How AI, analytics and collaboration solves professional drug spending challenges
Health planning and pharmacy welfare managers face numerous challenges in trying to manage their spending on specialized medicines. To address issues such as rising drug prices, data silos, and opaque clinical and financial risks, collaborative partners are critical to creating more sustainable approaches to drug management. Integrating artificial intelligence (AI) and advanced analytics into these collaborations is becoming best practices to uncover hidden cost drivers, simulate real-world impacts, and predict future spending with greater accuracy. Nicole Bulochnik, senior vice president of drug value strategy at Abarca, explains how the healthcare and pharmaceutical sectors work together to change the cost of specialized drugs.
To support the cost of specialty drugs, Bulochnik stressed that the biggest obstacle to the market is the high level of unknowns (unpredictable results, limited long-term data and uncertain patient volume), which amplifies all the cost pressures that are already at work. Prices continue to rise as payers and hospitals shoulder explosive budget efforts, such as growing gene therapy, positive revenue targets within pharmacies, and loose marketing regulations-driven budgets. Because specialized therapies serve relatively small patient populations, manufacturers set higher prices to recover their soaring R&D investment.
Spending on drug R&D increased by nearly 50% between 2015 and 2019, according to the Congressional Budget Office. It could cost as much as $2 billion to develop drugs.
Other factors to consider include:
- Complex Utilization Management – Develop and manage step-by-step therapy, prior authorization and formulation processes can be resource-intensive.
- Inconsistent or incomplete data from data silos from multiple sources can make things more complicated because they can hinder decision making.
- Ensuring that patients adhere to treatment while minimizing waste is an ongoing challenge.
- High variability and unpredictability of specialty drug initiation and adverse event rates make budget management difficult.
“From our perspective, the fastest win is to splice the data together – the claim, the lab, the genomics and the socially determined feed – and then let the prediction model surface, what really costs before hitting the ledger,” Bulochnik said.
“AI can now predict not only the number of new specialist therapies begins, but also the possibility of adverse events, helping to plan more accurately and intervene faster,” she added.
To develop a more sustainable approach to specialty drug management in specialty drug management, health programs, PBMs, drug manufacturers and specialty pharmacies, unified solutions must be designed in collaboration and jointly.
“We do not rely on static data interchange, but instead jointly develop AI-Ready data pipelines and build joint analytics solutions, providing common visibility into cost and quality drivers,” Bulochnik noted.
But the opportunity did not stop sharing data. Innovation strategies can be constructed by working closer together among healthcare entities.
Adherence interventions for AI health: Use AI to design patient-specific outreach strategies, limit drug waste and activate assistance programs. Pharmaceutical partners can enhance their impact by providing additional compliance resources.
SDOH and nursing resource integration: Incorporating social determinants of health (SDOH) resources into patient models to match individuals with food, transportation, and financial support programs. Share insights with pharmaceutical companies to expand patient access programs.
Pharma Co-innovation Pilot: Develop family diagnostic or follow-up kits for patients with active problems, using real-world data to perfect professional drug regimens.
Analytic-driven decision support: Implementing a system that crosses clinical criteria with coverage rules and predicts side effects risks, improving case management and appeals.
Comparative effectiveness modeling: Use AI to analyze clinical outcomes, economic incentives, and real-world use to recommend the most cost-effective treatments.
Dynamic Demand Forecast: Establish models for expected peaks in specialized drug utilization and implement early mitigation strategies, such as negotiating pharmaceutical price protection.
Switch Readiness Score: Applied machine learning to evaluate patient stability and transition preparation for biosimilars or low-cost therapeutic switches.
“From there we can develop and optimize contracts to maximize value-based opportunities with patient outcomes to minimize financial risk,” Bulochnik stressed.
“By allocating AI-driven insights to a collaboration framework, we are smarter and faster to manage professional costs and improve patient care methods at the same time,” she added.
Data analytics can provide actionable insights by predicting high-cost patients. Machine learning models can mark people who may need special treatments to support proactive interventions. Health plans can track real-time medication adherence in patients and correlate compliance gaps with projected downstream costs, enabling formal outreach to avoid avoidable complications.
With the expected increase in the cost of specialty drugs, coupled with NIH cutting funding for clinical research, collaboration between healthcare and pharmaceutical sector companies and organizations becomes critical to preventing the price of specialty drugs from getting out of control. Utilizing AI and a shared commitment to innovation will move leaders and laggards in specialized drug management forward.
photo: Vithun Khamsong, Getty Images
