Data and Analytics Drive Effort
to Curb Opioid Abuse: A Holistic
Approach for Health Systems
By Kapila Monga and Harpreet Singh
IN THE UNITED States, five percent of the world’s population
consumes about 80 percent of the world’s opioids, 1 a statistic
that merits the status of a public health crisis in which the demand and supply side of the issue deserve equal weight. Current efforts to address this crisis focus mainly on managing the
supply side of the equation through prescription guidelines and
relevant training programs. The reduction in demand that is
driven by awareness, treatment, and prevention of opioid abuse
through appropriate medical, social, behavioral, and policy interventions merits equal—if not more—focus.
While all players in the healthcare arena have a role to play, some
have a specific set of responsibilities. Physicians are uniquely positioned to address this epidemic as they are the first to observe
the signs and symptoms associated with opioid abuse and are
equipped to treat the illnesses that can result.
However, it must be acknowledged that this is a huge responsibility to place on the shoulders of one part of a complicated system. We need an effective combination of prevention and treatment strategy, which is why health information management
(HIM) can play a large role. They can assist with relatively simple
tasks such as altering electronic health records (EHRs) to help
physicians adhere to safe prescribing guidelines and converting
morphine milligram equivalents. Health IT systems can also help
identify opioid-seeking behaviors and patients at risk for overdoses and other adverse events. These and other data-centric solutions give health systems a needed boost in helping to predict,
plan for, and support prevention efforts to curb opioid abuse.
Prescription Opioid Abuse Engine
Two years ago, AHIMA launched an effort to improve docu-
mentation related to the use of opioids. 2 A review of physician
notes in EHR systems since reveals the significant improvement
that has happened in this regard. Inclusion of ICD- 10 codes
for opioid-use disorder and opioid overdose or poisoning has
helped further the cause. Much more can and should be done,
but it will take time to sort through the challenges inherent in
streamlining this documentation. The primary job function of
physicians is not medical documentation, and with EHRs be-
ing a contributing factor to burnout, asking more of them on the
documentation front is tricky.
At the same time, the healthcare industry today stands at a
crossroads when it comes to effectively analyzing and mining
EHR system data for reducing prescription opioid abuse, identi-
fying social determinants of health, slowing chronic disease pro-
gression, and reducing avoidable re-admissions, to name a few.
The diagram on page 37 represents a framework of a clinical
decision support system that utilizes data from EHR systems
to identify, predict, and eventually prevent prescription opioid
At the core of this decision support system is a quartet of engines: detection, intervention, monitoring, and prevention.
The detection engine identifies prescription opioid abuse
by tapping into the behavioral indicators hidden in physician notes, including pattern analysis and natural language processing, and predicting the same to help identify opioid abuse at an early stage.
The intervention engine recommends next-best actions
based on patients’ current needs, such as medication-assisted therapy, nonopioid pharmacological treatments,
and nonpharmacological treatments.
The monitoring engine ensures that the patients adhere
to the treatment provided and sends real-time alerts to