Skåne University Hospital, Lund, Sweden.
Baxter International Inc., Lund, Sweden.
PLoS One. 2021 Feb 25;16(2):e0247477. doi: 10.1371/journal.pone.0247477. eCollection 2021.
BACKGROUND/AIMS: Post-filter ionized calcium (iCa) measured on a blood gas analyzer (BGA) during regional citrate anticoagulated continuous renal replacement therapy (CRRT) are needed to control the regime. This increases the workload and requires attention including interpretation of blood analyses. Two algorithms were developed to calculate the post-filter iCa instead. The first algorithm used measured systemic total calcium and the second used a selected set of values from an initial blood gas sample as input.
Calculated post-filter iCa values were compared to real blood gas analyses. 57 patients treated at the intensive care unit at Skåne University Hospital in Lund during 2010-2017 were included after applying inclusion and exclusion criteria. Clinical and machine parameters were collected from the electronic medical records. Non-quality checked data contained 1240 measurements and quality checked data contained 1034 measurements.
The first algorithm using measured systemic total calcium resulted in slightly better precision and trueness with an average difference between the predicted and measured post-filter iCa concentration of 0.0185±0.0453 mmol/L for quality checked data, p<0.001. Neither algorithm could detect all instances requiring intervention.
The algorithms were able to estimate in range postfilter iCa values with great trueness and precision. However, they had some difficulties to estimate out-of-range postfilter iCa values. More work is needed to improve the algorithms especially in their citrate-modelling.
背景/目的：在局部枸橼酸抗凝连续肾脏替代治疗（CRRT）期间，需要在血气分析仪（BGA）上测量过滤后的离子化钙（iCa），以控制治疗方案。这增加了工作量，并需要关注，包括对血液分析的解读。因此，开发了两种算法来计算过滤后的 iCa。第一种算法使用测量的系统总钙，第二种算法使用初始血气样本中的一组选定值作为输入。
将计算出的过滤后 iCa 值与实际的血气分析结果进行比较。纳入标准为 2010 年至 2017 年在瑞典隆德大学医院重症监护病房接受治疗的 57 例患者。收集来自电子病历的临床和机器参数。非质量检查数据包含 1240 次测量，质量检查数据包含 1034 次测量。
使用测量的系统总钙的第一种算法具有更好的精度和准确度，质量检查数据中预测过滤后 iCa 浓度与实际测量浓度之间的平均差异为 0.0185±0.0453mmol/L，p<0.001。两种算法都无法检测到所有需要干预的情况。
这些算法能够准确地估计过滤后的 iCa 值，但在估计超出范围的过滤后 iCa 值时存在一些困难。需要进一步改进算法，特别是在枸橼酸盐建模方面。
