Case Study: Pharmacy Intelligence

How Pharmacy Intelligence Gave FGC Health the Predictive Power to Save Lives

Pharmacy Intelligence is an AI-powered intervention tool that harnesses a pharmacy visit to improve chronic disease management.


Imagine if you could change the world simply by filling out an online form.
Or perhaps just one part of the world: the part where patients with intricate combinations of diseases are at risk due to missed opportunities in patient care.
Dalbir Bains—CEO of FGC Health, a Canadian retail pharmacy chain—had this goal in mind when he took to the Internet.
His search brought him directly to the homepage of Lumiata, a disruptive new force in the healthcare industry, one that shared his goal of improving quality of care for patients. There, he filled out a form detailing his goal to leverage AI and machine learning to enhance the lives of his most at-risk pharmacy customers.
The completed form arrived at the desk of Ali Veldman, Senior Account Executive for Lumiata. As the initial point of contact for potential business users, Veldman was eager to show Bains how partnering with Lumiata could empower his organization to provide predictive insights that would make it easier to improve outcomes for patients.
“I reached out to [Bains] and scheduled an initial discovery call,” she says, “and it was like the stars had aligned.
Over the next four months, the two organizations would work together to hone in on Bains’ vision for his customers, collaborating on the complex process of turning FGC’s objective into a practical AI application.
By fall 2020, Pharmacy Intelligence would come out the other end with the predictive power to save lives.

The Challenge of Bridging the Ambulatory Care Gap

During those early meetings, FGC outlined what they termed the ambulatory care gap, or the fact that patients visit their doctors very infrequently, making it harder to track and monitor chronic conditions.
“In Canada, patients may be polychronic—managing multiple conditions and medications—but they’re not talking to their doctor that often,“ says Jenny Dann-Fenwick, Product Manager and Customer Success Representative at Lumiata.
“They may see their doctor once a quarter, or maybe even only twice or so a year.”
The problem is that for some patients, seeing their doctor on such rare occasions could have dire, if not deadly, consequences.
Patients with Type 2 Diabetes (T2D) are among those at the highest risk. According to the American Heart Association, cardiovascular disease (CVD) is the leading cause of death among those with T2D.
Furthermore, the combination of these two diseases requires special medical care and pharmaceutical attention. For example, diabetic patients who are also diagnosed with CVD may require a different medication regimen than what they took to deal with T2D alone.
However, these same patients are often missing out on the critical face-to-face interactions with their doctors that could help to catch the onset of CVD and better manage this complex combination of diseases. But what these patients are doing is coming to the pharmacy every month.
“That was the opportunity space that FCG saw to improve the care for these chronic patients,” Dann-Fenwick says.

“That was the opportunity space that FCG saw to improve the care for these chronic patients.”


The Vision of a White Glove Service

Bains knew that the Canadian healthcare market was ripe for a solution that employed AI and machine learning. Moreover, the best avenue would be to meet his patients where they already were: at the pharmacy.

“FGC was looking to create this sort of white glove service for the patients that were coming in to get their prescriptions filled at the pharmacy,” Veldman states. The service would target patients with chronic and complex disease groupings, like T2D and CVD, who are at the highest risk for complications due to gaps in patient care.
Unfortunately, as for many early stage startups looking to incorporate AI into their problem space, the barriers to entry seemed too much to overcome: no internal data science team, no experience deploying an AI solution… Not to mention the fact that their only dataset was an internal collection of patient pharmaceutical records.
It appeared that they had reached a gap of their own. For FGC, a partnership with Lumiata was the obvious solution.

The Path to Deploying Pharmacy Intelligence

“Lumiata has provided us with the predictive power to intervene and save lives. With three simple questions, our retail pharmacists can address health issues that have gone unnoticed and untreated.”​

Dalbir Bains, FGC Health Chairmain, President and CEO​

“I know they were really excited because we had a large internal data asset,” Dann-Fenwick explains. She’s referring to the 120+ million patient medical records that Lumiata uses to train their machine learning models. Lumiata’s internal dataset not only contains EHR claims and medical records on millions of patients, but it’s also enriched with the help of clinical IP and proprietary disease codes.

“I think that really differentiates us.”

That wasn’t the only defining factor, however: Lumiata brought a breadth of analytics, modeling, and data management capabilities to the table, which would drastically simplify the process of incorporating AI into any existing workflow.

FGC could choose from over 100 comprehensive pre-built models, trained on hundreds of diseases. They had the option to precisely tailor the chosen model to fit their specific healthcare needs.

Underneath it all was a solid foundation of in-house expertise: an elite team of developers, engineers, data scientists, and clinical advisors that ensured FGC would be as successful as possible at predicting disease onset in their patients.

The end result was Pharmacy Intelligence, an AI-powered intervention tool that harnesses a pharmacy visit to improve chronic disease management.

In Canada, patients who come into the pharmacy undergo a routine medication review. It is during this process that the pharmacist can deploy the Pharmacy Intelligence tool in real-time. With just 3 simple questions, the algorithm will predict within seconds the likelihood that a patient has both T2D and CVD.

If the model predicts that the patient has the combination of diseases and may therefore be a good candidate for medical intervention, then the tool produces a list of targeted screening questions for the pharmacist to ask the patient. Once the prediction has been validated, the tool offers targeted, actionable interventions for the pharmacist to pursue, such as a recommended new drug to add to the patient’s medication regimen.

Pharmacy Intelligence integrates with the pharmacist’s existing workflow and focuses outreach on only those patients who are at the highest risk of developing complications from the combination of CVD and T2D.

The Results

lumiata case study results

How Can You Leverage AI In Your Business?

Tech giants are always looking to disrupt major industries. Thanks to Covid-19, these efforts have sped up in recent months, with the rollout of services like Amazon PillPack shifting the way pharmacists connect with their patients.
Small businesses like FGC Health are keenly aware of the changing tides in healthcare, and they’re looking for ways to take on these corporate juggernauts and offer a more personalized experience to customers.
Dann-Fenwick sees boundless potential for Pharmacy Intelligence and similar AI-based intervention tools.
“AI is still so new, especially in healthcare, but it enables you to do things you probably couldn’t do otherwise,” she says. “With just a little ingenuity, you can truly help people.”
Still, there are hurdles to overcome when deploying AI and machine learning on a smaller scale. Collaborating with a team of trained AI experts will help you hit the ground running when it comes to leveraging AI and machine learning in your business.
Are you looking to innovate with machine learning and artificial intelligence, but aren’t quite sure where to start?
Lumiata can help you refine your vision and get an MVP to market so you can quickly see results. Our ready-made models can be deployed out-of-the-box or tailored to fit your specific use case. With our vast internal datasets, you’ll be able to generate accurate predictions for the onset of a multitude of diseases in no time.
AI doesn’t need to be a massive endeavor! With the right solution in place, you can begin to capitalize on the predictive potential of AI and make immediate improvements to outcomes in patient care.