Information content of stepped‐wedge designs when treatment effect heterogeneity and/or implementation periods are present

Kasza, Jessica; Taljaard, Monica; Forbes, Andrew

doi:10.1002/sim.8327

Cited by 20 publications

(30 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(This is a development of previous work in the discrete-time context, proposing that the information content of each cluster-period in a complete design could be quantified as the loss of precision when recruitment is suspended in that cluster period.) [ 13 , 14 ]…”

Section: Methodsmentioning

confidence: 99%

The hunt for efficient, incomplete designs for stepped wedge trials with continuous recruitment and continuous outcome measures

Hooper

Kasza

Forbes

2020

BMC Med Res Methodol

Self Cite

View full text Add to dashboard Cite

Background We consider the design of stepped wedge trials with continuous recruitment and continuous outcome measures. Suppose we recruit from a fixed number of clusters where eligible participants present continuously, and suppose we have fine control over when each cluster crosses to the intervention. Suppose also that we want to minimise the number of participants, leading us to consider “incomplete” designs (i.e. without full recruitment). How can we schedule recruitment and cross-over at different clusters to recruit efficiently while achieving good precision? Methods The large number of possible designs can make exhaustive searches impractical. Instead we consider an algorithm using iterative improvements to hunt for an efficient design. At each iteration (starting from a complete design) a single participant – the one with the smallest impact on precision – is removed, and small changes preserving total sample size are made until no further improvement in precision can be found. Results Striking patterns emerge. Solutions typically focus recruitment and cross-over on the leading diagonal of the cluster-by-time diagram, but in some scenarios clusters form distinct phases resembling before-and-after designs. Conclusions There is much to be learned about optimal design for incomplete stepped wedge trials. Algorithmic searches could offer a practical approach to trial design in complex settings generally.

show abstract

Section: Methodsmentioning

confidence: 99%

The hunt for efficient, incomplete designs for stepped wedge trials with continuous recruitment and continuous outcome measures

Hooper

Kasza

Forbes

2020

BMC Med Res Methodol

Self Cite

View full text Add to dashboard Cite

show abstract

“…Accordingly, the trial could be described as an "incomplete cross-forward cluster randomised design" [47] because although all sites cross over from the control to the intervention condition, the stepped wedge design is "incomplete" because measurements are taken in only a proportion of all possible cluster-periods. Another name for this design is a "staircase" design, which has already been suggested as a potential highly efficient trial design by Kasza and Forbes, and broadly similar designs have already been adopted in practice [48][49][50][51]. The rationale for this specific stepped wedge design is that data collection is concentrated in those cluster periods that provide the most information about the treatment effect (i.e.…”

Section: Methodsmentioning

confidence: 99%

“…The rationale for this specific stepped wedge design is that data collection is concentrated in those cluster periods that provide the most information about the treatment effect (i.e. close to the time the sites begin to receive the intervention) [48,49].…”

Section: Methodsmentioning

confidence: 99%

Design, rationale and analysis plan for the Stand Up for Health trial in contact centres: a stepped wedge feasibility study

Parker

Manner

Sivaramakrishnan

et al. 2020

Pilot Feasibility Stud

View full text Add to dashboard Cite

Background Contact centres are one of the most sedentary workplaces, with employees spending a very high proportion of their working day sitting down. About a quarter of contact centre staff regularly experience musculoskeletal health problems due to high levels of sedentary behaviour, including lower back pain. There have been no previous randomised studies specifically aiming to reduce sedentary behaviour in contact centre staff. To address this gap, the Stand Up for Health (SUH) study aims to test the feasibility and acceptability of a complex theory-based intervention to reduce sedentary behaviour in contact centres. Methods The Stand Up for Health study has a stepped wedge cluster randomised trial design, which is a pragmatic design whereby clusters (contact centres) are randomised to time points at which they will begin to receive the intervention. All contact centre staff have the opportunity to experience the intervention. To minimise the resource burden in this feasibility study, data collection is not continuous, but undertaken on a selective number of occasions, so the stepped wedge design is “incomplete”. Eleven contact centres in England and Scotland have been recruited, and the sample size is approximately 27 per centre (270 in total). The statistical analysis will predominantly focus on assessing feasibility, including the calculation of recruitment rates and rates of attrition. Exploratory analysis will be performed to compare objectively measured sedentary time in the workplace (measured using an activPAL™ device) between intervention and control conditions using a linear mixed effects regression model. Discussion To our knowledge, this is the first stepped wedge feasibility study conducted in call centres. The rationale and justification of our novel staircase stepped wedge design has been presented, and we hope that by presenting our study design and statistical analysis plan, it will contribute to the literature on stepped wedge trials, and in particular feasibility stepped wedge trials. The findings of the study will also help inform whether this is a suitable design for other settings where data collection is challenging. Trial registration The trial has been registered on the ISRCTN database: http://www.isrctn.com/ISRCTN11580369

