AI-Driven Strategies in Pharmacy Practice: Advocacy in Action

Pharma Now advocates for broader, more equitable access to AI applications within the pharmaceutical industry. Drawing on insights from previous articles, this piece aims to engage a diverse readership by exploring AI's potential, particularly in predicting, preventing, and managing drug-drug interactions (DDIs) in real-world pharmacy practice. Using the case of Glucagon-Like Peptide (GLP-1) receptor agonists (GLP-1RAs), we discuss key considerations for AI’s utility in effectively addressing such complex challenges.

GLP-1 receptor agonists, such as semaglutide (Ozempic®, Wegovy®, Rybelsus), have transformed the treatment of Type 2 diabetes and obesity. Previously, Pharma Now highlighted the increasing use of these drugs—compounded by supply shortages and compounding pharmacy practices— which have introduced significant challenges in regulation, clinical care, and logistics. In this paper, we build on these concerns by examining how GLP-1RAs impact gastric emptying, potentially altering the absorption of other medications. As patients, healthcare providers, and regulators face increasing demands and limited resources, AI tools can be crucial in predicting, preventing, and mitigating these drug-drug interactions (DDIs), particularly in a multilingual and diverse healthcare environment.

One of the biggest challenges to using AI effectively in the pharmaceutical field is that many professionals still don’t fully understand how powerful AI tools can be. Technology is advancing so quickly that many people in the industry, especially those who aren’t very familiar with technology, don’t realize the full potential of AI. Bridging this knowledge gap is crucial for successfully integrating AI into pharmacy. It involves education, mindset changes, and learning from real-world examples where AI has already succeeded. This article draws insights from a recent Pharma Now publication to explore how AI—through predictive analytics, natural language processing, and multilingual translation—could transform pharmacy practice into a proactive hub for identifying and resolving DDIs before they harm patients.

Another gap that requires to be addressed is referenced on a previous PharmaNow article “GLP-1 Agonists: From Diabetes Management to Weight Loss Solutions – FDA’s Role and the Compounding Controversy” states that “The challenge for regulators is that compounded medications are not subject to the same rigorous approval process as their commercially manufactured counterparts, which means the FDA must ensure the safety and efficacy of these alternative treatments.”

Given this anomaly with compounding pharmacies regulatory oversight and absent a typical ANDA submission stringent data requirements for safety and efficacy considerations, how then would an AI/ML generative reply insight fill in the gap with this inconsistency?

Can we truly rely completely with a purely AI/ML assisted introspection of the gaps or differences with US FDA regulations for compounded drugs versus drugs submitted with an ANDA submission? FDA regulations are not negotiable, complex and undeniably challenging to navigate.

A recent guidance published by the FDA Considerations for the Use of Artificial Intelligence To Support Regulatory Decision-Making for Drug and Biological Products states that if the COU or context of use involves a “human in the loop, we are to ensure that the evaluation methods for AI/ML model deployed consider performance of the human-AI team, , rather than just the performance of the model in isolation.

A solution maybe present where a pairing of AI/ML LLM model with a human-in-the-loop search platform can function as ‘dueling banjos. However, human handicaps quickly creep up when anyone becomes Google search-fatigued with peeking and poking search results links given the search’s platform’s web ‘noise and ads’ clutter these days. FDA.Gov’s website search, on the other hand, fails in its own way with its abysmal still meta-based search algorithms stuck in the Middle Ages with today’s big data and AI/ML-driven environment.

Why GLP-1 Therapy Puts the Spotlight on DDIs

GLP-1 receptor agonists work by slowing down gastric emptying by up to 70%. This change prolongs how long drugs stay in the stomach, which can affect how well they are absorbed. For medications with narrow therapeutic windows (i.e., drugs that can have serious side effects if not taken in the right amount), this is a big concern. For pharmacists, the stakes are high: if a patient taking GLP-1 therapy experiences altered absorption of other essential medications (such as anticoagulants, birth control, or immunosuppressants), the difference between effectiveness and toxicity can be minimal.

Moreover, GLP-1 agonists are increasingly being used for weight loss, not just diabetes, which increases the chance that patients will take multiple medications (polypharmacy). This complicates the risk of DDIs even more. Compounding pharmacies, which make customized medications for patients, also introduce more complexity, as their practices may not always meet the same quality standards.

The AI Opportunity: Revolutionizing DDI Prediction & Mitigation

AI has the potential to transform drug safety by enhancing the prediction, prevention, and management of drug-drug interactions (DDIs), particularly in multilingual environments. AI systems can analyze extensive patient data—such as electronic health records, medication histories, and real-time lab results—to identify at-risk patients. AI-powered pharmacovigilance platforms, like the Medication Extraction and Drug Interaction Chatbot (MEDIC), are being explored to improve drug safety and monitoring by leveraging prior knowledge, pattern recognition, and predictive modeling.

However, beyond addressing data quality and privacy concerns, the democratization of these efforts through large language models like ChatGPT requires careful consideration. Pharma Now's discussions on "SMART Tools," "SMART Practices," and "Good Linguistic Practices" emphasize the due diligence needed to ensure AI and machine learning are applied responsibly in complex environments. These considerations are crucial in real-world pharmacy settings, where they can help bridge the gap between compounding challenges, prescriber preferences, and patient safety.

So, is there a better way to bridge both gaps on democratizing AI/ML use and understanding for pharmacy stakeholders on one hand and the lack of FDA regulatory oversight on compounding pharmacies for GLP-1 drugs?

SmartSearch+ with its full text search (actual contents) capability rendered by human-in-the-loop side by side with an AI/ML LLM FDA Copilot for FDA.gov may be the answer. This unique ‘dueling banjos’ strategy can enhance or boosts gap finding within reach to the maximum possible when You- the-Human pharmacist are full-text searching Actual FDA.gov regulations for compounding drugs vs a drug sponsor filing its ANDA regulatory submission. Deploying both ‘banjos’ on one app cloud platform & user interface (laptops, mobile or tablets) can further boost newfound efficiencies to new heights.

As AI tools evolve, pharmacists must stay current on AI-driven analytics, communication practices, and emerging trends. The combination of advanced AI tools and pharmacist expertise can help create a proactive quality management system—what we call "SMART Shepherding"— throughout the medication process. Pharma Now, readers can better understand how to leverage the potential of AI and apply it practically in their pharmacy workflows by looking at real-world examples and comprehensive regulatory guidelines.

Conclusion

AI is not a distant future in DDI prediction and mitigation; it is already happening. The key to success lies in strong collaboration between pharmacists, healthcare providers, AI developers,

and regulators. Rather than focusing solely on the challenges of GLP-1 therapy, the industry can use this opportunity to rethink how we manage complex drug regimens.

By leveraging advanced AI tools—ranging from predictive analytics to multilingual support— pharmacists can shift from passive response to proactive prevention. With responsible use and a focus on good linguistic practices, these innovations can be democratized to improve patient care across diverse populations and healthcare systems worldwide.

In summary, the challenges posed by GLP-1RAs and DDIs are door openers or stepping stones toward a more agile, data-driven, and ethically guided pharmacy practice that prioritizes patient well-being while promoting continuous improvement across the industry.