The practice serves an elderly population, often in need of home visits. Our clinics are fit to burst, with complexity, frailty, mental health and social crises, sometimes all manifesting in one person. One evening, as I opened up my Docman screen to the reality of 141 unprocessed documents, it occurred to me the tasks ahead would test my reading and task sending skills, but little of my training as a GP.
Nonetheless it was critical I did not miss something of clinical importance, hidden within the swathes of administrative information. Decision fatigue negatively impacts on quality decision-making and I was left with a persistent thought of ‘If I know a significant number of these letters require no clinical input, why am I doing them?’
Moreover, a discussion with colleagues elsewhere highlighted that as GPs we vary markedly in the amount of time we each spend processing the same amount of administrative work – suggesting IT skills and risk perception were influencing how we dealt with the work, adding pressure to decision making rather than helping.
What we did
In Brighton and Hove we were involved in an Access Fund pilot, looking at testing new ways of working, which included a little understood programme called Workflow Optimisation.
We met with other partners in the area involved in the pilot to discuss how we might tackle our increasing administrative burden, using an administrator-led approach.
We agreed that we would train up administrators at each practice to read, code and action letters, passing on to GPs the ones that required clinical input. We posited that this could not only be safer and quicker, freeing up valuable GP time for clinical work, but could also help us to improve how we capture data for QOF and other quality purposes.
Our practice sent a member of the administrative team, who had strong working knowledge of the clinical system, and a history of note summarising to attend training courses at Here (formerly Brighton and Hove Integrated Care Service) who were leading the pilot. The course supported the administrator in developing their read code knowledge, and also taught a structured approach to specific high-volume, low-risk letter types. They learned where the areas of risk were, and a standard way to interact with a letter to ensure that all the key actions, read codes, and medication changes were detected and enacted.
We also nominated a GP to act as a Workflow Lead, responsible for supporting the administrator and collating feedback from partners. The training gave a suggested implementation programme, of specific letter types each week. This meant the start was phased which was meant us GPs felt in control of this change in how we work. It also meant the administrator had time to develop their confidence with the lower risk letter types.
The biggest challenge was ‘letting go’: putting our trust in our colleagues and handing over control of letters and other documents to them. Also there was no consistent approach to handling incoming letters so we had to develop a defined protocol that our Workflow team would adopt and implement.
Another was investment – in order for the Workflow changes to be successful, we had to increase the administrators’ hours to allow for the new work to be done, and support their development by enabling them to participate in the training and upskilling. The investment in monetary terms was 0.6 whole time equivalent administrator for a practice of 11,250 patients.
By six weeks we had reduced the number of letters each GP saw every day by around 50%, and by 18 weeks we had 80% of incoming correspondence processed in its entirety by our administrative team. We estimate this has released 40 minutes, each day, for every GP. Not only have we released clinical time, but we have sped up how long a letter spends in our system, and also enhanced our data quality. We now have a trained team, with dedicated time, capturing pertinent clinical information in a timely and consistent way.
As a consequence of the robust approach to data collection from incoming correspondence, we had our best QOF year last year, scoring 97.1%. This happened because we were capturing data from letters first time, using a template within the clinical system designed specifically with incoming correspondence in mind. The template was populated with common data entry points, using QOF and locally commissioned services read codes to maximise our QOF and LCS claims. We found our administrator was able to process letters faster, and capture data reliably and in domains that enhanced our claiming activity.
A small number of brief, targeted audits were undertaken by the Workflow GP lead, which demonstrated the administrators’ competency at handling the data and directing letters appropriately. The approach is incorporated in routine practice indemnity by our medical defence provider and we had no increase in premium in undertaking the Workflow programme.
We were one of the first group practices to take on the Workflow training; an evaluation of 63 practices in the original pilots around the country demonstrated sustained, safe impact and the programme has been included in the GP Forward View (GPFV) with £45 million ring fenced funding to allow every practice in England access to train and implement it.
The most recent evaluation, undertaken after training in excess of 800 practices, shows a full time GP reduces their administrative burden by 120 hours a year.
We have recently merged practices and are bringing in the workflow model with our newly expanded team, to ensure that we gain the maximum benefit from the merger. I would highly recommend that practices contacted their local Primary Care Commissioners or LMC to discuss accessing the GPFV funding for this programme of work. It is such a simple idea, but makes a real difference to everyday practice and helps to lighten the administrative load. It allows staff to be trained, provides a governance structure, and ensures it is implemented in a safe but effective way.
Dr Paul Deffley is a GP in East Sussex