Arthur Smith Institute for Urology, Hosftra North Shore-LIJ School of Medicine, New Hyde Park, New York.
J Urol. 2013 Dec;190(6):2106-11. doi: 10.1016/j.juro.2013.05.115. Epub 2013 Jun 11.
We determined the accuracy of 24-hour urinalysis in predicting stone type and identify the associations between 24-hour urine elements with stone type.
We performed a retrospective review of 503 stone formers with stone composition analysis and 24-hour urinalysis available. Analysis of 24-hour urine elements across stone types was performed using Fisher's exact test and ANOVA. Multinomial logistic regression was used to predict stone type based on 24-hour urinalysis.
A total of 280 (56%) patients had predominantly calcium oxalate, 103 (20%) had uric acid, 93 (19%) had calcium phosphate, 16 (3%) had mixed and 11 (2%) had other stone types. There were several significant patient characteristics and 24-hour urinalysis differences across stone type groups. The statistical model predicted 371 (74%) calcium oxalate, 78 (16%) uric acid, 52 (10%) calcium phosphate, zero mixed and 2 (less than 1%) other stone types. The model correctly predicted calcium oxalate stones in 85%, uric acid in 51%, calcium phosphate in 31%, mixed in 0% and other stone types in 18% of the cases. Of the predicted stone types, correct predictions were 61%, 69%, 56% and 71% for calcium oxalate, uric acid, calcium phosphate and other stones types, respectively. The overall accuracy was 64%. Plots were used to explore the associations between each 24-hour urine element with each predicted stone type adjusted for all the others urinary elements.
A 24-hour urinalysis alone does not accurately predict stone type. However, it may be used in conjunction with other variables to predict stone composition.
我们确定了 24 小时尿液分析预测结石类型的准确性，并确定了 24 小时尿液成分与结石类型之间的关系。
我们对 503 名结石形成者进行了回顾性分析，这些患者均有结石成分分析和 24 小时尿液分析。使用 Fisher 精确检验和 ANOVA 对 24 小时尿液成分进行分析。使用多项逻辑回归根据 24 小时尿液分析预测结石类型。
共有 280 例（56%）患者以草酸钙为主，103 例（20%）患者为尿酸，93 例（19%）患者为磷酸钙，16 例（3%）为混合结石，11 例（2%）为其他结石类型。各结石类型组患者特征和 24 小时尿液分析结果存在明显差异。该统计模型预测了 371 例（74%）草酸钙结石、78 例（16%）尿酸结石、52 例（10%）磷酸钙结石、0 例混合结石和 2 例（<1%）其他结石类型。该模型正确预测了 85%的草酸钙结石、51%的尿酸结石、31%的磷酸钙结石、0%的混合结石和 18%的其他结石类型。在预测的结石类型中，正确预测的草酸钙、尿酸、磷酸钙和其他结石类型的比例分别为 61%、69%、56%和 71%。总体准确率为 64%。我们绘制了散点图，以探索每个 24 小时尿液成分与每个预测结石类型之间的关系，该分析在调整所有其他尿液成分的情况下进行。
单独的 24 小时尿液分析不能准确预测结石类型。但是，它可以与其他变量一起用于预测结石成分。
