January 17 2019 | 0 Comments | 103 reads Average Rating: 3

Drowning Under a Deluge of Data? 5 Ways Machine Learning Can Help

by Lalithya Yerramilli in Health Analytics

The healthcare industry is amassing more data than ever before. Consider this: The U.S. healthcare system was already producing 150 exabytes of data several years ago and was expected to soon generate zettabytes (1021 gigabytes) and eventually yottabytes (1024 gigabytes) of data.1

The problem: It’s not humanly possible to leverage all this data for the greater good. Machine learning can help by quickly sifting through massive amounts of unrelated data and discovering patterns and trends that are beyond the grasp of mere mortals.

That’s good news for pharmaceutical companies. Instead of struggling with an overload of data, you can rely on machine learning to make the most of it by:

  • Showing the real-world value of drugs. With value-based models becoming more prevalent, pharmaceutical companies are under more pressure than ever to substantiate outcomes. It’s no longer enough to prove a drug’s value via clinical trials. Instead, you need to demonstrate value in real-world settings. Machine learning can accomplish this by combining medical and pharmacy data to deliver information on outcomes, such as total cost of care, rate of inpatient admissions, and emergency department visits over a period of time.
  • Optimizing sales during clutch periods. Life sciences companies typically need to optimize drug sales during two important periods: right after a drug is launched and then about six months after the drug comes off-patent. Machine learning can help by uncovering optimal target markets, such as neighborhoods with a highly probable concentration of undiagnosed chronic obstructive pulmonary disease (COPD) patients when the organization is introducing a COPD-related drug.
  • Supporting sales and marketing efforts. Machine learning can ensure your company is reaching out to the right physicians with the right message at the right time – and in the right way. For example, physicians in areas with a growing African-American population can be targeted with messages about asthma since African-Americans tend to have a higher instance of asthma overall and the death rate with asthma as the underlying cause for non-Hispanic Blacks is 2.5 per 100,000 deaths, compared to just 0.9 for non-Hispanic Whites.2
  • Discovering a different purpose. Machine learning can uncover off-label uses for drugs by comparing HCC and NDC codes, along with biometric data, to determine if there are unexpected correlations in use and outcomes. Once these uses are discovered, you can investigate them more thoroughly and test to determine if there is a new market for an existing drug.
  • Developing drugs for rare conditions. Feeding data about rare conditions into regression models typically doesn’t work, because the prevalence of such conditions is too small for traditional techniques to detect correlations. Machine learning, however, can identify the tiniest correlations in the largest amounts of data. As such, you can address these rare conditions with proven treatments.

These are just a sampling of the ways that machine learning can help your life sciences company succeed. Can you think of any other challenges that machine learning can help with?

The blog was based on “5 Ways Machine Learning Can Contribute to Life Sciences Organization Success,” an article that appeared in Becker’s Hospital Review.

References:

1. IHTT. Transforming Health Care through Big Data Strategies for leveraging big data in the health care industry. 2013.

2. U.S. Department of Health and Human Services Office of Minority Health. Asthma and African Americans. https://minorityhealth.hhs.gov/omh/browse.aspx?lvl=4&lvlid=15

Rate this Article:

Rating: 
Average: 3 (1 vote)

Author
Lalithya Yerramilli
Vice President, Analytics

Lalithya Yerramilli has 15 years of experience in analytics in insurance, healthcare, and life sciences industries working with customer info-base, transactional, physician level, patient level, claims and longitudinal datasets.

Read full profile and other posts

Log in to post comments

SEARCH BLOG

OUR THOUGHT LEADERS

Arun Rangamani
SVP, Care Optimization and SCIOXpert Services


Ben Steverman
Chief Technology Officer


Bob Abrahamson
Vice President, Product Management


David Hom
Chief Evangelist


Dr. Kevin Keck
Chief Medical Officer


Jen Cressman
Vice President, Professional Services


John Pagliuca
Vice President, Life Sciences


Jonathan Niloff, MD (Guest Author)
President, Niloff Healthcare Strategies, LLC


Lalithya Yerramilli
Vice President, Analytics


Lesli Adams, MPA (Guest Author)
Director of Population Health Strategy, Oracle Corporation


Leslie Strader
Project Manager


Linda Pantovic (Guest Author)
Director of Compliance & Risk Adjustment, Scripps Health Plan Services


Mark Feeney
Life Sciences Consultant


Monique Pierce
Vice President, Business Optimization


Nayfe Faillace
Chief Compliance & Privacy Officer


Nicole Cormier
Senior Manager, Home Health


Priyanka Rajkumar
VP - SCIOXpert and Solutions, Analytics


Rachel Hall
Senior Business Analyst


Rena Bielinski
SVP, Strategic Accounts


Rodger Smith
SVP, Payment Integrity


Rose Higgins
President, North America


Subha Vaidyanathan
VP, Technology and Data Management


Taryn Bevilacqua
Compliance Director


Tom Peterson
SVP, Risk Adjustment


ARCHIVES

Sign up to receive the latest SCIO news & insights, industry updates, event updates and more, right in your inbox.