Going hands-on with Codepilot in a real practice

January 6, 2022

As a GP with partner experience I resonate all too well with the sinking feeling that comes around this time every year when our Practice Manager hands us a list of patients to review in advance of the QOF deadline in March. The process of reviewing these patient records is tedious, demoralising and at the end of each opportunistic push between sessions we don’t know how much contribution we have made to practice finances or patient care. I love being a GP, but one thing is very certain: this is by far my least favourite part of the job.

For this reason, I was eager to see if Codepilot could alleviate this burden in a real-world benchmarking exercise. In doing this, I first recorded the time spent reviewing patient records the old fashioned way. Over the course of a month, I checked 96 records ‘manually’ for coding mistakes and oversights across 9 clinical domains. The list of the records to review was generated from tightly defined queries constructed by our Practice Manager, and represents a best-practice approach to tackling the task. 

I found that, on average, I spent 7.7 minutes on each record, though that varied somewhat by clinical domain. I then switched to using the Case Solver functionality of Codepilot. This tool ingeniously presents a curated summary of relevant information from each patient record alongside the pertinent QOF guidelines and recommendations for changes. 

For example, consider the patient record shown below (note that while this benchmarking exercise was performed on real patient data within my practice, this example is mock data for illustrative purposes only). This patient was surfaced as being potentially wrongfully missing from the diabetes register. Conventionally, I would have to trawl through consultation notes, letters, problem lists, prescriptions and investigations. If after all that it transpires the patient is indeed diabetic, I then need to remind myself of the active/qualifying QOF codes through SNOMED and add these codes to the patient's record.

Click to enlarge (opens new tab)

When using Codepilot, I am advised by its AI-generated suggestion to look straight at the Metformin prescription issued on 11/05/2021. I have saved time by going straight to the relevant consultation rather than trawling through the entire record. Having confirmed that the patient is diabetic, Codepilot highlights the correct codes that will include this patient on the diabetes register. I can see exactly why a record has been flagged and the correct codes are presented to me. 

By measuring the time taken to review 70 records across the 9 clinical domains using Codepilot, I found that the average time spent per record was reduced to 3.7 minutes - approximately half the time spent without the tool, or to put it another way, doubling my productivity. A full comparison of my time spent per record both with and without Codepilot is broken down by clinical domain in the graph below.

Click to enlarge (opens new tab)

As Medical Director for Codepilot, I have had the privilege of being one of the first to go hands-on with the software in a real-world setting. As a GP, the outcomes of this exercise were hugely satisfying as the software really does make life measurably easier both for my peers and me. The feedback I have received from fellow partners at other Codepilot pilot surgeries has been overwhelmingly positive - in particular, the support team has been complimented on being refreshingly responsive and professional, and the tool itself for being intuitive and easy to use.

There is no denying that we are well overdue for a technological overhaul in General Practice. The ever-changing demands on us all to do more in less time are simply not feasible if we continue to use antiquated software and rely on the ‘if it ain’t broke’ mentality. Codepilot presents a glimpse of what the future should look like on our desktops, freeing us up to spend more time looking after our patients and doing what we're trained to do.

If any colleagues would like more information or to discuss how Codepilot might benefit your practice, please feel free to email me personally at: Oli.Maunsell@codepilot.app

Dr Oli Maunsell is Medical Director for Codepilot, a GP at Ringmead Medical Practice in Berkshire, and a GP Trainer in the Oxford Deanery.