THE WHAT

MACHINE LEARNING AND

Prescription Optimization

Doctors aren’t perfect. Sometimes they prescribe the wrong drug and sometimes they are unwittingly complicit in drug abuse. The opioid epidemic that is ravaging the United States is a tragic consequence of medications that were inappropriately prescribed and a balkanized health care system that failed to identify dangerous behavior.


Data analytics is helping the health care system fight back against the types of mistakes that led to the opioid crisis:

  • Pharmacy benefit managers are increasingly able to alert pharmacies to patients who are most likely to be abusing drugs.
  • They can flag patients who have filled more prescriptions than is recommended or patients who have received the same prescription from multiple doctors, indicating that they are likely “doctor-shopping.”
  • Analytics are similarly helping health care systems to hold doctors accountable for dangerous prescribing practices, including writing opioid prescriptions to patients who should not qualify for such powerful medication or prescribing larger quantities of a medication than is necessary.
  • And as genetic analysis becomes more sophisticated, health care providers will be able to more easily predict how patients will react to certain drugs, including whether they are likely to have a dangerous allergic reaction or develop an addiction.


Take a look at this sample project predicting doctors' prescriptions using Medicare data: