Sample size re‐estimation for clinical trials with longitudinal negative binomial counts including time trends

Asendorf, Thomas; Henderson, Robin; Schmidli, Heinz; Friede, Tim

doi:10.1002/sim.8061

Cited by 14 publications

(17 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…An ideal time point for the sample size recalculation depends on several factors, such as pace of recruitment, trial duration, as well as uncertainty on initial choices of nuisance parameters (Chuang‐Stein, Anderson, Gallo, & Collins, 2006). Given that not all patients enrolled at interim might have completed follow‐up, approaches combining short‐ and long‐term data have been proposed to increase precision of the estimated variance components, which might be transferred to multicenter trials, see, for example, Asendorf, Henderson, Schmidli, and Friede (2019) for an application to longitudinal counts and Friede and Kieser (2006) for an overview. For further reading on practical guidance, we refer to the summary by Pritchett et al.…”

Section: Discussionmentioning

confidence: 99%

Sample size recalculation in multicenter randomized controlled clinical trials based on noncomparative data

Harden

Friede

2020

Biometrical J

Self Cite

View full text Add to dashboard Cite

Many late‐phase clinical trials recruit subjects at multiple study sites. This introduces a hierarchical structure into the data that can result in a power‐loss compared to a more homogeneous single‐center trial. Building on a recently proposed approach to sample size determination, we suggest a sample size recalculation procedure for multicenter trials with continuous endpoints. The procedure estimates nuisance parameters at interim from noncomparative data and recalculates the sample size required based on these estimates. In contrast to other sample size calculation methods for multicenter trials, our approach assumes a mixed effects model and does not rely on balanced data within centers. It is therefore advantageous, especially for sample size recalculation at interim. We illustrate the proposed methodology by a study evaluating a diabetes management system. Monte Carlo simulations are carried out to evaluate operation characteristics of the sample size recalculation procedure using comparative as well as noncomparative data, assessing their dependence on parameters such as between‐center heterogeneity, residual variance of observations, treatment effect size and number of centers. We compare two different estimators for between‐center heterogeneity, an unadjusted and a bias‐adjusted estimator, both based on quadratic forms. The type 1 error probability as well as statistical power are close to their nominal levels for all parameter combinations considered in our simulation study for the proposed unadjusted estimator, whereas the adjusted estimator exhibits some type 1 error rate inflation. Overall, the sample size recalculation procedure can be recommended to mitigate risks arising from misspecified nuisance parameters at the planning stage.

show abstract

Section: Discussionmentioning

confidence: 99%

Sample size recalculation in multicenter randomized controlled clinical trials based on noncomparative data

Harden

Friede

2020

Biometrical J

Self Cite

View full text Add to dashboard Cite

show abstract

“…Modeling the blinded sample through a mixture of two distributions has also been considered by Asendorf et al 15,16 in the context of longitudinal count data. Since we aim to only estimate the nuisance parameters, we replace the cumulative rate function

Λ_{T} (s)

in () under the assumption of a treatment effect

β_{H_{1}}

Λ_{C} (s) \exp (β_{H_{1}})

.…”

Section: Blinded Continuous Information Monitoring For Recurrent Evenmentioning

confidence: 99%

Blinded continuous information monitoring of recurrent event endpoints with time trends in clinical trials

Mütze

Salem

Benda

et al. 2020

Statistics in Medicine

Self Cite

View full text Add to dashboard Cite

Blinded sample size re‐estimation and information monitoring based on blinded data has been suggested to mitigate risks due to planning uncertainties regarding nuisance parameters. Motivated by a randomized controlled trial in pediatric multiple sclerosis (MS), a continuous monitoring procedure for overdispersed count data was proposed recently. However, this procedure assumed constant event rates, an assumption often not met in practice. Here we extend the procedure to accommodate time trends in the event rates considering two blinded approaches: (a) the mixture approach modeling the number of events by a mixture of two negative binomial distributions and (b) the lumping approach approximating the marginal distribution of the event counts by a negative binomial distribution. Through simulations the operating characteristics of the proposed procedures are investigated under decreasing event rates. We find that the type I error rate is not inflated relevantly by either of the monitoring procedures, with the exception of strong time dependencies where the procedure assuming constant rates exhibits some inflation. Furthermore, the procedure accommodating time trends has generally favorable power properties compared with the procedure based on constant rates which stops often too late. The proposed method is illustrated by the clinical trial in pediatric MS.

show abstract

“…For a relatively long time, the focus in the field of nuisance parameter‐based sample size reestimation has mainly been on continuous and binary endpoints with some more recent interest in recurrent events . Also, procedures for blinded continuous monitoring for trials with recurrent event endpoints have very recently been considered .…”

Section: Introductionmentioning

confidence: 99%

Blinded sample size reestimation in event‐driven clinical trials: Methods and an application in multiple sclerosis

Friede

Pohlmann

Schmidli

2019

Pharmaceutical Statistics

Self Cite

View full text Add to dashboard Cite

Motivated by a recently completed trial in secondary progressive multiple sclerosis, we developed blinded sample size reestimation procedures for clinical trials with time-to-event endpoint and assessed their properties in simulation studies. Assuming independent right-censoring and proportional hazards for the two treatment groups, we considered event-driven designs with fixed number of events, which guarantees the power to be at a desired level under a certain alternative. We develop reestimation procedures based on parametric models and show that these maintain the expected duration of the trial at a target length in flexible follow-up designs across a range of nuisance parameter values by adjusting the number of patients recruited into the trial based on blinded nuisance parameter estimates. Furthermore, we provide convincing evidence from a simulation study that such procedures proposed do not inflate the type I error rate in any practically relevant way, thereby satisfying the standards set by relevant international guidelines. Inspired by practical application of these procedures, we outline a number of extensions including methods for extrapolating the observed survival curve beyond the interim time point, application of reestimation procedures to interval censored data, and situations in which a confirmation of event is required leading to a certain lag time. KEYWORDSadaptive design, internal pilot study, multiple sclerosis, parametric models, sample size | INTRODUCTIONThe idea of using data from the first stage of a two-stage design to estimate a nuisance parameter and to adjust the sample size of the second stage accordingly in order to maintain the power of a hypothesis test based on all data goes back a long way and was described by Stein in the 1940s. 1 In contrast to Stein's proposal that used the variance estimate from the first stage data in the test statistic of the final analysis, it is common practice to use the data of both design stages to estimate the nuisance parameters in the final analysis. In their seminal paper, Wittes and Brittain 2 referred to the latter as the internal pilot study design. The advantages and disadvantages of both approaches were, for example, investigated by Proschan and Wittes, 3 considering sample size reestimation based on unblinded nuisance parameter estimates. Already, from the early paper by Wittes and Brittain, 1 it was clear that the approach using naively a nuisance parameter estimate based on data aggregated across both design stages inflates the type I error rate above the nominal level and various ways of adjusting for this in the final analysis have been proposed over the years. [4][5][6][7][8] Blinded procedures, which use interim data pooled across treatment code and therefore do not require breaking the treatment code at interim, were found to be equally effective in adjusting the sample size, 9 but with much smaller,

show abstract

Sample size re‐estimation for clinical trials with longitudinal negative binomial counts including time trends

Cited by 14 publications

References 40 publications

Sample size recalculation in multicenter randomized controlled clinical trials based on noncomparative data

Sample size recalculation in multicenter randomized controlled clinical trials based on noncomparative data

Blinded continuous information monitoring of recurrent event endpoints with time trends in clinical trials

Blinded sample size reestimation in event‐driven clinical trials: Methods and an application in multiple sclerosis

Contact Info

Product

Resources

About