate in clinical workflows to facilitate care management and identify and address the needs of at-risk individuals
  • Understand the treatments with the potential for better outcomes
  • Use to perform risk adjustment on quality measures to account for patient severity and allow benchmarking between providers

Today, when a patient is admitted to the hospital for a heart attack, their individual case is compared with millions of prior cases whose outcomes are already known. Predictive analytics allows caregivers to input all of the patient’s unique data and generate a plan that is most likely to result in the best possible outcome. In this patient’s case, this would mean preventing hospital re-admission, choosing the appropriate level of post-acute care, and allowing them to return home as healthy as possible.

The full power of clinical prediction is best realized when the computational question is carefully defined, specific variables are gathered, a targeted need is met, and participants are willing to act.

The key to successful predictive analytics implementation is rooted much more in upfront planning than harnessing big data. It actually begins well upstream of the predictor and implementation.

First, it is essential to accurately model the workflow and detail the specific question at hand that you want the computer to address.

Second, collect the necessary data specific to and characteristic of that problem space such as:

  • What is known about the specific patient?
  • What is known about that population?
  • What supplementary data can be leveraged from external and public sources?

Third, recognize the weaknesses and leverage the strengths of various algorithm approaches. Difference algorithms cater to specific data sets. One size doesn’t fit all.

Last but not the least, find the appropriate clinical group and environment to implement the strategies.

Prediction should not be done for prediction sake. The not-so-obvious irony is that without having the proper framework in place, with willingness to intervene and context for meaningful use, prediction is really not very useful. In fact, it is often a waste of time and money.

In healthcare, the trade-off of a more generalized prediction model that inputs Big Data and global features is that targeted utility is lost or diluted. The very features that characterize a condition well are the attributes that can train an accurate predictor. But if those features do not stand out above the background noise, the predictor only finds the noise well. For this reason, prediction focused on a specific clinical setting or patient need will always trump a generic predictor in terms of accuracy and utility.

For predictive analytics, it’s the intervention that matters most. After all, it’s the intervention–not the predictor–that will improve patient care.