The Longitudinal PTH-Study

The main study aim is to quantify the agreement between the analytical results provided by two third generation and two second generation Parathyroid hormone (PTH) assays. The primary comparison will be performed between the second-generation PTH assay"Intact PTH assay" from Siemens Healthcare Diagnostics Inc. and the third-generation PTH assay "biointact (1-84)" from Roche Diagnostics in terms of a Bland-Altman analysis. Several studies have evaluated the correlation between various PTH assays at a single time-point, but no previous study has tested the hypothesis that longitudinal changes in PTH levels, which are important for making treatment decisions, can be monitored by several PTH assays alike. To this aim, the key secondary objective is to analyze the longitudinal variance in PTH over the course of 1 year, using each of two assays of the second and third generations, respectively. Other secondary objectives include determining changes in serum phosphate, serum calcium, fibroblast growth factor 23 (FGF23), with respect to treatment decisions. For clinical applicability of the results to be obtained here, an important goal of the present study will be not to influence treatment decisions, which will remain independent of the study investigators, at the full responsibility of the hemodialysis physicians.

At every quarterly blood draw over the course of one year, the investigators will freeze the serum from 100 patients, and at the end of 4 quarters the investigators will analyze PTH-levels using the following assays: Intact Parathyroid Hormone (Advia Centaur, Siemens Healthcare), PTH-Intact (Cobas, Roche), PTH (1-84) - The agreement between the PTH assays will be analyzed at baseline, as well as at the subsequent quarterly evaluation time-points by Bland-Altman analysis and complemented by Passing-Bablok regression. The longitudinal changes in PTH will be displayed graphically and analyzed by estimating the within-patient variance across time, the between patient variance at each time-point as well as effects on the mean log-PTH level due to course of disease and therapeutic interventions from a linear mixed model.