Consistency of alerts generated by, and implementation of, the NHS England acute kidney injury detection algorithm in English laboratories
2024

Evaluating the NHS England Acute Kidney Injury Detection Algorithm

Sample size: 475634 publication Evidence: high

Author Information

Author(s): Aylward Ryan, Casula Anna, Tiffin Nicki, Ben-Shlomo Yoav, Rayner Brian, Birnie Kate, Caskey Fergus John

Primary Institution: Bristol Medical School, Population Health Sciences, University of Bristol

Hypothesis

How consistently is the NHS England acute kidney injury detection algorithm implemented across English laboratories?

Conclusion

The study found that while alerts submitted to the UK Renal Registry are generally valid for AKI surveillance, there are significant concerns regarding inconsistent laboratory practices and incomplete adoption of the detection algorithm.

Supporting Evidence

  • 1,579,633 local alerts were generated compared to 1,646,850 central alerts.
  • Agreement between local and central alerts was high, with Gwet’s AC1 ranging from 0.97 to 0.98.
  • Lower agreement was observed in laboratories with high baseline serum creatinine levels.

Takeaway

This study looked at how well hospitals in England are using a system to detect kidney problems. It found that while most hospitals do a good job, some are not following the rules properly.

Methodology

The study analyzed 9,096,667 serum creatinine results from 29 laboratories over 15 months, comparing local alerts with centrally generated alerts using inter-rater agreement methods.

Potential Biases

Some laboratories may suppress alerts for certain patient groups, leading to potential underreporting of AKI cases.

Limitations

The study only included data from laboratories that consistently submitted complete data, which may not represent all laboratories in England.

Participant Demographics

The median age of participants was 72 years, with 47% being female.

Digital Object Identifier (DOI)

10.1007/s40620-024-02030-6

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication