As a society, we increasingly rely on digital technologies in most aspects of our life, such as social media and online banking. These technologies have had a significant impact on our personal and business interactions.
However, what can be less obvious is the extent to which digital technologies underpin critical services, whose failure can lead to human harm. These ‘safety-critical systems’ have traditionally been dominated by aviation, rail and nuclear power generation; all industries with an impressive track record in achieving high levels of safety.
As a researcher in safety engineering, after joining academia from industry, I have had the opportunity to work closely with leading engineering organisations such as Rolls-Royce, NASA and Jaguar Land Rover. I thought that such domains provided the ultimate challenge for an engineer. However, this all changed when I discovered healthcare!
The complex clinical environment and the multidisciplinary nature of healthcare services, blending of technical, personal and organisational aspects, fascinated me. I saw healthcare as a highly adaptive system, where some sources of uncertainty and variation are necessary to treat patients with different health conditions, personal preferences and social situations.
At the heart of healthcare is the need to manage and communicate the right information to the right people, at the right time and in the right format. This is increasingly being enabled through digital information technologies.
These technologies have become critical to healthcare, for example, transferring and maintaining electronic health records and electronic prescribing. Recently, the healthcare landscape has expanded with the use of mobile health apps, empowering patients to take a more active role in their care.
Patient safety is a fundamental concern, though. While these new technologies can potentially improve patient safety, they may also introduce new hazards. Electronic prescribing can reduce transcription errors (due to doctors’ famously challenging handwriting!), but may also increase the risk of ‘alert fatigue’, with patients becoming desensitised to the urgency of alerts.
In another example, a recent assessment of 46 mobile health apps for calculating insulin dosage concluded that the majority did not include protection mechanisms against incorrect dose recommendations, putting users at risk of potentially-fatal overdoses.
Engineering a solution
Two years ago, I was fortunate to receive a Royal Academy of Engineering Industrial Fellowship, which funded a collaborative project with NHS Digital on patient safety and digital technologies. This meant that I was able to explore healthcare from the inside and interact with clinicians and engineers from different NHS trusts and technology firms.
Working with NHS Digital, and jointly reviewing the safety evidence of different national and local health information systems, we discovered that the dynamic characteristic of healthcare, combined with insufficient funding, has made it difficult for organisations to generate the safety evidence to the required level of rigour and clarity. A primary contributory factor was the lack of a common language and sufficient engagement between the clinicians who use the technology and the engineers who provide it.
In collaboration with colleagues at the University of York and NHS Digital, I developed the Safety Modelling, Analysis and Reporting Toolset (SMART). SMART is a partially automated technique that models both what the digital health technology does and the care pathways in which the technology is used. For example, SMART can be used in diagnosis or treatment, whether in a hospital or in a GP surgery.
The resulting models allowed clinicians and engineers to work together to identify hazards such as how patient safety can be compromised. They also paved the way for reasoning about how the risk of these hazards can be communicated, assessed, and mitigated.
Engineering brings structure, clarity and simplicity through modelling. SMART has brought these principles to the complex and critical care settings in which digital technologies are used. SMART is currently being trialled by a number of NHS trusts and technology firms, and NHS Digital has committed to work with me to develop it further and promote its use across the country.
My fellowship project has ended, but my journey is still in its early stages. I am now exploring the benefits and challenges to patient safety from wearable devices and artificial intelligence. Being embedded within the NHS has been a privilege and a hugely rewarding experience, which highlighted the direct benefits of engineering for assuring patient safety and improving the quality of care.
Image credit: NEC Corporation of America