A Practical AI Roadmap for Pharmacy Adoption: From product to clinical services
Most “AI in pharmacy” pitches begin at the wrong end, making sweeping clinical promises while overlooking the foundation that pharmacists actually work on. Pharmacy is a two-part business: a product model (safe and efficient medication delivery) and a service model (clinical care that improves patient pharmaceutical outcomes). The core issue is that the impact and potential of AI differ significantly depending on which model of pharmacy is being considered. This lack of clarity stems from differing interpretations of terms such as 'AI' and 'pharmacy,' which leads to fragmented discussions and mixed expectations about AI's value and utility.
Pharmacy is not a homogeneous field. We have diverse interest groups and stakeholders, from regulatory agencies to corporations and professional identities. Each group has opinions about how AI should advance pharmacy. Ultimately, pharmacy is a business, and pharmacists are involved in the process from drug discovery to patient delivery. With that said, when people claim AI can have massive implications in pharmacy, I would agree, but only if we are precise about which aspect we are discussing. Clarifying this is crucial, as the true impact of AI hinges on the specific area of pharmacy under consideration.
For example, most presentations and conversations I’ve encountered focus on pharmaceutical dispensing in the community setting. This emphasis is unsurprising, as it shapes the most prominent public image, whether positive or negative, of the pharmacy profession. While topics such as increasing medication access, reducing costs, and building patient engagement are all important, they only scratch the surface of AI's potential in pharmacy. For example, I will hear talks on the use of AI to better choose and select the best drug for a patient, which, in theory, makes a lot of sense and is what we’ve been aiming for in the past century with EBM. However, our daily workforce is burdened by more mundane issues that hinder the pharmacist's ability to address this issue for every patient.
Several models that have helped guide the general approach of pharmacy.
To move beyond this surface-level focus, let’s take a broader view: What is the role of pharmacists, and where do we fit within the healthcare system? One of the most seminal position statements addressing this comes from Hepler and Strand in 1990, when they introduced the concept of “Pharmaceutical Care,” which is defined as follows:
“Pharmaceutical care is the responsible provision of drug therapy for the purpose of achieving definite outcomes that improve a patient’s quality of life. These outcomes are (1) cure of a disease, (2) elimination or reduction of a patient’s symptomatology, (3) arresting or slowing of a disease process, or (4) preventing a disease or symptomatology.”
They go on further to state that pharmaceutical care involves the patient and health team in designing, implementing, and monitoring a therapeutic plan to achieve these outcomes, which include (1) identifying potential and actual drug-related problems, (2) resolving actual drug-related problems, and (3) preventing potential drug-related problems, to be enacted across all practice settings.
I would like to emphasize here that this position paper arose as the profession was grappling with a professional identity crisis. For all too long, the business of pharmacy ruled supreme, with a singular focus on getting a product to the patient. The idea of a pharmacist being involved beyond assuring clinical oversight was still in its nascent phase.
A decade later, the pharmacy profession would undergo significant changes, with a push in the USA towards the PharmD model, with the expectation of more clinical duties for the profession in light of growing changes in reimbursement strategies for pharmacists' duties and an expectation of a booming pharmaceutical field with many new therapies coming to market.
However, even over 30 years later, the profession remains closely tied to a product-based model. Pharmacists generate revenue primarily by dispensing medications, whether in community, hospital, specialty, or long-term care settings. This business approach continues to shape external perceptions and day-to-day operations.
Nonetheless, the concept of a service-based model has been around for more than 50 years, and more pharmacists than ever now hold non-dispensing roles that involve clinical care and administrative duties, facilitating the safe and efficacious use of pharmaceuticals in society. The service-based model we have discussed further, whether it involves preventative care, vaccination programs, test-to-treat, or low-to-moderate acuity disease management, has indeed expanded the scope of practice.
Recognizing these distinctions is critical: the challenge for AI in pharmacy is to navigate and bridge both the product- and service-based models that define our profession. Focusing innovation solely on dispensing overlooks key aspects of modern pharmacy, especially as decision-makers seek solutions that enhance not only medication distribution but also the expanding clinical care and quality standards for which pharmacies are now responsible. Retail and health-system pharmacies face sustained reimbursement and staffing pressures. The durable path for AI is the unglamorous one: remove friction in getting the right medication to the right patient at the right time, and let that same infrastructure emit the clinical data you need to justify services. By integrating AI into this dual model, we can create a loop where data generated from both dispensing and services can feed continuous improvement, potentially enhancing both operational efficiency and patient care outcomes.
Taking all of this into account, I then pose the following question: Which AI tools will be easier to adopt into pharmacy practice, and which may be further out than anticipated? To answer this, let’s break the issue into manageable parts.
