As Chief Digital Officer of NHS Blood and Transplant, Aaron Powell is in the happy position of not just helping the service save lives but also adding functionality that allows the service to extend its life-saving capabilities.
Take NHSBT's cloud-based organ donation and transplantation hub, which manages organ donation and transplant offers. In the past 12 months working with IBM, Powell and his team added lungs and livers to its support for hearts – a move anticipated to save an additional 50 lives a year. Likewise, the launch of a digital transplant patient registration portal in the hub now allows clinicians to access and maintain a real-time transplant waiting list; previously, clinicians went around with a printed copy of the current waiting list in their back pocket.
In the blood systems programme, Powell has used the customer interaction elements of a CRM tool to automate communications with donors, including the extremely well-received SMS 'thank you' messages when blood is issued to hospitals. The CRM also lays the foundation for issuing flexible, personalised invitations to donate, matching the frequency of donation to an individual donor's physiology, while ensuring the right donors with the right blood groups are attending to match the demand for different blood groups in hospitals.
Powell rejects the idea of top-down digital transformation in favour of a cascading snowball model. His idea is to start small, develop and prove a capability, and then extend and apply it to new challenges. As clinical colleagues see the success that others have in identifying, deploying and extending innovations, they copy the approach. In the process a critical mass is created, with the entire organisation undergoing digital transformation.
His big innovation focus in the past year has been a series of Big Data and machine learning trials with partners to build confidence among the clinical and scientific communities in the real value of data and machine learning. One trial has focused on donor behaviour (whether external data such as weather and sporting events affect donor attendance); with all donation sessions currently overbooked by a standard percentage – and often overrunning as a result – its value is obvious.
Another trial has seen the creation of a neural network to develop personalised predictions of waiting times for a kidney transplant, which has improved predictability by 7%. Meanwhile a trial to integrate genomic data with other data seeks to identify the relative safety of donation for individual blood donors and identify the 'best' and 'worst' likely donors for current demand.