Comparison of GFR Measurements Assessed From Single Versus Multiple Samples
Received 20 August 2008; accepted 20 March 2009.
Background
Many previous studies have evaluated single-sample glomerular filtration rate (GFR) against multisample GFR, of which the single sample was a member, but none have compared single and multisample GFRs against an independent reference method. We therefore performed this comparison by using simultaneous independent multisample GFR measured with a different indicator.
Setting & Participants
University hospital: patients and healthy volunteers (95 studies in 60 patients and 20 healthy participants). Healthy volunteers were studied fasting and after food; 10 of them had a repeated fasting study.
Study Design
Diagnostic test study.
Index Test
Single-sample GFR.
Reference Test
Multisample GFR with a different indicator.
Measurements
GFR was measured by using chromium-51 (51Cr)-EDTA and iohexol, injected into opposite arms and scaled to 1.73 m2. Blood samples, obtained bilaterally 20, 40, 60, 120, 180, and 240 minutes after injection, were assayed for indicator injected contralaterally. Single-sample GFR (Jacobsson method) was calculated from indicator concentrations at 3 and 4 hours. Single-sample GFR from 1 indicator was compared with multisample GFR from the other and vice versa, as well as from the same indicator. Differences were expressed as limits of agreement between paired measurements in Bland-Altman plots. Precision was expressed as the SD of the mean difference between paired measurements.
Results
Limits of agreement between multisample GFRs measured by using 51Cr-EDTA and iohexol (−12 to 20 mL/min) were similar to the corresponding limits for single-sample GFR at 3 (−16 to 17 mL/min) and 4 hours (−11 to 17 mL/min). The precision of single-sample GFR at 4 hours by using 51Cr-EDTA for predicting iohexol multisample GFR (6.9 mL/min) was better than that of multisample GFR with 51Cr-EDTA (7.9 mL/min). When analysis was limited to patients with GFR less than 60 mL/min, single-sample GFR was slightly inferior to multisample GFR. In healthy participants, single-sample GFR with 51Cr-EDTA at 3 and 4 hours showed repeatability (SD of change, 9.4 and 9.3 mL/min) similar to multisample GFR with 51Cr-EDTA (10.7 mL/min). Single-sample GFR at 4 hours by using 51Cr-EDTA detected a food-induced increase in GFR (4.4 ± 5.9 mL/min; P < 0.001) with more confidence than multisample GFR by using 51Cr-EDTA (4.6 ± 7.5 mL/min; P < 0.01).
Limitations
No separate gold standard (eg, inulin) to facilitate interpretation of observed differences between 2 markers.
Conclusions
Single-sample GFR is as reliable as multisample GFR for measuring GFR, especially when GFR is greater than 60 mL/min.
ABSTRACT