AWS for Industries
Pharmacy in the Cloud: AWS transforms medication safety and operations
Pharmacists and pharmacies are on the front lines of ensuring the right medications reach the right patients safely. By digitizing and streamlining pharmacy workflows, operational efficiency and safety are improved. Consequently, it also empowers pharmacists to focus more time on patient-centered activities (such as clinical decision making, patient counselling and safety oversight). Digitalization enhances the central role in delivering high-quality healthcare and safeguarding patients’ well-being.
We will take you through a medication’s journey within a pharmacy—from prescription processing to dispensing and monitoring. We will also highlight key technology intervention points where cloud innovation is transforming pharmacies.
Prescription processing
Prescription processing refers to the steps involved in managing a prescription including:
- Receiving a prescription and checking for its completeness
- Reviewing the prescription for information, such as potential drug interactions
- Entering the prescription details into the pharmacy management system
Prescription processing may be a bottleneck step, especially if prescriptions are received in handwritten format, fax or entered as free text in the system. These formats require manual entry into the pharmacy management system.
Amazon Pharmacy processes large volumes of prescriptions each day and utilizes over 50 services powered by Amazon Web Services (AWS). Amazon Pharmacy uses AWS HIPAA-eligible services (such as Amazon Bedrock and Amazon Comprehend Medical) to automate prescription processing and support Pharmacists at Amazon Pharmacy in streamlining their work. The services can identify prescription images from documents received by fax and extract text from the prescription images. They can then associate the extracted health data with known drug properties (such as drug ingredients, strength and dosage form) and identify the drug from the Amazon Pharmacy drug catalog.
The prescription notes are converted into a machine-readable format and streamlined human readable instructions. For example, the original copy of 1 tab PO 8am and 1 tab PO 8pm becomes translated to the human instructions of Take 1 tablet by mouth twice daily at 8 am and 8 pm. The instructions are also transformed for machine readability where dose per day = 2, and the dose times = 0800 and 2000.
By leveraging AWS services to automate prescription processing, Amazon Pharmacy has significantly scaled its manual operations efficiently, reduced costs, improved speed, and minimized prescription errors.
Pricing transparency
Current medication pricing is largely hidden until point-of-sale and patients often discover costs only at checkout. This can contribute to diminished trust in the healthcare system, and reduced adherence to prescribed therapies. Patients may even delay or not take medications when confronted with prices they did not anticipate.
Using HIPAA-eligible generative AI and machine learning (ML) tools and services on AWS, Amazon Pharmacy is providing its customers with greater price transparency. In 2023, Amazon Pharmacy began offering insurance pricing estimates using deep learning statistical models hosted on Amazon SageMaker. Amazon SageMaker is a service where data scientists and ML engineers can build, train, and deploy ML models for virtually any use case with fully managed infrastructure, tools, and workflows.
The price estimates help customers to make more informed decisions. Once a customer’s prescription is sent to Amazon Pharmacy, the customer can access clear, transparent, personalized insurance and cash pay options. Customers can also view discounts available through PrimeRx, and other potential ways to save before checkout.
The outcome is that operational efficiency has been optimized and 99% of prescriptions include up-front pricing estimates.
Customer care: Clinical customer agent
Pharmacy staff often have to break their flow of work to answer general questions regarding inventory or billing from patients, either in person or through the phone. To reduce interruptions and enhance operational efficiency, Amazon Pharmacy built the Clinical Customer Agent, an agentic AI solution that integrates several AWS services to support such queries.
When a customer calls, a customer care representative queries the Clinical Customer Agent, which uses the natural language processing capabilities of Amazon Bedrock to understand the customer’s question. The Clinical Customer Agent then accesses the relevant Amazon Pharmacy systems and knowledge bases in real-time to gather the necessary information.
The Clinical Customer Agent has unified access to prescription and order status, standard operating procedures (SOPs) and knowledge base content. The agent application runs on the scalable compute power of AWS Fargate for Amazon ECS, while Amazon CloudFront delivers the intuitive agent interface. Customer feedback is stored in Amazon DynamoDB, enabling continuous improvement of the solution. Amazon Simple Storage Service (Amazon S3) manages the session data by providing consistent experience for each customer.
The Clinical Customer Agent has resulted in a significant reduction in the average handling time of queries by customer care representatives. The time savings have reduced operational costs and consequently reduced the cost to fill a prescription at Amazon Pharmacy.
Medication adherence
According to the WHO report Adherence to long-term therapies, about 50% of patients in developed countries follow long-term treatments for chronic illnesses, with rates even lower in developing countries. Medication non-adherence contributes to disease progression, increased mortality and, at the healthcare system level, it leads to increased hospitalization and higher healthcare costs.
