The advent of digital has transformed marketers’ ability to build segmented audiences and deliver highly targeted messages. Yet health care marketing has lagged far behind in data segmentation because of the complex requirements of HIPAA (Health Insurance Portability and Accountability Act).
The issue isn’t a lack of data but the time and resources needed to deidentify or anonymize patient information to make it compliant and usable for audience analytics.
Compliance costs money. And that high cost means most health care marketers work with relatively small data sets, which severely limits the accuracy of their audience analytics insights. That slow deidentification process also makes it hard to incorporate new information – meaning they’re often working with “old” data. Costly, unwieldy processes have historically constrained what marketers can do with health care audience data.
So how do we create a more agile system while still respecting the privacy of our audience? Let’s start by talking about what’s broken in the status quo.
Using AI to accelerate health care data deidentification
Health care audience development began with a system control process of deidentification. Using manual workflows, teams reviewed methodologies to remove anything that would cause reidentifiability. Then, third-party statisticians would audit the data and provide guidelines on how to combine it while still respecting privacy law.
Eventually, we’d arrive at a HIPAA-compliant audience. The process was very slow. And if an audience did not perform well, going back to try again was a long, costly process.
Today, marketers in other sectors are already using AI to “clean” and prepare data for analytics – and the deidentification processes described above present ideal applications for rapidly advancing AI tools. For example, by scanning documents for preset word groups, AI can group together matching information to produce targeted data. These AI-produced data sets can then be sent to a third party to check for compliance.
Most importantly, this AI-powered data deidentification can be done at a speed and scale that humans could never match – and for a fraction of the cost.
Insurance claims: an untapped trove of audience insights
The slow, costly deidentification process bred the other problem that’s plagued health care audience development: Marketers typically work with very small data sets. Unfortunately, the quality of analytics insights is directly correlated with the size of the data set.
This problem can be solved using data that already exists: health care insurance claims. From the data fields themselves to the interaction with insurance companies, health care claims data paints a picture of what’s occurring in the health care space for every therapeutic area.
When it comes to claims data, there is a lot of information. PurpleLab found 50 to 60 billion claims in total over the past decade. Unfortunately, to parse through these claims would require potentially thousands of hours of work. This is, again, where AI can help.
We created an AI that’s tackling both the data deidentification and data interpretation challenges for health care claims data. We’re seeing health care marketers use our toolset to build large-scale data sets that can deliver powerful insights while remaining broad and general enough to be HIPAA-compliant.
The final piece of the puzzle: Democratizing audience analytics
Removing deidentification bottlenecks and opening broader data sets gives health care marketers a powerful new foundation to build better audiences for effective campaigns. But more data doesn’t drive better results unless you’re able to interpret the insights at speed and scale.
Here again, traditional audience-building workflows hit a bottleneck: The specifics of health care patient information and claims data have historically required highly specialized data analysts. This drives up the cost of health care audience development and slows the data-insights-action pipeline, meaning marketers aren’t working with fresh insights or adapting to audience changes in a timely fashion.
There’s a clear need to democratize how we interpret that data. The problem is that the tools needed most for audience development are designed for data analysts, and no organization has enough of those folks.
Here’s what we wondered: What if we could remove that bottleneck and give everyone in the organization direct access to HIPAA-compliant health care claims data?
The democratization of health care audience analytics has to start with a straightforward user interface that’s accessible and understandable to everyone – not just health care data experts.
This idea allows the agencies and ad tech companies that partner with health care marketers to develop ideas and models on their own, tailored for each audience – patients, their families and caregivers, as well as providers and other stakeholders in the health care delivery ecosystem.
For example, at PurpleLab, we’re giving health care marketers and their agency and ad tech partners the ability to index for any condition or prescribing behavior. This can be used as the basis for consumer audience modeling, as well as building a targeted audience of health care providers.
The core idea here is simple: Open up access to intuitive tools that let people ask questions. Let the curiosity and wisdom of a broader crowd drive better questions and more insightful answers.
Giving patients the privacy-safe personalization they want
The conversation in health care marketing has traditionally framed what patients want (privacy) in opposition to what marketers want (targeting insights). But the consumerization of the health care world today sees patients taking a more active role in their health care decisions. They still expect their privacy to be respected, but they also want their decision-making process to be enhanced by information and offers that are personalized and relevant.
The future of health care audience development will bring together the interests of marketers and patients. Automated deidentification – built with privacy-safe measures and HIPAA compliance in mind – will put more information in people’s hands at each stage of the process. Empowering creatives to know the patient better, strategists to plan better and the media team to test more tactics to find the best performance.
This change isn’t just about improving ROI. When more of the right people are informed about the right treatments, we all win.
For more articles featuring Ted Sweetser, click here.
