Case study: Reproducible Analytical Pipeline (RAP) for Rapid, Sustained and Statistically-Significant Improvement in MRSA Admission Screening Compliance

6 Aug 2026

Excellence in Clinical Audit Awards, part of Clinical Audit Awareness Week 2026, Innovation winner

This MRSA Admission Screening Compliance project involved the creation of a Reproducible Analytical Pipeline (RAP) for Rapid, Sustained and Statistically-Significant Improvement in MRSA Admission Screening Compliance, using open-source software. It integrated Electronic Patient Record (EPR) and laboratory data, to identify screening-eligible inpatients and determine compliance status, enabling targeted interventions where they were needed most.

Discover much more about the impact of this work in the case study below:

The project was delivered by the team at Newcastle upon Tyne Hospitals NHS Foundation Trust (NUTH), including Daniel Weiand, Paul Bradley, Caroline Cullerton, Geoff Dines, Joanne Field, Peta Le Roux, Jamie Lupton, Julie Samuel and Chris Plummer. It brought together colleagues from multiple teams, including medical microbiology, infection prevention and control (IPC), data science, and information services.

Systematic weekly, patient-level surveillance of MRSA admission screening compliance began on 7 January 2025. The results presented in the award submission at CAAW26 covered the period from 7 January 2025 to 16 June 2026 (approximately 18 months). Weekly reports continue to be produced and shared as part of routine surveillance and improvement activities. The approach has also been adapted to support other clinical and operational priorities at NUTH.

Who is the project designed to support?

The project was designed to support ward and IPC teams in delivering safer care by improving the processes involved in timely MRSA admission screening. For ward teams, the RAP provided reliable, patient-level intelligence so that improvement activity could be focused where it was needed most. For leaders and assurance groups, it provided a trustworthy and fully auditable view of compliance with the Trust’s MRSA policy over time, supported by Statistical Process Control (SPC) charts that visualised changes over time.

What did it set out to achieve?

The project set out to develop and implement a RAP that could combine data from separate systems and triangulate findings to generate new clinical insights. It also aimed to prospectively evaluate, report on, monitor and improve compliance with MRSA admission screening practice.

How did the project work?

At NUTH, MRSA admission screening is required under Trust policy for almost all inpatients, with only a small number of exceptions. Screening is expected to be undertaken between 18 weeks before admission and the date of admission. Prior to this project, organisation-wide screening compliance rates were unknown, and available data could not be reported with sufficient reliability or timeliness to support targeted improvement activity.

The team built a RAP using open-source software and reusable, modular, Git version-controlled code. Each week, the RAP identified current inpatients eligible for MRSA admission screening and determined whether screening had been undertaken in compliance with Trust policy. The pipeline extracted data from the secure data warehouse using SQL and R, combining EPR information with laboratory data to generate new clinical insights. Data were cleaned, checked and analysed in R. Reports were then generated in a standardised HTML format using Quarto and uploaded to a secure central intranet hub.

The weekly RAP outputs provided teams with direct and timely access to consistent, actionable, patient-level data. They highlighted overall compliance, variation between clinical areas, and patient-level gaps requiring action. Statistical process control (SPC) methods were used to identify and visualise meaningful changes beyond normal week-to-week variation.

Did the project involve changes to processes in order to implement?

Yes, the main process change was moving from largely unknown levels of MRSA admission screening compliance to prospective, weekly, patient-level reporting of this metric. This created a regular feedback loop that meaningfully supported targeted local action. For example, ward-level data enabled teams to see where screening was not consistently happening and to focus staff education, reminders and workflow changes in those areas. The project also supported leadership engagement by providing timely, auditable information for governance and assurance.

Did the project succeed and, if so, what is the evidence?

Yes, the project achieved rapid, sustained and statistically significant improvements in MRSA admission screening compliance.

The evidence submitted for the award included prospective weekly surveillance from 7 January 2025 to 16 June 2026, analysed using statistical process control (SPC) methods. The RAP successfully generated reports in 75 of 77 weeks, representing a reliability rate of 97.4%. The two missed reporting weeks were due to summer holiday staffing constraints within the data science team and unscheduled downtime of the laboratory information management system.

A clear example of impact was demonstrated in the Trust’s Medical Assessment Suite, a key area for patient flow where many patients are admitted following initial assessment in the Emergency Department. Using RAP-generated, patient-level data, this clinical area improved from not ranking among the top 10 screening locations in the week ending 3 February 2026, to 6th place one week later, joint second place two weeks later, and first place three weeks later. By 16 June 2026, it had become the leading screening location by a substantial margin.

Compliance with MRSA admission screening

FIG: Compliance with MRSA admission screening (%) (higher is better) – annotated

The RAP also monitored broader MRSA measures, which provided additional context for infection prevention activity.

Other insights or learning points

A key learning point is that improvement depends not only on collecting data, but on ensuring it is timely, trusted and actionable. The RAP provided a fully auditable “single source of truth”, enabling teams to act immediately on the latest Trust-wide, patient-level information rather than relying on delayed reports.

The project demonstrated that cost-effective, fully open-source software (FOSS) can be used within a secure NHS environment to deliver practical and scalable quality improvement gains. Using Git version-controlled R code to run the RAP made the process transparent and reproducible, while enabling automation that substantially reduced manual effort and helped maintain the quality and consistency of outputs.

Another important insight was that ward-level feedback can encourage local ownership. The data did not simply describe performance; it enabled teams to understand where action was needed and supported targeted interventions to achieve rapid, sustained improvement.

Other supporting materials:

Further resources from HQIP