Stability of ionized calcium measurements at concentrations less than 0.3 mmol/L by point-of-care blood gas instruments: application for postfilter calcium quality control in patients with citrate anticoagulation during extracorporeal circulation

Introduction

Extracorporeal anticoagulation with sodium citrate is used in critically ill patients with continuous renal replacement therapy. ¹ Citrate is administered in the extracorporeal circulation to prevent activation of blood coagulation by binding ionized calcium (Ca²⁺). The known complication of this procedure is hypocalcaemia. Therefore, monitoring of ionized calcium is of great importance for evaluation of the extracorporeal citrate anticoagulation. Blood samples are usually analysed by point-of-care blood gas instruments. In order to achieve sufficient extracorporeal anticoagulation, the postfilter Ca²⁺ must be maintained at extremely low concentration (0.25–0.34 mmol/L).^1–3 There is no information about stability of Ca²⁺ measurements at this concentration. Recently, concern has been expressed about the discrepancy of postfilter Ca²⁺ measurements by different blood gas instruments. ⁴ The aim of this study was to evaluate the stability of Ca²⁺ measurements at concentrations below 0.3 mmol/L by blood gas instruments ABL800 and ABL90, Radiometer.

Methods

Ca²⁺ was measured by the potentiometric method with ion-selective electrodes by ABL800 and ABL90 instruments, Radiometer Medical, Copenhagen, Denmark. The measurement interval for Ca²⁺ was 0.20–9.99 mmol/L for both instruments.^5,6 Altogether four identical ABL800 instruments were used in the study: two of them for real postfilter samples, while two other instruments were used for other routine measurements at the clinical laboratory. Mean bias between different ABL800 instruments was 0.01 mmol/L (95% CI: 0.00–0.02). To evaluate stability of Ca²⁺ measurements below 0.3 mmol/L, we used the external standard Qualicheck level 3 s7950 (Radiometer Medical, Copenhagen, Denmark) as a quality control.

The assigned value of the external standard was 0.22 mmol/L for ABL800 and 0.25 mmol/L for ABL90. For each measurement, a new ampoule of the external standard was opened and an adaptor was used to connect the original glass ampoule to the Radiometer instrument. The external standard Qualicheck level 3 s7950 was measured once per day at all four ABL800 instruments and one ABL90 instrument. Mean weekly control values were based on these daily measurements.

Two different lots of Ca²⁺ membranes were tested for the same ABL800 instruments, each lot for the period of 12 weeks. The Ca²⁺ membranes were replaced after four weeks at two instruments that were routinely used for the postfilter samples, while two other ABL800 instruments had Ca²⁺ membranes for 12 weeks, which is the recommended lifetime of Ca²⁺ membrane for ABL800. ⁵ For the ABL90 instrument, the sensor cassettes were replaced after four weeks, as required. ⁶

Further, weekly mean deviations from the assigned control value for Ca²⁺ measurements with two different Ca²⁺ membrane lots were calculated for ABL800 instruments for weeks 1–12. These means were presented with standard deviations (SDs) to show lot-to-lot variation.

The deviation of Ca²⁺ control value >0.05 mmol/L was considered as clinically significant (>20% deviation from 0.25 mmol/L). The definition of the clinically significant drift was based on the Fresenius Medical Care instructions for stepwise adjusting of the citrate anticoagulation, ³ as well as on the recommendations for the drift tolerance limits (±0.05 mmol/L) for the Ca²⁺ analysis set by the Radiometer Medical, Copenhagen, Denmark.^5,6

Results

The stability of Ca²⁺ measurements below 0.3 mmol/L is presented in the Figure 1. Figure 1 shows the deviations (in %) of the Ca²⁺ control weekly means from the assigned control value with SDs that reflect lot-to-lot variation for ABL800 instruments (Figure 1). During the first four weeks, all Ca²⁺ measurements were within the acceptable range and showed the same results for all Radiometer ABL800 instruments. Further, we have observed a continuous increasing of Ca²⁺ values at the ABL800 instruments. Already after four weeks, we observed a deviation of the control over 0.05 mmol/L from the assigned value. The mean increase of the Ca²⁺ control values was 28% at the beginning of the 5th week. At the beginning of the 9th week, the increase of the Ca²⁺ values was over 40% with lot-to-lot variation 5%. The deviation from the assigned control value was 42% at the end of the 12th week with lot-to-lot variation 15%. OPEN IN VIEWER

When the same two lots of the Ca²⁺ membrane were replaced after four weeks at the ABL800 instruments, the control values remained within the acceptable range. The mean deviation from the assigned value was 9%.

The control Ca²⁺ measurements were more stable when measured by ABL90 instrument. The mean variation was 0.005 mmol/L; the largest control deviation did not exceed 0.01 mmol/L (4% of the assigned control value).

Additionally, the routine Ca²⁺ quality control (S7745, Radiometer) was measured at the same time at these instruments. This control had higher Ca²⁺ assigned values: 0.51 mmol/L for ABL800 and 0.37 mmol/L for ABL90 (data not shown). The ABL90 control had little deviation, mean CVa was 0.7%. The ABL800 control has shown a tendency to increase, but remained within acceptable range during 12 weeks. The deviation from the assigned value was less than 0.05 mmol/L. Mean CVa was <3% for all ABL800 instruments. Thus, the routine Ca²⁺ quality control for the ABL800 has not reflected the drift of Ca²⁺ measurements at lower concentrations.

Discussion and conclusions

ABL800 instruments were not stable in measuring Ca²⁺ at concentrations less than 0.3 mmol/L when Ca²⁺ membrane was used for the usual lifetime of 12 weeks. Ca²⁺ measurements at ABL90 were more stable compared with the ABL800 instruments. For ABL90 instrument, the sensor cassette was replaced after four weeks, as required. Based on our data, we recommend to replace Ca²⁺ membrane on the ABL800 instruments after four weeks. Otherwise, it may result in a clinically significant overestimation of Ca²⁺ at concentrations below 0.3 mmol/L by the ABL800 instruments. It is alarming, because postfilter Ca²⁺ evaluation in patients with citrate anticoagulation during extracorporeal circulation requires precise Ca²⁺ measurements at these concentrations. The drift that we observed in quality control measurements may reflect possible Ca²⁺ overestimation in the real postfilter samples. Another analytical problem is a definite lot-to-lot variation of the Ca²⁺ measurements at these concentrations. It is necessary to use an extra quality control in addition to the usual quality controls to monitor stability of Ca²⁺ measurements below 0.3 mmol/L.