Case Study: Using Machine Learning and Advanced Analytics to Optimize Physician Outreach
- After experiencing a slowdown in new prescribers post-launch, a mid-sized biopharmaceutical company wanted to understand how multiple factors contribute to new prescriptions based on advanced analytics and data modeling.
- Medidata Acorn AI’s team of in-house data scientists and statisticians created advanced machine learning algorithms to search the sponsor’s expansive reservoir spanning 40 features including specialty pharmacy (SP), labs, call, email, and trigger data to test their hypotheses and uncover trends.
- In just two months, Acorn AI found some crucial factors that influence prescribers. These insights can help the company target the right physicians most likely to prescribe their medication in order to increase sales and improve health outcomes.
A huge part of maintaining post-launch momentum is developing an evidence-based marketing strategy, particularly for physician outreach. The first step to optimizing strategy is having the right guiding information. Some questions companies may have are:
- Who is prescribing my medication? Are marketing and sales activities reaching them?
- What factors influence a prescriber to give a patient my medication?
- Are there untapped educational channels we can leverage to increase physician disease and therapy awareness?
Medidata Acorn AI Commercial Data Solutions (CDS) partnered with a mid-sized biopharmaceutical company to help them answer these pivotal questions through advanced analytics and machine learning modeling.
This global biopharmaceutical company develops life-transforming medicines for patients with serious and rare diseases. Their FDA-approved treatment is a specialty medicine for patients with a rare bacterial lung infection.
Since receiving FDA approval and launching a few years ago, the company noticed a slowdown in new prescribers. Though this slowdown was expected, the commercial team was very interested to understand which prescribers were affecting the decline and how to address it. After purchasing third-party lab data, the company found that there is a significant population of physicians who test for this lung infection but have not prescribed their treatment.
The nature of this disease area makes it difficult for treatment to reach key specialty groups. Because the infection is often misdiagnosed as a condition presenting similar symptoms and progression, there are likely a higher number of patients going without proper treatment than known to physicians.
These factors made it challenging to gain an accurate understanding of prescription patterns. Although the company had some theories about what influences a prescriber, they wanted to definitively understand how multiple factors contribute to a new prescription based on advanced analytics and data modeling.
The company partnered with Medidata’s Commercial Data Solutions team to perform advanced commercial analytics in order to gain deeper insight into the prescriber journey. CDS uses a data-agnostic approach, integrating data from multiple, disparate sources to provide meaningful insights to field sales, market access, and commercial leadership.
First, Acorn AI’s team of in-house data scientists and statisticians compiled a list of every possible factor that could affect a prescriber’s decision-making. Next, the team created advanced machine learning algorithms to search the sponsor’s expansive reservoir spanning 40 features, including specialty pharmacy (SP), labs, call, email, and trigger data, to test their hypotheses and uncover trends. With the power of the robust Medidata data platform and innovative commercial data model—regularly processing 1B+ data points daily—the CDS team delivered granular insights that would greatly improve the sponsor’s physician outreach strategy in just two months.
Medidata’s analysis found some crucial factors that influence prescribers, including some findings that the sponsor had not previously considered. Here is a sampling of the findings:
- As the sponsor predicted, a higher number of sales calls indicated whether a prescriber would likely prescribe the medication.
- Although sales calls were an effective marketing method, the data suggested that approximately 20 calls was the saturation point. After 20 calls, the likelihood of prescribing no longer changes.
- Infectious disease specialists are more likely to prescribe the medication than pulmonologists.
- As infectious disease experts receive more calls, their likelihood of prescribing the medication goes down, though they are the majority of prescribers.
Armed with these insights, the sponsor is now able to refine their physician outreach strategy to optimize effectiveness and resource efficiency. These insights can help them target the right physicians most likely to prescribe their medication in order to increase sales.
Top 15 of 40 Features that Affect the Biotech’s Physician Outreach
This graph shows us the features that were most important when predicting the outcome—new prescribers. The features are ranked from top to bottom by order of importance. The graph shows that calls and newly diagnosed patients emails-opened are highly indicative of a physician prescribing the drug.
About Acorn AI
Acorn AI Commercial Data Solutions provides commercial data management technology and advanced analytics capabilities that give drug manufacturers a 360-degree view into the performance of their product from day one, making sure the right patient gets the right treatment at the right time.
Whether launching a new product or supporting a mature one, commercial launch teams need powerful AI and machine learning analytics capabilities that can quickly ingest data and deliver fast, accurate insights to stakeholders across the organization – no matter where they are in the product life cycle. CDS’s data agnostic philosophy, deep life sciences domain, and data ecosystem expertise let manufacturers turn their data into a competitive advantage and reach peak sales faster.