show abstract

“…As has been shown previously (eg, Kasza et al), the variance of the generalized least‐squares estimator of θ is given by:

\begin{align} normalvar ((), \overset{θ}{true}) = {({}, true \sum_{k = 1}^{K} {boldX}_{k}^{T} normalvar {false({\overset{boldY}{true}}_{k} false)}^{prefix- 1} {boldX}_{k} prefix- true \sum_{k = 1}^{K} {boldX}_{k}^{T} normalvar {false({\overset{boldY}{true}}_{k} false)}^{prefix- 1} {([], true \sum_{k = 1}^{K} normalvar {false({\overset{boldY}{true}}_{k} false)}^{prefix- 1})}^{prefix- 1} true \sum_{k = 1}^{K} normalvar {false({\overset{boldY}{true}}_{k} false)}^{prefix- 1} {boldX}_{k})}^{prefix- 1}, \end{align}

where

X_{k}^{T} = (X_{k 1}, \dots, X_{k T})

is the vector of treatment assignments for cluster k , and

var ({\overline{Y}}_{k})

is the T × T variance matrix of the vector

{\overline{Y}}_{k} = {({\overline{Y}}_{k 1 •}, \dots, {\overline{Y}}_{k T •})}^{T}

with elements …”

Section: Sample Size Formulas For Open Cohort Longitudinal Cluster Ramentioning

confidence: 99%

Sample size and power calculations for open cohort longitudinal cluster randomized trials

Kasza

Hooper

Copas

et al. 2020

Statistics in Medicine

Self Cite

View full text Add to dashboard Cite

When calculating sample size or power for stepped wedge or other types of longitudinal cluster randomized trials, it is critical that the planned sampling structure be accurately specified. One common assumption is that participants will provide measurements in each trial period, that is, a closed cohort, and another is that each participant provides only one measurement during the course of the trial. However some studies have an "open cohort" sampling structure, where participants may provide measurements in variable numbers of periods. To date, sample size calculations for longitudinal cluster randomized trials have not accommodated open cohorts. Feldman and McKinlay (1994) provided some guidance, stating that the participant-level autocorrelation could be varied to account for the degree of overlap in different periods of the study, but did not indicate precisely how to do so. We present sample size and power formulas that allow for open cohorts and discuss the impact of the degree of "openness" on sample size and power. We consider designs where the number of participants in each cluster will be maintained throughout the trial, but individual participants may provide differing numbers of measurements. Our results are a unification of closed cohort and repeated cross-sectional sample results of Hooper et al (2016), and indicate precisely how participant autocorrelation of Feldman and McKinlay should be varied to account for an open cohort sampling structure. We discuss different types of open cohort sampling schemes and how open cohort sampling structure impacts on power in the presence of decaying within-cluster correlations and autoregressive participant-level errors. K E Y W O R D Scluster crossover trial, intra cluster correlation, mixed effects models, stepped wedge INTRODUCTIONCluster randomized trials are randomized trials in which clusters of participants, rather than the participants themselves, are randomized to particular treatments. 1 Longitudinal cluster randomized trials extend standard cluster randomized trials in time: clusters are now randomized to a sequence of treatments, and may switch between intervention and control conditions over the course of the trial. 2,3 Particular examples of such trials include cluster randomized cross-over trials, 4 stepped wedge trials, 5 or even parallel cluster trial designs, in which measurements are taken at several time pointsThis is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

Information content of stepped‐wedge designs when treatment effect heterogeneity and/or implementation periods are present

Cited by 20 publications

References 20 publications

The hunt for efficient, incomplete designs for stepped wedge trials with continuous recruitment and continuous outcome measures

The hunt for efficient, incomplete designs for stepped wedge trials with continuous recruitment and continuous outcome measures

Design, rationale and analysis plan for the Stand Up for Health trial in contact centres: a stepped wedge feasibility study

Sample size and power calculations for open cohort longitudinal cluster randomized trials

Contact Info

Product

Resources

About