What We Learned by Applying the NIEM Standard to the FAACT Platform

What We Learned by Applying the NIEM Standard to the FAACT Platform

Adam Roy
Robert Reynolds
Robert Reynolds

We had a chance to sit down with Qlarion’s CTO Adam Roy and Robert Reynolds, Qlarion’s lead data engineer for the Framework for Addiction Analysis and Community Transformation (FAACT) program. They gave us an overview of the FAACT platform and explained the challenges and benefits of applying the National Information Exchange Model (NIEM), an XML-based framework that enables information sharing.

Q: What can you tell us about the FAACT platform? Just how big is it and where do the datasets come from:

AR: FAACT is one of Virginia’s premier cross-agency data-sharing initiatives. It is a secure data-sharing platform that was designed to help communities in the Commonwealth combat Virginia’s opioid and addiction crisis.

RR: The FAACT platform includes tens of millions of records. It includes anonymized information on drug use, types and length of treatment, demographic information, hospital admissions, state and local police data, and forensic analysis of compounds found at crime scenes.

Q: How can you draw insights if the data is anonymized?

AR: Data security and privacy are two of our top concerns. We don’t collect or share any information that could be used to identify a specific individual who may be suffering from substance use disorder. Individual records are anonymized and integrated so users of the platform can look for trends and patterns but can’t identify a particular patient or criminal offender.

With the millions of records that currently reside in the FAACT platform, users are still able to see patterns and identify insights. Qlarion has created some pre-built visualizations, analyses, and templates, but any of the users can also create their own reports and visualizations that are useful to their organizations.

Q: What is the NIEM standard and why is it relevant to the FAACT platform?

AR: The NIEM standard provides a common vocabulary that allows us to categorize data in the same way for vastly different datasets. The most common example is the field “Last Name.” In other datasets, this field could be called “Surname,” “Family Name” or, simply, “Name.” The NIEM standard allows us to clearly define data elements and eliminate ambiguity.

The Commonwealth actually required use of the NIEM standard when we started working on the FAACT platform. The goal of the NIEM standard is to create a common lexicon so that users from government, public safety, social services and healthcare can look at data from other domains and gain insights that could help them better serve the citizens of the Commonwealth.

Q: What were some of the challenges of applying the NIEM standard to these datasets?

RR: It’s more complex than you would think. For instance, how do you define an incident? Depending on the dataset, an incident could be any situation when:

  • The police or an ambulance is called
  • A person is admitted to the hospital
  • A patient receives a diagnosis
  • A patient is admitted to a treatment facility

By using the NIEM standard, each of these “incidents” can be categorized separately.

The way the different datasets categorize overdoses can also be highly complex. Most providers log overdoses as either fatal or non-fatal, but some datasets include a lot more granularity than others. This creates many questions, including:

  • How do you log overdose victims that have multiple health conditions?
  • How do you treat multiple admissions to a treatment facility? Is that considered one incident or many?
  • How do you log patients that may have multiple compounds in their system – but only one of those compounds is an opioid?

These are just some of the scenarios that we’ve had to consider when applying the NIEM standard to these various datasets.

On the plus side, the NIEM standard has helped us create data hierarchies. For instance, we can show crime scenes that took place on streets within neighborhoods are also part of towns and counties. We use this same hierarchical approach to make time and date information relatable. For example, if some datasets only provide monthly data, and others provide a date/time stamp, we can create hierarchies to show that incidents that happen on a particular day roll up into days of the week, months and years.

Q: In addition to standardizing the data, what is Qlarion’s role?

AR: Qlarion’s role is multi-faceted. We’re working with the state government to roll out the platform across the Commonwealth. At the same time, we’re working with other government organizations and first responders to join a collective data trust so that they feel comfortable providing data that we can input into the platform.  As such, we are responsible for data governance and security.

We’ve spent a lot of effort working to ensure that FAACT is a self-service platform. We don’t want the user to have to go back to the data steward to understand the data. In addition to the NIEM standard, we’ve created a data dictionary that defines each dataset and provides more granular information about how the data is structured; this ensures that the data is used correctly to make accurate decisions.

Finally, we help with tailored analytics so that various groups within the Commonwealth can make sense of the data. We provide them with valuable insights to effectively understand and proactively tackle the opioid and addiction crisis within their communities.