POS-020 quality control results from StatSensor Xpress™ point-of-care creatinine meter under real-world usage

Introduction The International Society of Nephrology’s (ISN) 0 by 25 policy aims to eliminate preventable deaths from acute kidney injury (AKI) worldwide by 2025. One of the barriers to this aim is detection of AKI in a timely fashion prior to developing complications. The ISN Kidney Care Network (KCN) project is trialling the use of AKI screening and point of care (POC) creatinine testing to expedite detection of AKI. POC testing is increasingly used with glucometers and haemoglobin measurement but creatinine has remained primarily a lab based test. Recently, several POC creatinine meters have become available1. Studies these have shown appropriate detection of AKI but mixed correlation with lab results2, 3. The promise of hand-held creatinine results to allow rapid detection of AKI has led to their inclusion in the KCN study including the South African site which is described below 4 We aimed to determine if the StatSensor XpressTM POC creatinine meter quality control results remained consistent with the manufacturer’s target range, and variability was appropriate despite the extreme conditions associated with extended use in rural clinics. Methods Over a 6 month period of the Kidney Care Network project in Mseleni, South Africa investigators collected weekly quality control readings from 7 StatSensor XpressTM creatinine meters tested against two control serums (Nova Linearity solutions C1 and C3) provided by the manufacturer. C1 has a creatinine of 84 μmol/L with an acceptable range of 44 to 120 and C3 has a creatinine of 530 μmol/L with an acceptable range of 398 to 630. Results Accuracy: We calculated the difference for each sample compared to the reference mean and then worked out the bias and the limits of agreement (1.96 times SD) C1 bias is 4.3 +/- 24.4 and C3 bias is -35.5 +/- 136.4. Results outside reference range: C1 151/156 results were within reference range of 44 to 120 (96.79%), limits 39-132. C3 147/156 results were within the reference range of 398 to 630 (94.23%), limits 338-638. Change over time: The C1 difference is normally distributed and a regression model of the c1 difference against time is significant (coefficient -0.05, 95% CI -0.09, -0.01), p=0.01), implying the mean difference fell over time (by -0.05/day over the study period). The C3 difference is not normally distributed. Modelling data as linear showed significant association with time (coefficient -0.61, 95% CI -0.82, -0.41, p<0.0001). Conclusions The results overall showed reasonable accuracy consistent with previous studies on this meter, these meet some but not all of the rigorous standards set for laboratory based testing. During the study on several occasions the solution and or the test strips were left outside of fridges for extended periods due to human error, and on two occasions there were powercuts leading to loss of the cold chain. Variability over time could be expected due to degradation of the strips, the control solution, or the machine itself, especially when subjected to harsh environments of the study field. Results are encouraging that despite time exposed to these conditions there was decreased variation from the target value, thought this could be explained by regression to the mean. Reassuringly, in none of our cases did the meter provide a normal result from C3, in keeping with its current use as a screening tool for detection of severe AKI.

Citation

Fredlund, M., Nyawo, M., Hamilton, A., Rocco, M., & Cullis, B. (2021). POS-020: Quality control results from StatSensor Xpress™ point-of-care creatinine meter under real-world usage. Kidney International Reports, 6(4, Suppl.), S9. https://doi.org/10.1016/j.ekir.2021.03.026

URI

https://hdl.handle.net/20.500.14709/1105