University Hospitals Coventry and Warwickshire NHS Trust digitizes and improves patient experience with AWS

Giving patients and their families access to the information they need in ways convenient for them is a key focus for one of the UK’s largest National Health Service (NHS) trusts. Like many healthcare providers, University Hospitals Coventry and Warwickshire (UHCW) NHS Trust, which manages two major hospitals and serves a population of more than one million, has operated with legacy technology that relies heavily on phone calls and manual processes for contacting patients. Recognizing an opportunity to modernize, the Trust linked up with IBM Consulting for an innovative pilot project to digitize patient engagement channels using Amazon Web Services (AWS).

Most industries have moved to digital self-service platforms that make life easier for users and improve their satisfaction with services, whilst enabling staff to focus on more complex queries. UHCW NHS Trust is on its own digitization journey and has recognized that it needs to provide more convenient self-service options for patients to improve their experience of its services.

At UHCW NHS Trust, an influx of phone calls and emails created challenges for staff responding to a specific sub-set of patients. The Trust saw an opportunity to quickly test and validate new digital technologies to address this before committing to long-term investments. It chose IBM Consulting as a partner to help with this, based on their healthcare expertise and technical delivery skills on AWS.

This pilot project ties into UHCW NHS Trust’s broader innovation initiative, UHCWi, which focuses on making continuous improvements that put patients first and empowers staff to eliminate waste within their processes and deliver safer care.

Collaborative innovation

IBM Consulting and AWS provided investment and technology for UHCW NHS Trust to pilot new solutions focused on digitizing patient interactions to improve their experience and free-up staff time. Taking an eight-week approach, IBM Consulting used design thinking and the technical knowledge of their Healthcare and Life Sciences team to quickly build initial prototypes on AWS.

The team chose four priority solutions to demonstrate the art of the possible and quantify potential benefits:

1. A virtual assistant (VA) on Amazon Lex to handle frequently asked questions (FAQs) from patients and appointment queries before handing off to a live agent when needed.

2. An SMS text message referral confirmation system using Amazon Pinpoint to notify patients when UHCW NHS Trust receives their GP referral.

3. An automated email responder using Amazon Simple Email Service (Amazon SES) to ask patients to provide missing information so their queries can be resolved more quickly.

4. An automated outpatient appointment reminder and change request system, built using a serverless architecture on AWS. This makes it easier for a specific subset of patients to receive reminders and reschedule appointments – further improving patient satisfaction and clinical outcomes.

This collaborative way of working combined IBM Consulting’s technical and citizen engagement expertise with UHCW NHS Trust’s insights into patient needs to drive quick results. The Trust provided important information about patient journeys and operational challenges, and coordinated access to internal stakeholders and systems.

A closer look at the technology behind the scenes

Although patients only see simple, user-friendly communications, the underlying AWS-based architecture illustrated how multiple services work together to provide intelligent functionality.

Virtual assistant

Amazon Lex powers the VA with automatic speech recognition (ASR) and natural language understanding (NLU) to interpret patient queries. User inputs are matched against a knowledge base to provide standard responses, or trigger fulfillment code to gather more details, before handing off complex requests to a live agent in UHCW NHS Trust’s Patient Access Team. That team is provided with full conversation context through Amazon Connect.

The VA covers FAQs, appointment changes, cancellations, referral status checks, and patient record inquiries. There is also logic to handle errors and notify users if the Trust’s Booking Centre is closed. By using self-service and intelligent routing, the VA aims to reduce the total number of calls that the Patient Access Team needs to handle, reducing operational time and costs.

Referral confirmation

For the SMS text message referral confirmation system, Amazon Pinpoint is used to let patients know when the hospital has received their referral from the GP, with the patient then put on the waiting list for an outpatient appointment.

Amazon Pinpoint offers targeted personalized communication at scale combined with two-way messaging. Automated workflows trigger appointment reminders and notifications while intelligently routing patient replies based on keyword analysis.

As shown in Figure 1, the referral confirmation workflow starts by uploading a contact list to Amazon Simple Storage Service (Amazon S3). AWS Lambda then processes this list to create personalized communication campaigns in Amazon Pinpoint. As Pinpoint sends out text messages, it generates metrics that are captured in real-time by Amazon Kinesis. These metrics are then stored in a data lake within S3. Finally, AWS Glue processes this data, allowing Amazon QuickSight to create insightful dashboards. These dashboards link engagement metrics to operational KPIs, demonstrating quantifiable benefits. Throughout the process, robust access controls ensure patients’ data remains secure and their privacy is protected.

