Abstract: PUB311
Report on STOP-ACEi Trial: Reviewing Creatinine Intraclass Correlation Coefficients
Session Information
Category: Pathology and Lab Medicine
- 1800 Pathology and Lab Medicine
Authors
- Spencer, Sebastian, University of Hull, Hull, England, United Kingdom
- Desborough, Robert, Hull University Teaching Hospitals NHS Trust, Hull, England, United Kingdom
- Mehta, Samir, University of Birmingham Clinical Trials Unit, Birmingham, England, United Kingdom
- Bhandari, Sunil, Hull University Teaching Hospitals NHS Trust, Hull, England, United Kingdom
Background
The STOP-ACEi trial assessed whether discontinuing ACE inhibitors (ACEi) or angiotensin II receptor blockers (ARBs) slows renal decline in patients with advanced chronic kidney disease (CKD). Serum creatinine, a critical biomarker for estimating glomerular filtration rate (eGFR), was measured at local laboratories throughout the study. Inter-laboratory variation may impact outcome reliability. This secondary analysis evaluated agreement between local creatinine results and a centralised re-analysis using re-run serum samples, exploring implications for clinical trials and kidney function monitoring.
Methods
Stored serum samples were re-analysed centrally using the same assay platform at baseline, year 1, year 2, and year 3. Agreement was assessed using Bland-Altman plots, scatter plots, and Pearson’s correlation. A repeated measures mixed-effects model estimated intraclass correlation coefficients (ICCs) at the patient, centre, and measurement levels. Sensitivity analyses assessed the impact of storage duration, freeze-thaw cycles, and patient-level factors on variability.
Results
Creatinine values from the centralised and local laboratories showed strong agreement (Pearson’s r = +0.91), with minimal systematic bias on Bland-Altman analysis. Most variability arose at the patient level (ICC = 0.48; 95% CI: 0.43–0.53). Centre-level variability was low (ICC = 0.05; 95% CI: 0.02–0.11), while measurement-level variability was negligible and could not be reliably estimated. Sensitivity analysis revealed no significant effects of storage conditions or freeze-thaw cycles on measurement reliability.
Conclusion
This study confirms the reliability of local laboratory creatinine measurements, with minimal inter-laboratory variability unlikely to affect trial outcomes. Patient-level differences accounted for most observed variability.