Number and frequency of albuminuria measurements in clinical trials in diabetic nephropathy

Abstract

Background and objectives: Albuminuria change is often used to assess drug efficacy in intervention trials in nephrology. The change is often calculated using a variable number of urine samples collected at baseline and end of treatment. Yet more albuminuria measurements usually occur. Because albuminuria shows a large day-to-day variability, this study assessed to what extent the average and the precision of the antialbuminuric drug effect varies with the number of urine collections at each visit and the number of follow-up visits.

Design, setting, participants, & measurements: This study used data from three randomized intervention trials (Aliskiren Combined with Losartan in Type 2 Diabetes and Nephropathy, Selective Vitamin D Receptor Activation for Albuminuria Lowering, and Residual Albuminuria Lowering with Endothelin Antagonist Atrasentan) including patients with type 2 diabetes and macroalbuminuria. Albuminuria-lowering drug effects were estimated from one, two, or three urine collections at consecutive days before each study visit and reported as albuminuria change from baseline to end of treatment or the change over time considering an average of all follow-up albuminuria measurements.

Results: Increasing the number of urine collections for an albuminuria measurement at baseline and end of treatment or using all study visits during follow-up did not alter the average drug effect. The precision of the drug effect increased (decreased SEM) when the number of study visits and the number of urine collections per visit were increased. Using all albuminuria measurements at all study visits led to a 4- to 6-fold reduction in sample size to detect a 30% albuminuria-lowering treatment effect with 80% power compared with using baseline and end-of-treatment albuminuria measurements alone.

Conclusions: Increasing the number of urine collections per study visit and the number of visits over time does not change the average drug effect estimate but markedly increases the precision, thereby enhancing statistical power. Thus, clinical trial designs in diabetic nephropathy using albuminuria as an end point can be significantly improved, leading to smaller sample sizes and less complex trials.

Keywords: albuminuria; diabetic nephropathy; proteinuria; randomized controlled trials.

Figure 1.

Albuminuria change between baseline and end of treatment. The top panels show albuminuria reduction (in percentages and 95% CIs) according to the number of consecutive collected urine samples considering the change from baseline and the end-of-study visit. The bottom panels show the SEM of the treatment effect (Δ log albuminuria) according to the number of consecutive collected urine samples considering the change from baseline to the end-of-study visit. 95% CI, 95% confidence interval; AVOID, Aliskiren Combined with Losartan in Type 2 Diabetes and Nephropathy; JPN, Japan; RADAR, Residual Albuminuria Lowering with Endothelin Antagonist Atrasentan; VITAL, Selective Vitamin D Receptor Activation for Albuminuria Lowering.

Figure 2.

Albuminuria-lowering treatment effect and SEM of the treatment effect as a function of number of follow-up visits. The top panels show the albuminuria-lowering treatment effect. The bottom panels show the SEM of the treatment effect (log scale). The x axis shows the number of follow-up visits that were used to calculate the treatment effect and the time interval during which these follow-up visits were conducted. A maximum of five follow-up visits were scheduled in the (A) AVOID trial (at weeks 4, 8, 12, 16, and 24), whereas six follow-up visits were performed in the (B) VITAL (at weeks 4, 8, 12, 16, 20, and 24) and (C and D) RADAR/JPN (at weeks 2, 4, 6, 8, 10, and 12) trials.