NHS General Practice in England faces multiple challenges as evidenced from multiple national organisations, policy groups, think tanks, and universities. Some of these challenges include skill shortages, staff retention, increased workloads, and financial burdens. Various strategies have been proposed to ameliorate these challenges, from increasing financial investments and renewing infrastructure to reorganising and redesigning staff. There is certainly no single solution that will create a robust, sustainable, and high performing general practice of the future. The research program detailed in this report outlines the development and interpretation of an evidence base to inform the conversation about the use of technological support within primary care. Specifically, to what extent can currently available automation technologies be applied in general practice? And, what kinds of work conducted by which type of staff are most amiable to automation? By understanding the scope of automation and the location of where to best apply potential technologies, primary care organisations, policy makers, and industry can make an informed decision on work digitisation and workforce transformation.
We employ a machine learning methodology and expert survey estimates of the state-of-the-art in automation technologies, in order to model the extent to which tasks can be technically automated within primary care today. To improve the robustness of this analysis, our data describing tasks and work in primary care is based on ethnographic observations at six primary care health centres in England, involving over 350 hours spent in understanding the scope of work for each staff type commonly found in primary care. From this analysis we develop an “automatability score” to identify the specific healthcare tasks that have the potential to be mostly or completely replaced by automation techniques. We further describe which skills, knowledge and ability characteristics are likely to complement such work.
Our principal finding is that approximately 44% of all administrative and clerical tasks in general practice can be either mostly or completely automated. We are excited by this finding because it shows the potential for rethinking the future of primary care, and also the potential (beneficial) impact of appropriately implemented technologies on the workforce. We also find that no single occupation can be entirely automated, but that different staff types will be able to have more or less of their work automated. This is another exciting finding because it provides evidence to consider how occupations may be redesigned: from considering new ways of working to adding tasks that bring more value to the patient and the community. Lastly, we focus on a few select exemplar tasks that illustrate the need to carefully consider and think through the implications of automating work in general practice. For example, new ways of working with a machine may necessitate new ways of training and educating staff. The introduction of technology into organisations will always have intended and unintended consequences; we try to highlight and anticipate the former throughout our investigation.
Throughout the project we have found that the automation technologies that promise to make a positive improvement in the workforce are those which automate mundane tasks. Automating and removing some of these tasks could free up countless hours where more urgent, rewarding, or patient-centred tasks can take their place. Automation technologies that assist with tasks such as telephone answering, letter writing, document scanning, email monitoring, and information filtering show the most potential, and would represent a remarkable change to the way all staff work in primary care. However, we note that the work in general practice, much of which seems mundane and routine, is full of complexities and exceptions. We hope that adopting a cautious and informed approach to task automation in general practice will allow for greater opportunities for mentorship, communication, multi-professional training, and increase opportunities to experiment with new ideas to improve patient care. At a minimum we believe reconsidering how the tasks we have highlighted are performed, and by whom, could contribute to an effective and efficient use of resources in primary care.
Willis, M., Duckworth, P., Graham, L., Osborne, M., Meyer, E.T., Coulter, A. (2020). The Future
Of Healthcare: Computerization, Automation, and General Practice Services. Retrieved from: https://bit.ly/foh-report-jun2020