Almost half of all GP Admin Tasks Are Automatable—Why Aren’t We Acting Faster
Two million more GP appointments per month are being delivered since the pandemic as part of the NHS primary care access recovery plan1. However, with more appointments and current processes being manual and time intensive, there is an ever-increasing administrative burden across the whole of primary care. This is putting pressure on the workforce, with over 20% of GP trainees being at a high risk of burnout (2).
This administrative burden is not inevitable as 44% of all general practice administration tasks could be automated using currently available technology (3).
Now more than ever is an appetite, need and opportunity to adopt digital technologies to automate time-consuming yet mundane processes, releasing time for patient care, reducing overtime and improving job satisfaction.
To address the severity of the administrative challenge, the Oxford Internet Institute (University of Oxford) conducted a study to identify the extent of achievable automation3. The study involved focus groups, case studies, interviews and online surveys with 65 clinical and non-clinical primary care staff participating across six GP practices, as well as a survey completed by over 150 automation technology experts. A predictive machine learning model was then used to apply the automation experts’ assessments to primary care tasks.
The model predicted that 44% of administrative tasks are ‘mostly’ or ‘completely’ automatable using currently available technology.
So, from a real-world perspective, what kind of administrative tasks take place in General Practice?
Before a patient’s appointment, a lot of time has already been spent completing administrative tasks whether it be manually triaging a patient at the first point of contact, booking a patient into the clinic itself or re-directing a patient to more appropriate services.
Then comes the clinic itself…so let’s take the example of a 3-hour clinic in a busy GP surgery, made up of 18 patients, the expectation being that each patient has a 10-minute appointment. Within that 10-minute period, excluding the actual patient assessment, the following administrative tasks take place:
Review the clinical notes before and during the appointment to understand the patient’s medical history, drug history, test results, recent clinic letters and previous attendances.
Detailed documentation of the encounter itself including; history, examination, observations, management plan and safety-netting advice.
Request investigations (e.g. blood tests, scans) including the labelling of any samples that were taken during the appointment.
Generate and sign off any prescribed medications.
Complete any referral letters or forms.
Send tasks/messages to colleagues, for example, asking the medical secretaries to chase a previous discharge letter.
To add to this, every patient is unique with their own biopsychosocial complexities and to manage their care safely often involves running over the allocated ’10-minutes’, which if the case for most patients, results in a clinician often working well beyond their 3-hour clinic.
The above only paints a picture of the admin generated during a clinic, but in fact there is additional admin outside of the clinic including:
Review and filing of test results for other patients.
Review and sign-off of prescription requests.
Review and action secondary care letters including medication changes and follow-up instructions.
Complete additional paperwork such as death certificates, insurance forms, and cremation forms.
In the event that there is a medical emergency or ‘lunchtime’ home-visit, it is no surprise that the primary care workforce is experiencing unprecedented levels of stress and burnout. This is not sustainable for staff or for patient care, and with the digital technologies available today it is inexcusable to not be using automation.
Technologies are advancing at a rapid rate and confidence in using them is growing. Within primary care, we are now starting to see new and innovative technologies to help alleviate the administrative workload and although in their infancy, they are yielding some exciting results.
One example is ‘Heidi’ an ambient AI medical scribe that seeks to automate clinical documentation by listening and transcribing, allowing the clinician to focus solely on their patient. Within primary care, having seen first-hand consultations documented using Heidi, they are very thorough.
Automation has a role to play at every point in the patient’s journey and the technology is available NOW, so why haven’t these tools been universally adopted by primary care?
Risk: There is likely to be a concern that these technologies may pose a risk to the patient if they falsely document or interpret information. This misunderstands the role of automation, it is there to help reduce the admin burden on the clinician not replace them. Clinicians will need to review the content generated by the automation, and the overall responsibility still falls with them.
Confidence: ‘AI’ is a buzzword that has become the hot topic of conversation in the media for both its perks, but also its safety concerns. Patient safety is the number one priority and anything that poses even the slightest theoretical risk is going to be a cause for concern. However, any technologies available within the NHS will have been evaluated against strict NHS clinical safety, data protection and cyber security criteria; to ensure patient safety, service quality and regulatory compliance. This is the green light to say that it is safe to use!
Experience: Digital literacy varies from person to person and this lack of experience using digital tools is likely to cause a resistance to change. As with all new ways of working, people need to be supported to familiarise themselves with automation tools and how to adjust their practice to get the full benefit from them.
The administrative burden facing primary care is extensive, but the bigger the challenge, the greater the opportunity for change. With solutions readily available, there is no time like the present to adopt automation! Going forward, the focus needs to be on continuing to develop, test and refine digital technologies to support primary care staff with their profound workload, exploring and alleviating their concerns, with the end result being a rejuvenated workforce and ultimately better patient outcomes.
Author
(1) GMC National Training Survey
(2) Study run by the University of Oxford and published in the British Medical Journal (BMJ).