This article was originally published in Drug Channels.
In a competitive market, pharma companies are increasingly relying on various datasets to drive their decision-making. With so many disparate sources, however, data standardization presents a challenge, especially for companies eager to use predictive analytics tools. In a recent survey of 125 pharma executives, nearly half cited data integration and cleanliness as the primary roadblocks to adopting technologies like AI.
Harmonized data is a prerequisite for generating actionable insights, with or without sophisticated data science engines. For pharma companies, integrating internal and external datasets to create a single source of clean, unified data is an imperative. Without visibility across the product lifecycle, manufacturers are climbing a mountain without a guide—as those with more efficient data strategies forge ahead.
Making the Case for Data Integration
An integrated data model is especially important for market access teams, as they study the impact of payer and provider behavior on utilization. Integrating real-world datasets with payer policy, restriction, and formulary data reveals the full scope of the patient journey, helping pharma identify and mitigate the specific barriers impeding access to their therapies.
By bridging medical and pharmacy claims to policy and restriction data, pharma can explore the difference in how payers say they’ll manage a drug versus the reality. Every day, thousands of claims are processed for drugs that are technically not covered on published formularies. Integrated claims and coverage data quantifies how medical exceptions, new-to-market policies, and unpublished policies affect patient access. By tracking the impact of payer restrictions on time to treatment, manufacturers can advocate for adjustments to speed access.
The addition of other real-world datasets (RWD), like lab and EMR data, completes the picture, showing how patients proceed from symptoms to diagnosis, treatments and outcomes. EMR data reveals the nuances of inpatient care, while unstructured clinical notes help pharma pinpoint specific findings, biomarkers and genetic variants. Lab test results can serve as real-time trigger events to help pharma target prescribers before they make treatment decisions. They can also be used to track disease progression over time, helping manufacturers amass efficacy data.
All of this RWD enriches the traditional market access trifecta of coverage, restriction and pathways data, enabling new commercial insights.
Establishing a Feedback Loop
Historically, the pharma pipeline functioned in silos. Clinical development focused solely on the data fueling and emerging from their trials, while commercial teams used market access data—supplemented with RWD—to drive utilization.
In recent years, however, pharma companies have strived to create a 360-degree feedback loop to ensure that RWD from their existing brands is incorporated back into development. Along with ensuring that new drugs are effective and accessible, they also want to establish a more sustainable pipeline, which requires a holistic view of how patients receive care.
With the addition of forecasting and clinical trial data to RWD-informed coverage data, pharma companies can start looking across both sides of their house at once. For example, market access data is increasingly being considered during clinical trial design, as payer preference and reimbursement decisions depend in large part on trial outcomes.
Integrating trial intelligence with payer policy data provides unique insights for clinical teams. Historically, how has the achievement of payer-preferred endpoints impacted performance? Payers tend to cover a new drug to label, unless there is a significant differential in trial results within a category. As the tipping point is typically the achievement of specific endpoints, knowing which ones are most likely to drive preferential coverage can impact trial choices.
Similarly, the marriage of forecasting and market access data helps pharma see not only a drug’s performance, but also its associated sales projections. This unified data provides a better perspective for market access teams, as they can now determine how various contracting decisions will impact projected forecasts.
Predicting Potential Outcomes
Once the right datasets are harmonized, pharma companies can begin to layer in AI-driven analysis for guidance. For example, manufacturers could use AI models trained on aggregated trial intelligence to generate recommendations on everything from the best I/E criteria to the best investigators to use in a trial. Essentially, these models leverage historical data to select the ideal parameters for a future trial of choice, which is an excellent example of predictive analytics driving pharma decision-making.
AI models can also be used to predict the coverage uptake curve for drugs still in development, helping manufacturers to make better go-to-market decisions. By parsing historic market access data for relevant drugs—and making connections between endpoint selection, payer behavior, and hospital/physician utilization—a predictive analytics tool can generate precise recommendations for each step of the drug life cycle.
In an ideal world, pharma companies would know in advance not only the optimal design for their trials, but also the precise payers, PBMs and IDNs to target for maximum access. With an integrated data model fueled by powerful data sources from pipeline to prescription—and the addition of technologies like AI—manufacturers can move toward a more predictive, patient-centered future.
Learn more about NorstellaLinQ, pharma’s first fully integrated data asset combining claims, labs, and EMR data with forecasting, clinical, payer and commercial intelligence.