Causes of medication non-adherence include factors such as:
- Cost
- Side effects
- Forgetfulness
- Complex regimen
- Misunderstanding of instructions
- Patient beliefs
To address these challenges, MedAdhere AI by Mindbowser Inc. (powered by AWS) combines:
- AI-powered personalized, multi-channel reminders (SMS, voice, app notifications) and automated follow-ups for patients to take medications
- Predictive analytics and actionable dashboards to identify patients at high risk of non-adherence using real-time data
- Seamless integration with electronic health records (EHR), remote patient monitoring platforms, and AWS-powered data lakes for a unified patient view
With the support of AWS services, MedAdhere AI verifies patients stay on track with treatments, reduces hospital readmissions and strengthens provider-patient communication.
Inventory management
Inventory management in pharmacies is challenging due to the need to balance stock levels to avoid shortages or waste. This is needed with unpredictable demand and especially with perishable medications or proper storage of temperature-sensitive medications. Adopting digital, data-driven systems is key to improving visibility and control across the supply chain.
Alpro Pharmacy, Malaysia’s largest prescription pharmacy, experienced a 400% surge in demand for its product and services. Its basic inventory and on-premises point of sales system could not keep up with the demand, leading to inaccurate inventory reporting, product stock outs and long waiting times for patients.
To transform its operations, Alpro Pharmacy engaged with AWS Partner, QR Retail Automation (Asia) Sdn Bhd (QRRA), which developed QR AgoraCloud on AWS, a solution that is equipped with tools to help customers manage online, offline and back-office operations. Alpro Pharmacy, QRRA and AWS collaborated to develop a solution for Alpro Pharmacy that would automate its warehouse operations and streamline its supply chain with a centralized inventory system across 200 stores.
As a result, Alpro Pharmacy can fulfill orders on the same day, within a two-hour target. Order fulfillment volumes can now handle increases from 500 to 4000 each day and have provided an increase in its picker output by 80%, while reducing its operational cost. The processing and delivering of prescriptions are faster and more accurate. Alpro Pharmacy pharmacists can make better decisions about medication management and provide personalized care to patients.
Labor planning
While increased use of automation and digital tools are alleviating the issue of labor shortages in the pharmacy workforce, staffing challenges remain significant. Labor planning is critical to ensure efficient and safe operations of the pharmacy.
Amazon Pharmacy has adopted the Demand Planning application of AWS Supply Chain. Information such as order volume and form factor (packet, vial unit of use) are input into a model to generate a demand plan output. It can be used to plan daily capacity at each fulfillment site. Understanding the daily capacity, labor planners can schedule the appropriate staff for each site every day.
Prior to AWS Supply Chain, different teams managed separate forecasting processes, which resulted in longer cycle times, higher resource needs, and limited data granularity. With AWS Supply Chain, Amazon Pharmacy has saved approximately 13% of weekly planning time through reduced manual efforts. They have also achieved 50% better forecasting accuracy than the industry standard mean absolute percent error (MAPE) target.
Pharmacy management systems on AWS
HealthTechs globally build their solutions on the cloud because they are able to scale solutions faster, while experiencing better availability and higher security. AWS Partner, Pharmony, created Pharmony One, an all-in-one cloud-based pharmacy management system (PMS) on AWS. Pharmacists can redirect the time spent managing time-consuming server and hardware updates to focus on their role as healthcare advisors. Pharmony One is delivered as a software-as-a-service (SaaS), meaning pharmacies can access it through the Internet without having to host their own server.
The key benefits of Pharmony One are:
- It centralizes wholesalers ordering, inventory tracking and product search through integrated databases
- Offers full traceability of stock and is compliant with data sovereignty regulations in Europe
- Suitable for single pharmacies or pharmacy chains, as it allows for group orders
- Lowers infrastructure costs from savings due to less in-house servers and hardware, while enjoying higher data security since no data can be robbed onsite
- Its interfaces are designed for different devices and companion mobile apps for on-the-go access.
With the Africa-based HealthTech proprietary web app, Bloom by mPharma, pharmacists can access real-time drug information, treatment guidelines, and patient medication history. This gives them the ability to provide more informed consultations and personalized care. In addition, Bloom facilitates better communication between pharmacists and patients.
Patients can refill prescriptions, schedule appointments, and receive medication reminders through a user-friendly web interface. This is more convenient and also helps improve patient adherence to treatment plans, leading to better health outcomes. Since adopting AWS, mPharma has grown from serving a single pharmacy in Accra to a network of over 500 pharmacies in Ghana, Nigeria, Kenya, Uganda, Zambia, Ethiopia, Rwanda, Togo, and Benin.
The firm’s vendor management inventory (VMI) system is used in over 500 pharmacies in these markets to reduce risks in medicine stocking. AWS enabled Bloom to scale smoothly without significant upfront infrastructure investments.
Conclusion
Through various examples we have shown how AWS is helping pharmacy customers around the world enhance efficiency, strengthen medication safety, and accelerate innovation. As cloud technology, AI, and data-driven insights continue to evolve, the opportunities for transforming pharmacy operations and improving patient care are virtually limitless.
Contact your account representatives or an AWS Representative to learn how we can help solve your unique pharmacy challenges.
We would like to acknowledge and thank Alexandre Alves, former Sr. Principal Engineer for Amazon Pharmacy, for his contribution to the content of this blog.