Abstract: PUB089
In Vitro Evaluation of a Continuous Potassium Monitoring System in Simulated Dialysis Conditions
Session Information
Category: Bioengineering
- 400 Bioengineering
Authors
- Valsamis, Jean-Baptiste, Proton Intelligence Australia, Pty Ltd., Melbourne, Victoria, Australia
- Roshni, Rita, Proton Intelligence Australia, Pty Ltd., Melbourne, Victoria, Australia
- Devendran, Citsabehsan, Proton Intelligence Australia, Pty Ltd., Melbourne, Victoria, Australia
- Lin, Fenge, Proton Intelligence Australia, Pty Ltd., Melbourne, Victoria, Australia
- Wang, Kaixuan, Proton Intelligence Australia, Pty Ltd., Melbourne, Victoria, Australia
- Lensing, Cody J., Proton Intelligence, Inc., Vancouver, British Columbia, Canada
- Ranamukha, Sahan, Proton Intelligence Australia, Pty Ltd., Melbourne, Victoria, Australia
- Cadarso, Victor J, Proton Intelligence Australia, Pty Ltd., Melbourne, Victoria, Australia
Background
Potassium (K+) dysregulation is a life-threatening condition. During a dialysis session, a patient’s K+ levels often begin above the normal range and end below the normal range, before rebounding after the session. We have developed a continuous K+ monitor (CKM) capable of measuring dynamic K+ levels in real-time. In this report, we mimic dialysis conditions in vitro to characterize the sensors’ ability to measure rapidly shifting K+ levels.
Methods
The potentiometric sensor probes consist of a working electrode (WE) that measures K+ directly, selectively, and sensitively and a reference electrode (RE). Eight sensors were immersed in fluid controlled by a pump system able to control K+ levels. We simulated dialysis conditions over 5 hours by changing K+ concentration from 7 mmol/L to 4.5 mmol/L and back to 6 mmol/L. The electrical potential was recorded every 30 seconds, and analyzed by retrofitting to the K+ concentration using a 3 parameters function to account for sensitivity, offset, and drift. Parameters were estimated for each sensor individually using least square fit method.
Results
All sensors responded rapidly during K+ changes (Figure). One sensor showed a faulty RE and one showed a time lag of 15 minutes, and were excluded. With the 6 remaining sensors, the average sensitivity was 48 ± 3.5 mV/dec, the drift was 0.84 ± 0.24 mV/h, and the offset was 2.87 ± 14.2 mV. Compared to the programed K+ profile, MARD was <1% and average accuracy was <0.1 mmol/L (defined as the standard deviation of the difference between estimated K+ and reference K+).
Conclusion
The CKM sensors responded accurately to changing K+ levels across the physiological range during a simulated dialysis session. In vivo testing is needed to further support these results.
Figure: CKM sensor probes tested in a simulated in vitro dialysis session. Profile represents the programmed K+ concentrations.
Funding
- Commercial Support – Proton Intelligence, Inc.