Automated waiting list validation powered by AI

EBO developed an AI-driven waiting list validation tool for Betsi Cadwaladr University Health Board (BCUHB) to validate patient waiting lists in a cost-effective manner at scale.

Betsi Cadwaladr University Health Board

Betsi Cadwaladr University Health Board (BCUHB) is the largest Health Board in Wales employing over 19,000 staff. It delivers primary care, hospital, community, and mental health care services to more than 700,000 people in North Wales.

The local IT infrastructure consists of multiple national systems provided by Digital Health and Care Wales (DHCW), such as the Welsh Patient Administration System (WPAS), integrated with bespoke systems to provide a cohesive local eco-system.

Due to the impact of the COVID-19 Pandemic on NHS services, like many other NHS organisations, BCUHB faced the challenge of unprecedented waits for planned care services. As a result, in April 2022, the Welsh Government published ‘Our programme for transforming and modernising planned care and reducing waiting lists in Wales’, which set out the expectation for all Health Boards to constantly review and validate their waiting lists “to ensure that the list is accurate and up to date through administrative validation”.

Summary of outcomes

0 %
Response rate
0 %
Conversation completion rate
%
Patient satisfaction rate
£ k
Estimated savings per cycle

The Case for Change

To ensure waiting lists are accurate, patient validation needs to take place every 3-months during a patient’s wait. This will ensure that patients that need to be seen, receive an appointment, or procedure as quickly as possible and according to clinical need. Through validation, patients can tell BCUHB whether they still need an appointment.

The current patient validation activity is being delivered using a call centre model. This activity is being undertaken in tranches and often requires the services of external companies to call patients and record their responses in the WPAS.

This process presents a number of challenges:

  • Costly one-off validation exercises
  • Limited number of contacts during the patients’ wait
  • Poor patient feedback and experience
  • Quality of the calls and recording of responses difficult to standardise
  • External suppliers’ numbers are not always recognised and are sometimes reported as fraudulent
  • Tranche-based validation means manual updating of WPAS can be delayed and inefficient, resulting in poor appointment utilisation and impacting waiting times
  • Difficult to manage the process across the call centre and internal speciality teams, often with multiple users updating Excel spreadsheets resulting in poor data quality and auditability

[EBO] more than rose to the challenge. It felt like a true development partnership with staff from both EBO and BCU forming a cohesive team

Danielle Edwards

Danielle-Edwards

The Opportunity

Working in partnership with EBO, BCUHB developed and undertook a proof of concept (POC) of the EBO Waiting List Validation Solution which, through integration with WPAS, would allow BCU to:

  • Replace up to 76% of costly one-off validation call tranches with automated and continuous digital patient validation
  • Initiate contact with patients via SMS directing them to a virtual conversation at a time convenient for them
  • Capture the responses for the remaining patients that are contacted by telephone within the same database as the virtual responses
  • Automatically update the WPAS to reduce administrative costs and provide equitable high standards of communication with patients
  • Create an enriched patient experience and digital-first uptake, by signposting patients to key information and services via the Virtual Assistant

Deployment

The POC project was delivered over two phases. In Phase 1, BCUHB wanted to understand whether patients were willing to engage with a Virtual Assistant. Phase 2 focused on making improvements in line with patient feedback, developing the waiting list view and reporting dashboards for staff.
 

PHASE 1

Assessing Patient Appetite to Engage with a Virtual Assistant

Phase 1 included 685 Ear Nose Throat (ENT) patients with mobile numbers in WPAS that were:

  • Stage 1 Category (waiting for a first appointment)
  • Had been waiting between 26-51 weeks for an appointment
  • Un-booked
  • Un-validated in the previous 3 months
  • 16years old and over

PHASE 2

Development of the EBO Dashboard

The focus for Phase 2 was to develop the product dashboards and reporting. An additional 448 ENT patients were validated in Phase 2. This phase focused on the development of EBO dashboards for monitoring and reporting. Improvements were made to the EBO Solution following Phase 1 that were then deployed in Phase 2. These included:

  • Improved two-factor authentication
  • Concatenated text messages 
  • Detailed Q&A on the Virtual Assistant landing page
  • New dashboards and reports

Evaluation

The POC project was delivered over two phases. In Phase 1, BCUHB wanted to understand whether patients were willing to engage with a Virtual Assistant. Phase 2 focused on making improvements in line with patient feedback, developing the waiting list view and reporting dashboards for staff.
 

Removal rates

In Phase 1, 6% of patients were removed from the waiting list because they did not require an appointment. Similarly in Phase 2, 7% of patients responded saying they did not require an appointment. The breakdown for all responses for each phase is shown figure 1.

Patient feedback​

Patient feedback was requested at the end of each conversation with patients choosing between a score of like, neutral, or dislike. Patients were also invited to leave a comment to explain the reason for their score.

Key Benefits

The main benefits achieved from the POC cohorts for patients and the Health Board were:

Benefits to Patients

Benefits to Patients

Extrapolating the 4% average removal rate for the POC to the 18,495 patients on BCUHB’s current waiting list for first appointments (waiting over 36 weeks) at the time of the evaluation, the potantial is

750 appointments freed up with circa £44k DNA avoidance

Compared to the current external call centre validation model the potential system efficiency for 1 validation cycle could be in the region of £137K

Next Steps

The POC project was delivered over two phases. In Phase 1, BCUHB wanted to understand whether patients were willing to engage with a Virtual Assistant. Phase 2 focused on making improvements in line with patient feedback, developing the waiting list view and reporting dashboards for staff.
 

PHASE 3

Continuous Validation with Bi-directional Integration to the WPAS (Welsh EPR)

This phase will focus on deployment of the solution in BCUHB.
The solution enhancements planned for Phase 3 will allow:

  • Automated upload of data from WPAS to EBO via integration
  • Speciality clinical review processes agreed and reflected in workflow
  • Continuous trigger built into EBO database
  • Bi-directional integration to auto update WPAS
  • Handling of patients on multiple pathways
  • Full audit trail and report suit

Conclusion

The POC project was delivered over two phases. In Phase 1, BCUHB wanted to understand whether patients were willing to engage with a Virtual Assistant. Phase 2 focused on making improvements in line with patient feedback, developing the waiting list view and reporting dashboards for staff.
 

Trusted by leading healthcare organisations

Back to top