Tuesday 15 April 2025

A 900 Document Backlog to Same-Day Processing: How NHS Practices Transformed Their Workflows with AI

A 900 Document Backlog to Same-Day Processing: How NHS Practices Transformed Their Workflows with AI

Managing the flood of clinical documents is a daily struggle for GP surgeries across the UK, often resulting in backlogs, inefficiencies, and risks to patient safety.

We spoke with Uzo Chukwunonye, Practice Manager for Lawson and Spring Hill Practices (combined 34,000 patients), about how they transformed their document workflows using Anima.

The Challenge: Drowning in Documents

Before implementing Anima, Lawson and Spring Hill Practice faced several critical challenges that will sound familiar to many:

  • Overwhelming volume and backlog: 150-200 documents daily arriving through multiple routes, reached close to 1000 documents of backlog at times

  • Duplicate processing: Documents arriving through mesh, post, and email, requiring manual checking.

  • Inconsistent processing: Staff members process documents differently.

  • Clinical safety concerns: Having to allocate uncoded, unsummarized documents directly to clinicians.

  • Staffing challenges: Four full-time clinical coders still unable to manage the workload, along with difficulties in finding and retaining skilled coders.

Uzo described the situation: "at times we'll see 900 to 1000 documents just waiting to be processed, it was mostly driven by sickness absence. This posed a lot of issues especially with clinical safety".

The Transformation: Implementing Anima's Solution

Seeking a solution, Uzo's team discovered Anima at a conference. While initially skeptical, the potential to address their burning pains and create system stability led them to trial the software. Key steps in their implementation included:

  • Unified Input: Consolidated all document sources in Anima to prevent duplicate systems.

  • Verified AI: Manually checked AI outputs against source documents to build trust.

  • Phased Workflow: Limited initial Anima access to coders (review, process, save to EMIS, task via EMIS), requiring clinician adaptation.

  • Rapid Iteration: Close collaboration with Anima led to quick feedback implementation and customisations.

The Results: Measurable Improvements & Reduced Burden

After implementing Anima in April 2024, the practices saw significant benefits:

  • Processing time per document: Reduced from 4 minutes 24 seconds to just 2 minutes 3 seconds after a three-month trial. The decreasing trend continues post trial to the record month being sub 1 minute.

Document processing time in minutes by quarter

  • End-of-day backlog: Reduced from 800-900 documents to as low as 20.

  • Cost efficiency: Estimated yearly cost of £19,301 compared to £30,800 for a full-time coder.

Uzo summarised the impact: "the efficiency of our coding operation you know vastly improved".

The Human Element: Staff Adoption and Support:

Technology adoption is always a journey. Initially skeptical about AI, Uzo's team addressed concerns by introducing the platform in a step-wise deployment assisted by the Anima team, he states:

"It was very important at the beginning... let's not fight about the AI system. Let us just look at the trial, at the end of the trial, we'll decide where we go." 

This focus on data and results gradually transformed uncertainty into enthusiasm. One coder evolved from initial resistance to becoming a "super user" who took ownership to drive improvements. 

Key Takeaways for Your Practice:

  • Understand your problem: Have a clear understanding of what problem you want to solve 

  • Centralize your approach: Avoid having two different document processing procedures 

  • Learn from others: Shorten your learning curve by building on others' experiences 

  • Remember human oversight: Staff are still needed to check, summarize, and assign work—Anima enhances rather than replaces 

Ready to Streamline Your Document Workflow?

Anima has transformed document processing at Uzo's practices, cutting document processing to a quarter of the time and reducing backlogs by over 90%. Staff now focus on high-value work while the AI handles routine tasks, improving both efficiency and clinical safety.