Sample size recalculation using conditional power

Denne, Jonathan

doi:10.1002/sim.734

Cited by 111 publications

(86 citation statements)

References 21 publications

(27 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…But, if the unexpected occurs, adaptive methods are available to deal with this. The approach based on maintaining conditional type I error probability put forward by Denne (2001) and by Müller & Schäfer (2001) is particularly promising as it has the potential to be used with error spending designs that already adapt to unpredictable information sequences and, possibly, update sample size in response to estimates of a nuisance parameter.…”

Section: Discussionmentioning

confidence: 99%

Adaptive and nonadaptive group sequential tests

Jennison¹,

Turnbull²

2006

Biometrika

View full text Add to dashboard Cite

SUMMARYMethods have been proposed to re-design a clinical trial at an interim stage in order to increase power.This may be in response to external factors which indicate power should be sought at a smaller effect size, or it could be a reaction to data observed in the study itself. In order to preserve the type I error rate, methods for unplanned design change have to be defined in terms of non-sufficient statistics and this calls into question their efficiency and the credibility of conclusions reached. We evaluate methods for adaptive re-design, extending the theoretical arguments for use of sufficient statistics of Tsiatis & Mehta (2003) and assessing the possible benefits of pre-planned adaptive designs by numerical computation of optimal tests; these optimal adaptive designs are concrete examples of optimal sequentially planned sequential tests proposed by Schmitz (1993). We conclude that the flexibility of unplanned adaptive designs comes at a price and we recommend the appropriate power for a study should be determined as thoroughly as possible at the outset. Then, standard error spending tests, possibly with unevenly spaced analyses, provide efficient designs but it is still possible to fall back on flexible methods for re-design should study objectives change unexpectedly once the trial is under way.

show abstract

Section: Discussionmentioning

confidence: 99%

Adaptive and nonadaptive group sequential tests

Jennison¹,

Turnbull²

2006

Biometrika

View full text Add to dashboard Cite

show abstract

“…The interim analysis aimed to evaluate the feasibility of the study with a futility guideline, and possible sample size adjustment, as well, based on the conditional power at the interim mark. 11 The conditional power is the probability of observing a statistically significant treatment effect at the end of a trial, conditional on the data observed at interim and under specific assumptions on the true treatment trends for the remaining 50% patients to be enrolled. A conditional power Ͻ40% at interim was the cutoff at which the DSMB could recommend stopping the study for futility.…”

Section: Data Managementmentioning

confidence: 99%

A Prospective, Randomized Clinical Trial of Hemodynamic Support With Impella 2.5 Versus Intra-Aortic Balloon Pump in Patients Undergoing High-Risk Percutaneous Coronary Intervention

et al. 2012

View full text Add to dashboard Cite

Background— Although coronary artery bypass grafting is generally preferred in symptomatic patients with severe, complex multivessel, or left main disease, some patients present with clinical features that make coronary artery bypass grafting clinically unattractive. Percutaneous coronary intervention with hemodynamic support may be feasible for these patients. Currently, there is no systematic comparative evaluation of hemodynamic support devices for this indication. Methods and Results— We randomly assigned 452 symptomatic patients with complex 3-vessel disease or unprotected left main coronary artery disease and severely depressed left ventricular function to intra-aortic balloon pump (IABP) (n=226) or Impella 2.5 (n=226) support during nonemergent high-risk percutaneous coronary intervention. The primary end point was the 30-day incidence of major adverse events. A 90-day follow-up was required, as well, by protocol. Impella 2.5 provided superior hemodynamic support in comparison with IABP, with maximal decrease in cardiac power output from baseline of −0.04±0.24 W in comparison with −0.14±0.27 W for IABP ( P =0.001). The primary end point (30-day major adverse events) was not statistically different between groups: 35.1% for Impella 2.5 versus 40.1% for IABP, P =0.227 in the intent-to-treat population and 34.3% versus 42.2%, P =0.092 in the per protocol population. At 90 days, a strong trend toward decreased major adverse events was observed in Impella 2.5–supported patients in comparison with IABP: 40.6% versus 49.3%, P =0.066 in the intent-to-treat population and 40.0% versus 51.0%, P =0.023 in the per protocol population, respectively. Conclusions— The 30-day incidence of major adverse events was not different for patients with IABP or Impella 2.5 hemodynamic support. However, trends for improved outcomes were observed for Impella 2.5–supported patients at 90 days. Clinical Trial Registration— URL: http://www.clinicaltrials.gov . Unique identifier: NCT00562016.

show abstract

“…As several authors have pointed out, it is good practice to design a study as efficiently as possible given initial assumptions, so the benefits of this design are obtained in the usual circumstances where no mid-course change is required. However, if the unexpected occurs, adaptations can be made following the methods described in Section 8 or, more generally, by maintaining the conditional Type I error probability, as suggested by Denne(2001) and Müller and Schäfer (2001). Finally, the use of flexible adaptive methods to rescue an under-powered study should not be overlooked: while it is easy to be critical of a poor initial choice of sample size, it would be naive to think that such problems will cease to arise.…”

Section: Discussionmentioning

confidence: 99%