Figure 1. Referral confirmation high-level architecture. The major components are an Amazon Simple Storage Service (Amazon S3) bucket, AWS Lambda, Amazon Pinpoint, Amazon Kinesis, AWS Glue, and Amazon QuickSight.

Email responder

Managing patient appointment changes can be an administrative challenge. Patients often send incomplete information when they ask for appointments to be rescheduled, which means staff have to spend additional time checking details such as date of birth or medical record numbers. This results in delays to patients getting the updates they need.

To improve the effectiveness of email communication with patients, an automated system was designed to analyze incoming emails. This system enables the UHCW NHS Trust team to speed up response rates by providing intelligent email categorization and automated responses to patient emails. More specifically, the system identifies when certain patient identification details are needed to fulfill the patient’s request.

As shown in Figure 2, this system is powered by an AWS Lambda that connects to the Patient Access Team’s email inbox using the IMAP protocol, which allows the automated processing of incoming emails from patients. Then, an AWS Step Functions workflow uses Amazon Comprehend to analyze the content of these and classify them into a set list of categories. Amazon Comprehend also determines whether the email contains sufficient information for the Patient Access Team to process the request or more details need to be collected from the patient. Finally, Lambda functions sort messages based on the analysis results and automatically send personalized replies through Amazon SES to request missing details to fulfill the patient’s request.

Figure 2. Email autoresponder high-level architecture.

As outlined in Figure 3, the system detects the email’s language, categorizes it, extracts and analyzes relevant information like names and addresses, and then sends an automated response. This process is designed to efficiently handle various types of incoming emails, including appointment-related queries and general inquiries, while also identifying urgent requests.

Figure 3. Email responder classifier workflow.

This automated approach simplifies the patient experience by immediately asking for additional information from the patient. This results in the Patient Access Team having to send fewer follow-up emails requesting additional information from patients and allows them to process patient requests more quickly.

Appointment reminder

Giving patients the ability to manage their medical appointments more conveniently improves their satisfaction with services and clinical outcomes. The solution implemented an automated appointment reminder and change request system using a serverless architecture on AWS.

As shown in Figure 4, the system uploads patient contact data and upcoming appointments to Amazon S3. This triggers a Lambda function to move the data into an Amazon DynamoDB table. Then the solution uses Amazon Pinpoint to automatically send text reminders to patients about their upcoming appointments. Patients can easily request a change by replying to the text. This invokes a Lambda function to generate a unique link for patients to select a new time and submit the request to the Patient Access Team. Appointment change requests are handled by Amazon API Gateway and Lambda, saving information to the DynamoDB database. A daily Amazon EventBridge job compiles the change requests into an Excel file for the Patient Access Team to review and approve.

This fully automated, serverless system makes it effortless for patients to receive appointment reminders and request changes. By using AWS services such as Amazon Pinpoint, Lambda, API Gateway, and DynamoDB streams, the system enhanced the patient experience while increasing operational efficiency.

Figure 4. Appointment reminder high-level architecture.

Significant progress

In only eight weeks, UHCW NHS Trust and IBM Consulting delivered the virtual assistant and SMS solutions, which immediately began helping patients. The email responder and phone upgrades were created as interactive prototypes to showcase future capabilities and design patterns.

During this period, the new solutions dealt with 3,500 patient interactions, reducing calls to the booking centre by an estimated 33 percent, and sent more than 10,000 referral text messages. This translated to forecasted annual savings of more than £270,000 from prevented calls alone. After seeing the early results, Professor Andy Hardy, CEO of UHCW NHS Trust, said, “Working with AWS has enabled us to combine process research and cutting-edge technology to identify areas we can further improve patient experiences and outcomes. This offers massive potential for the future, and we are excited to see what we may be able to achieve in the future.”

The fast path to healthcare modernization

UHCW NHS Trust’s approach demonstrates the speed of cloud-based IT transformation, with a focused investment on rapid prototyping of solutions directly targeting patient needs. Within two months, real systems were already transforming the patient experience and generating a return on investment. More importantly, the technical foundations have been established on AWS and the Trust now has validated designs which it could use to guide future expansion of these pilot efforts across more types of patient interactions. What started as a short-term, experimental endeavour has evolved into a long-term modernization initiative with the potential to digitally transform patient engagement.

Learn more about how healthcare and life science organizations can accelerate innovation with AWS on the AWS for Healthcare and Life Sciences homepage.