Data analytics help Express Scripts predict, address the problem of patients who aren't taking their medications as prescribed.
The problem of medication non-adherence--patients not taking their medication as prescribed--is a significant problem, one that cost the United States healthcare system more than $317 billion last year due to the treatment of medical complications that could have been avoided. But Express Scripts, the largest pharmacy benefits provider in U.S., hopes to reduce non-adherence costs with the help of big data analytics.
Express Scripts manages the pharmacy benefits for approximately 100 million Americans, which comes out to nearly 1.4 billion prescriptions per year. The company uses a data analytics system from Teradata to address medication non-adherence, which can lead to expensive emergency room visits, hospitalizations, and extra tests.
"We've developed predictive models that allow us to estimate, with a high degree of accuracy up to a year in advance, the chance that any individual patient will be non-adherent to therapy," said Dr. Bob Nease, Express Scripts chief scientist, in a phone interview with InformationWeek.
The company runs predictive models that look at more than 400 variables, many of which reveal actionable information to battle non-adherence.
"We know, for example, that people who are married or partnered do better than people who appear to be single," said Dr. Nease.
Children, age, and gender are important factors as well. "Couples who have small children in the home tend to have slightly lower adherence, and there appears to be a gender discordance between the patient and the doctor," Dr. Nease said.
"Male patients who have female physicians have slightly worse adherence than male patients with male physicians," he added. However, that doesn't hold true for female patients, who seem to do well regardless of the sex of their physician.
Zip codes and income levels also play a role in whether men take medications prescribed by female doctors.
"For male patients who live in relatively well-to-do zip codes, there's almost no effect based on the sex of the physician," said Dr. Nease. But non-adherence is magnified for male patients who live in poorer neighborhoods.
It's not any one variable that's particularly informative, but rather the entire portfolio, Dr. Nease said. Express Scripts uses a different set of models for each disease state, as well as models that predict cessation--stopping your medication entirely--and compliance. Other models study patients who've just started taking their medication, and people who've been taking a medication for a while.
To dive this deeply into its data, Express Scripts enlisted the help of data scientists trained in the financial services industry.
"We process over a billion claims a year, and probably cover the care for 1 of 3 Americans. So the data is flat-out enormous," said Dr. Nease. "And extracting insights from the data requires a fair degree of sophistication, which is where the folks from the financial services industry really helped us."
These analytical insights have enabled Express Scripts to launch Screen RX, a voluntary service that helps identify patients who are at increased risk of non-adherence, and do so before they run into trouble.
"It uses predictive models the way a doctor would use a screening test for high blood pressure, high blood sugar, or even for cancer," Dr. Nease said.
For instance, if a patient often procrastinates on refills, Express Scripts may try to get that person into a medication home-delivery program. Patients who often forget to take their medications may receive a beeper that goes off when it's time to take their medicine. Patients may also receive reminder text messages, or even electronic pillboxes that dispense medications automatically.
"We're very bullish on data and the ability to glean actionable insights, when done properly," said Dr. Nease. "Healthcare is desperately in need of this."
6 Tools to Protect Big DataMost IT teams have their conventional databases covered in terms of security and business continuity. But as we enter the era of big data, Hadoop, and NoSQL, protection schemes need to evolve. In fact, big data could drive the next big security strategy shift.
Big Data Brings Big Security ProblemsWhy should big data be more difficult to secure? In a word, variety. But the business won’t wait to use it to predict customer behavior, find correlations across disparate data sources, predict fraud or financial risk, and more.