One general area that I think has seen steady strides over the years is “The Seven Rights” (or Five, if you go back years), which emphasizes the safe use of medications for patients. For those unfamiliar, those rights include:
The right Patient
The right Drug
The right Dose
The right Route
The right Time
The right Reason
The right Documentation
Traditionally, it was the first 5 items listed that we followed, which also goes with most prescriptions, for example:
John Doe is prescribed Ozempic 1mg SubQ QW.
Now, we could expand upon that to ensure it was properly administered (such as in an institutional setting through barcode scanning) and given for the right reason (such as diabetes via EHR records).
I want to start here, as this is, at the very least, the responsibility of any pharmacist to ensure. Mess this up, and too many problems can occur, because after all, we are generally the last check before a patient receives their medication throughout their health journey. It impacts the product- and service-based models that we have previously described.
We have made significant efforts to optimize this process, and modern technology has undoubtedly streamlined it. However, AI offers multiple avenues to enhance these processes through behavioral incentives. For instance, AI can employ nudges, such as alerts and default checks, to significantly reduce the cognitive load on pharmacists. By reminding pharmacists of the Seven Rights in real-time, these AI notifications can help reinforce safer habits and ensure adherence to best practices. Furthermore, aligning these AI features with incentives that promote compliance can strengthen the argument for early adoption, as they not only enhance safety but also improve workflow efficiency.
As we move past the fundamental seven rights, our attention shifts back to the product-based model of pharmacy. I likely don’t need to elaborate at length: The more drugs we sell, the more revenue we generate. However, scaling this simple on-paper process becomes challenging given reimbursement issues and market pressures. As a result, many pharmacies struggle to sustain the model.
Hence, we have seen a high interest in leveraging technology to help the business of pharmacy. More automation, central fills, better PMS tools that can maximize inventory, reducing friction points with payers, engaging patients, reducing medication churn, etc. For the pharmacy to survive in today's market, adopting a consumer-driven mentality is crucial. Reducing the friction of getting a medication to the patient is key to retaining customers and continuing to operate and grow. For example, average pharmacy profit margins have declined over the last few years due to increased competition and reimbursement cuts. This creates a pressing need for AI-driven efficiencies to help recover and stabilize these margins. By implementing AI solutions, pharmacies can more effectively navigate these challenges, ensuring operational sustainability.
As such, AI has a tremendous role to play, and it has for many years found a home in early business intelligence and ML tools to help with the operational and administrative aspects of pharmacy. Vision technology has been highly important for the rise of pharmacy automation, and the primary reason for the growth of central fill pharmacies, which has helped reduce the need for local pharmacies to spend labor time filling general medication refills, for example. This has now pushed mail-order business for companies like Amazon and its related services.
However, AI innovators face a competitive landscape in supporting medication dispensing, as established organizations are already integrating advanced technologies to enhance their operations. For example, introducing a new AI tool for pharmacy automation, such as streamlining IVR systems, requires navigating complex PMS integrations and adapting to varied customer requirements.
Mergers and acquisitions are likely to occur as companies expand their AI offerings and seek to improve efficiency. Pharmacy operators will evaluate new AI solutions critically before adoption. Key pain points remain, but operational improvements are possible, especially when solutions enhance pharmacy workflows and support medication sales.
I say all this not to discount the value of AI at this point, just to highlight that a perceived need may be an overemphasized shortcoming of current practice that could be mitigated through other issues that AI alone cannot solve. Keep this in mind and be mindful to consult with pharmacies across multiple locales and populations, especially when creating or training any new models you plan to deploy. To make this assessment more actionable, pharmacy professionals can start by conducting a needs analysis within their current operations to identify specific areas where AI could be beneficial. Engaging in cross-departmental discussions or workshops can help in pinpointing gaps and opportunities. Additionally, participating in AI model development workshops or pilot programs can not only enhance understanding but also tailor applications to meet specific organizational needs. Collaborating with technology partners who can provide insights into best practices and the latest advancements in AI can also be a critical factor in successful implementation.
Now, let's move on to the final component and focus on the pharmaceutical care aspect of pharmacy practice. I have already touched on this in a previous article, but it's worth mentioning some aspects here as well. The previous definition of Pharmaceutical Care underwent significant changes over the past decade, with the introduction of the Pharmacist Patient Care Process (PPCP) by the JCPP in 2014 and its subsequent update in 2025. The 5 aspects of the PPCP include:
Collect. Ensure necessary subjective/objective information from multiple sources (patient, caregiver, records, other professionals).
Assess. Assess the collected information to identify and prioritize needs, informing a care plan.
Plan. Develop a person-centered, evidence-based, and cost-conscious plan in collaboration with the patient/caregiver and team.
