Desgagnés Noémie, King James A, Kline Gregory A, Seiden-Long Isolde, Leung Alexander A
Department of Medicine, University of Calgary, Calgary, Alberta, Canada.
Data and Research Services, Alberta Strategy for Patient-Oriented Research Provincial Research Data Services, Alberta Health Services.
JAMA Netw Open. 2025 Jan 2;8(1):e2455251. doi: 10.1001/jamanetworkopen.2024.55251.
Using albumin-adjusted calcium is commonly recommended for for measuring calcium, but with little empirical evidence to support the practice.
To assess the correlation between total calcium measurements (with or without adjustment) vs the ionized calcium level as a reference standard.
DESIGN, SETTING, AND PARTICIPANTS: This was a population-based cross-sectional study in the province of Alberta, Canada, including adults tested for serum total calcium and ionized calcium simultaneously between January 1, 2013, and October 31, 2019. Statistical analysis was performed from March 2023 to October 2024.
The correlation between unadjusted and adjusted total calcium measurements (using 10 formulas) and the ionized calcium level was evaluated, along with the potential association with the classification of calcium status.
Among 22 658 patients included, 11 889 (52.5%) were female and 10 769 (47.5%) were male; the median (IQR) age was 60 (47-72) years. The unadjusted total calcium (R2 = 71.7%; 95% CI, 71.1%-72.2%) had a stronger correlation with ionized calcium than the commonly used simplified Payne formula (ie, total calcium [mmol/L] + 0.02 [40 - albumin (g/L)]) (R2 = 68.9%; 95% CI, 68.0%-69.6%) and correlated similarly to other formulas (Payne: lowest R2 = 60.3%; 95% CI, 59.3%-61.3%; and James: highest R2 = 76.7%; 95% CI, 76.1%-77.3%). When classifying patients into categories of hypocalcemia, normocalcemia, or hypercalcemia, unadjusted total calcium had the best overall agreement (74.5%) with ionized calcium compared with albumin-adjusted calcium using the original Payne and simplified Payne formulas (agreement 63.0% and 58.7%, respectively). Misclassification using the adjustment formulas was worse in the presence of hypoalbuminemia (albumin level <30 g/L).
In this cross-sectional study drawn from a contemporaneous population, there appeared to be heavy reliance on adjustment formulas for calcium in clinical practice with little gain but considerable risk of misclassification of true calcium status, especially in the presence of hypoalbuminemia. These results suggest that unadjusted total calcium was the best and most practical alternative to ionized calcium.
测量钙时通常建议使用经白蛋白校正的钙，但几乎没有实证证据支持这种做法。
以离子钙水平作为参考标准，评估总钙测量值（无论是否校正）与之的相关性。
设计、地点和参与者：这是一项基于加拿大艾伯塔省人群的横断面研究，纳入了2013年1月1日至2019年10月31日期间同时检测血清总钙和离子钙的成年人。2023年3月至2024年10月进行了统计分析。
评估了未校正和校正后的总钙测量值（使用10种公式）与离子钙水平之间的相关性，以及与钙状态分类的潜在关联。
纳入的22658例患者中，11889例（52.5%）为女性，10769例（47.5%）为男性；年龄中位数（IQR）为60（47 - 72）岁。未校正的总钙（R2 = 71.7%；95%CI，71.1% - 72.2%）与离子钙的相关性比常用的简化佩恩公式（即总钙[mmol/L] + 0.02[40 - 白蛋白（g/L）]）（R2 = 68.9%；95%CI，68.0% - 69.6%）更强，且与其他公式的相关性相似（佩恩公式：最低R2 = 60.3%；95%CI，59.3% - 61.3%；詹姆斯公式：最高R2 = 76.7%；95%CI，76.1% - 77.3%）。将患者分为低钙血症、正常钙血症或高钙血症类别时，与使用原始佩恩公式和简化佩恩公式的经白蛋白校正的钙相比，未校正的总钙与离子钙的总体一致性最佳（74.5%）（一致性分别为63.0%和58.7%）。在存在低白蛋白血症（白蛋白水平<30 g/L）的情况下，使用校正公式的错误分类更严重。
在这项来自同期人群的横断面研究中，临床实践中似乎严重依赖钙的校正公式，收益甚微，但存在对真实钙状态错误分类的巨大风险，尤其是在存在低白蛋白血症的情况下。这些结果表明，未校正的总钙是替代离子钙的最佳且最实用的选择。
