The role of data and analytics in addressing the US opioid crisis
Three years ago, prior to leaving practice as a pediatric emergency physician in a suburb of St. Louis, I coded (resuscitated) the last patient in my clinical career. Unfortunately, I had to give the parents the devastating news that their 18-year-old son had died of a heroin overdose. A life extinguished far too soon. This horrifying scenario plays out every day across the United States, from the big cities to the heartland. This is the front line of the opioid epidemic—a battle the country is losing.
Opioid addiction is a multidimensional and complex phenomenon. There is no silver bullet fix. Addressing the nation’s opioid crisis demands addressing multifactorial causes and impacts, which is not easy.
The best way to do this is with comprehensive data insight into risk factors, behaviors, patterns and profiles that inform effective intervention, education and prevention strategies. The good news is that local governments and organizations across the health and human services spectrum—from public health institutions and behavior health entities to pharmacies and providers—possess relevant data.
The bad news is that this data is isolated as individual datasets across multiple organizations. Complicating things even further, policies often prohibit agencies from sharing data with each other, and people are often ambivalent about sharing their personal data. Despite these barriers, accessing and assembling disparate data is critical to paint a full picture of all the factors driving the opioid problem. Progress does not come from having data. Progress comes from how organizations use it.
So, what that would this look like in practice? Take the example of babies born with neonatal abstinence syndrome (NAS). These babies become addicted to opioids in the womb. NAS is a lead indicator of women who maybe addicted to opioids. NAS data can be correlated with other risk factor data, including social, criminal justice and health data, along with clinician prescribing behavior.
By pulling all these together and using advanced analytics tools such as machine learning and predictive modeling, organizations can identify the nature of problems at a more granular level than ever before. Using data and analytics, it is possible to understand the story of specific clusters—or even a single individual—and predict the best possible measures to support them and target resources. Combining and analyzing data in new ways can trace not only the factors leading to addiction, but also the costs of all the services an individual may require as a result. By seeing the picture at this level of granularity, agencies and local governments can pinpoint where their resources, interventions and programs designed to address specific causal factors can be focused, with the greatest chance of success.
Once a program is in place, it is vital to measure its results. Constantly. Diligently. Continuous reporting of progress gauges the efficacy of opioid addiction programs and indicates where and how they may need to be adjusted.
By using data and analytics to create new insights, this nation can come one step closer to mitigating, and even preventing, the spread of this epidemic.