Implement. Provide person-centered care by implementing the prioritized plan in partnership with the patient/caregiver and team.
Follow-Up: Monitor and Evaluate. Follow up to monitor and evaluate plan implementation and the patient’s overall health with the patient/caregiver and team.
The PPCP is applicable in all areas of pharmacy practice, emphasizing the key points of ensuring the safe and efficacious use of pharmaceuticals for patients in various settings. However, the PPCP definitely has a focus on the clinical care aspect of patients that pharmacists provide, and fits into the expanding service-based model that the profession is aiming for at this time. Selling drugs alone won’t keep pharmacies open, and engaging in some form of service will be key in the coming decades to meet a changing health consumer market.
Nonetheless, a service-based model still has yet to reach a large enough business stance (across the breadth of pharmacists) to be a sustainable model at scale. Even in the use of clinical pharmacists in many health systems, their work is generally focused on cost-containment and reduction strategies, minimizing waste expenditure, and ensuring the safe and efficacious application of pharmaceuticals. It is not often that even these individuals can demonstrate a direct profit to their institution (aside from some settings). As such, the caveat is that while a service-based model is a goal, the product-based model that ensures some semblance of business cannot be abandoned lest the whole operation fall apart.
To navigate this challenge, pharmacies can employ several strategies to demonstrate the value of a service-based model. Tracking metrics such as quality-adjusted life years (QALYs), cost savings from reduced hospital admissions, and improvements in patient outcomes can provide tangible evidence of value to administrators and payers. Furthermore, implementing programs that highlight the positive impact of pharmaceutical care on patient satisfaction and adherence to medication regimes can strengthen the business case. By showcasing these outcomes, pharmacies can appeal to stakeholders and secure support for expanding service-based models.
This then brings me to the seminal argument we face with the application of AI in the pharmacy space: Where does it make sense to utilize it? When speaking with early adopters and those considering the use of AI tools, the items most discussed are those that can help maximize their current workforce and address current needs that have not been met, in part due to cost-cutting, staff burnout, and limited capabilities. I often hear, “Yes, we’d like to do X, but cannot due to Y.” Can AI then help with this?
My working theory thus suggests that AI tools that are likely to gain the most traction are those that can at least meet the minimum requirements of ensuring the Seven Rights and addressing pharmaceutical dispensing issues currently seen in pharmacy practice, because once these are more manageable, we can advance to further services. Going straight for service alone won’t win, unless it overlaps with the product and safety aspect of care.
At this point, the growing question is whether service can truly become a core focus for the pharmacy profession. Restating what I’ve thought for years: We lack sufficient data and outcomes to demonstrate the economic value of our services at scale to payers and administrators. While the VA and select institutions have made progress, expansion remains the challenge. I believe AI can help us reach this goal, provided we understand our limitations and use the tool thoughtfully. This is where we must start—before aiming for transformative pharmacy services, let’s ensure our foundation is in place to help drive meaningful innovation and adoption.
To bridge this gap, it is crucial to focus on measurable outcomes that can convincingly demonstrate the value of our services. Some key metrics could include a reduction in hospital readmissions through effective medication management, improvements in patient medication adherence rates, and the cost per quality-adjusted life-year (QALY) gained by AI-assisted pharmaceutical services, especially if they support the product-based model. By targeting these metrics, we can lay the groundwork for viable AI-enabled pilot programs that appeal to payers and stakeholders, highlighting the potential for significant impact on healthcare costs and patient outcomes.
To bring these pilot programs to fruition, we can follow a series of practical steps. First, select a primary outcome metric that aligns with organizational goals, such as reducing readmission rates, enhancing medication adherence, or addressing a staff need that has been growing for the past few years. Next, select the AI tools that are most likely to have a positive impact on the chosen metric. Following implementation, systematically measure the impact using predetermined criteria. Involve cross-functional teams to monitor progress and adapt strategies as needed to enhance effectiveness. Finally, compile the outcomes to deliver a comprehensive report that demonstrates the pilot's success and outlines recommendations for scaling the initiatives across broader operations.
Now, many questions remain unanswered, such as how exactly we implement AI tools, the barriers to overcome, what the ‘best’ tools and evidence to look for are, KPIs and metrics to track, and the overall business and economic impact of AI integration. To be honest, some of these aspects are beyond me at this point, which is why I think a broader discussion on this topic should occur across the pharmacy profession.
Ultimately, AI is on the horizon. How and when will vary. Currently, I believe much of it is talk, and developing action will ramp up, especially as we start to see evidence from organizations discussing their successes and failures with AI integration. These are important milestones that will help us determine what to do and how to build upon developing the future of pharmacy. It will come in small iterations over time, but so is the nature of anything in health tech. Here's to the future.