Fractioned urine collections to assess urine supersaturation in urolithiasis; a pathway worth exploring
- PDF / 354,493 Bytes
- 2 Pages / 595.276 x 790.866 pts Page_size
- 85 Downloads / 225 Views
LETTER TO THE EDITOR
Fractioned urine collections to assess urine supersaturation in urolithiasis; a pathway worth exploring Maria Chalkidou1,2 · Eva Pella2 · Dimitrios Hatzichristou3 · Pantelis Sarafidis2 Received: 24 March 2020 / Accepted: 4 April 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020
Dear editor We read with great interest the study from Rodriguez et al. [1] that analysed which of 8 fractioned urine collections covering different periods of the day and night was better associated with the 24-urine collection, with regards to supersaturation of calcium oxalate, calcium phosphate, and uric acid. Without doubt, urine supersaturation levels are predictive of stone formation, thus the 24-h urinalysis is considered an important tool in the management of high-risk stoneformers [2]. As the authors correctly point out, performing a proper 24-h collection is not an easy task for the majority of patients, resulting in limited accuracy. Although valid diagnostic methods evolved for other urine parameters; i.e. the albumin-to-creatinine ratio to assess albuminuria [3], a “spot” urine specimen cannot directly apply for urine supersaturation, as urine concentration in the above substances may vary significantly throughout the day corresponding to different crystallization risks. There are, however, a number of issues of this innovative study that came to our attention. Throughout the manuscript, the authors refer to the “variability” of 24-h supersaturation, without reporting any variability index (i.e. coefficient of variation or relevant). Instead, they use mean or log-transformed values of calculated supersaturation of calcium oxalate, calcium phosphate, and uric acid as the dependent variables. Since the study included 4 different 24-h collections from each of their 12 subjects, at least the intra-individual * Pantelis Sarafidis [email protected] 1
variability of the supersaturation between these 4 collections could have been estimated. Instead of that, the authors use each of the four 24-h samples from each participant as a separate sample, in order to increase the study power. This, however, rather does not allow the use of linear regression modelling to quantify the individual contribution of each fractioned sample to the 24-h information value (i.e. what percentage of the R2 or “variation” of 24-h supersaturation can be attributed to the supersaturation of each sample) simply because the samples from the same subject would largely reflect a similar pattern, and, therefore, these 48 samples would be different from samples of 48 separate individuals, as the authors also acknowledge. Finally, the authors calculate the percentage of the R2 attributed to each individual sample starting from a univariate linear regression analysis in Table 2; to our view, as some of the individual samples do not associate with the 24-h urine supersaturation (p up to 0.70) it would be more appropriate to start from a multivariate regression. By all means, the study of Rodriquez et al. is an important additio
Data Loading...
