Using Data To Understand Medical Treatment Effectiveness

We are living in a world where the power of data analysis is transforming everything – manufacturing, finance…you name it. Now, it’s time for it to consider what data analysis – or in this case, data science – can do for health care.

In an intriguing article entitled Data Science and the Health Care Revolution, a team of writers from O’Reilly Radar provide an insightful glimpse into the many ways that data science can impact health care. The authors of this article – Tim O’Reilly, Julie Steele, Mike Loukides, and Colin Hill – chose to pointedly focus on how data science can help us use health care spend more effectively. They also explore how a better use of data can help people know in advance what kind of treatment will work for them.

The authors quickly point out we really don’t understand the relationship between treatments and outcomes. Hence, with over $2.6 trillion spent yearly in the U.S. on healthcare, it’s critical – even vital – to have a better understanding of treatment effectiveness, particularly on an individual level.

There’s a whole lot more health-related data available today than years ago, including, for instance, biological data such as gene expression and next-generation DNA sequencing. There is also more clinical and health outcome data available through electronic health records (EHRs) and drug and medical claims than we had before. All this data opens the door for new questions to be asked about what treatments work and for whom.

What We Can Do Now That We Couldn’t Before

Before, the answers all came from clinical studies, with treatments based on what worked for the statistically average patient. Now, there is much more knowledge about everything, from drugs, surgery, and disease management to medical devices and care delivery.

The authors present several examples of how data can be used more effectively in treatments, including this one:

“For a long time, we thought that Tamoxifen was roughly 80% effective for breast cancer patients. But now we know much more: we know that it’s 100% effective in 70 to 80% of the patients, and ineffective in the rest.”

 The article also shares insights on many other fascinating topics as well, including how data can help us understand the origins of disease and make hospital systems more efficient.

The authors also explore the myriad of non-traditional types of data that are now available to us that weren’t before, including data from patient networks and mobile apps. Then, acknowledging that data science is no longer optional in health care reform, the article begins to look at how to enable the data.

It’s time to unlock the silos

The problem with unleashing the true promises of data science in health care is one common to all industries. As the authors say, “…90% of the work is getting the data in a form in which it can be used; the analysis itself is often simple.”

According to the article, there are two main issues that are blocking the effective use of health care data today. The first one is simply the shape medical records are in. There are still a lot of handwritten notes that are not computable. Secondly, data is scattered in too many silos that aren’t sharing information.

The authors believe that better electronic health records will be a big step in the right direction. So will the ability to link those records, from the doctor’s office and labs to hospitals and insurers into a data network that stores all the patient data. The technology is here, say the authors, but the systems are not yet in place. The good news for us and for the health care system itself is that they are really close.

The full article, available through the link below, explores all of the aforementioned topics in much more depth and is well worth the read for anyone interested in today’s – and tomorrow’s – health care system.

Read the full article Data Science and the Health Care Revolution now! (No registration required.)