Lessons for dialysis investigators from the Steno-2 Study
Philip A. McFarlane · Journal of Nephrology · 2008
People undergoing dialysis have a substantially shortened survival and a high rate of cardiovascular events. Recent clinical trials have failed to demonstrate a survival benefit of reducing traditional or dialysis-specific cardiovascular risk factors. The reasons for the failure of these clinical trials are unclear. One candidate explanation is that they lacked sufficient statistical power to detect important outcome differences. Several errors in trial design or execution can lead to trials being underpowered. An overestimation of the attributable risk of the condition of interest is a common error. Statistical models that partition attributable risk can be invalid when multiple risk factors are present, as is the case in dialysis patients. Diabetes investigators faced similar challenges in their early clinical trials. However, using only 160 people with diabetes, the Steno-2 Study demonstrated a 50% reduction in cardiovascular risk in people. This impressive result was achieved because patients in the experimental arm of the Steno-2 Study underwent reduction of multiple cardiovascular risk factors simultaneously. A Steno-2 Study approach would be an attractive trial design for dialysis investigators. It could be done with fewer resources than conventional randomized trials, and a positive result would strongly argue against therapeutic complacency in the dialysis unit. However, a variety of limitations to this approach exist. Most significantly, if positive, it would not be possible to determine which individual components of the intervention led to the improved outcomes. Despite the limitations, the Steno-2 design should currently be an attractive option for dialysis investigators.