Search

Longitudinal studies are frequently used in medical research and involve collecting repeated measures on individuals over time. Observations from the same individual are invariably correlated and thus an analytic approach that accounts for this clustering by individual is required. While almost all research suffers from missing data, this can be particularly problematic in longitudinal studies as participation often becomes harder to maintain over time. Multiple imputation (MI) is widely used to handle missing data in such studies. When using MI, it is important that the imputation model is compatible with the proposed analysis model. In a longitudinal analysis, this implies that the clustering considered in the analysis model should be reflected in the imputation process. Several MI approaches have been proposed to impute incomplete longitudinal data, such as treating repeated measurements of the same variable as distinct variables or using generalized linear mixed imputation models. However, the uptake of these methods has been limited, as they require additional data manipulation and use of advanced imputation procedures. In this tutorial, we review the available MI approaches that can be used for handling incomplete longitudinal data, including where individuals are clustered within higher-level clusters. We illustrate implementation with replicable R and Stata code using a case study from the Childhood to Adolescence Transition Study., Comment: 35 pages, 3 figures

Background. The DURATIONS design has been recently proposed as a practical alternative to a standard two-arm non-inferiority design when the goal is to optimise some continuous aspect of treatment administration, e.g. duration or frequency, preserving efficacy but improving on secondary outcomes such as safety, costs or convenience. The main features of this design are that (i) it randomises patients to a moderate number of arms across the continuum and (ii) it uses a model to share information across arms. While papers published to date about the design have focused on analysis aspects, here we show how to design such a trial in practice. We use the REFINE-Lung trial as an example; this is a trial seeking the optimal frequency of immunotherapy treatment for non-small cell lung cancer patients. Because the aspect of treatment administration to optimise is frequency, rather than duration, we propose to rename the design as Multi-Arm Multi-Stage Response Over Continuous Intervention (MAMS-ROCI). Methods. We show how simulations can be used to design such a trial. We propose to use the ADEMP framework to plan such simulations, clearly specifying aims, data generating mechanisms, estimands, methods and performance measures before coding and analysing the simulations. We discuss the possible choices to be made using the REFINE-Lung trial as an example. Results. We describe all the choices made while designing the REFINE-Lung trial, and the results of the simulations performed. We justify our choice of total sample size based on these results. Conclusions. MAMS-ROCI trials can be designed using simulation studies that have to be carefully planned and conducted. REFINE-Lung has been designed using such an approach and we have shown how researchers could similarly design their own MAMS-ROCI trial., Comment: 25 pages, 1 table, 5 figures

Background: Multiple imputation is often used to reduce bias and gain efficiency when there is missing data. The most appropriate imputation method depends on the model the analyst is interested in fitting. Several imputation approaches have been proposed for when this model is a logistic regression model with an interaction term that contains a binary partially observed variable; however, it is not clear which performs best under certain parameter settings. Methods: Using 1000 simulations, each with 10,000 observations, under six data-generating mechanisms (DGM), we investigate the performance of four methods: (i) 'passive imputation', (ii) 'just another variable' (JAV), (iii) 'stratify-impute-append' (SIA), and (iv) 'substantive model compatible fully conditional specifica-tion' (SMCFCS). The application of each method is shown in an empirical example using England-based cancer registry data. Results: SMCFCS and SIA showed the least biased estimate of the coefficients for the fully, and partially, observed variable and the interaction term. SMCFCS and SIA showed good coverage and low relative error for all DGMs. SMCFCS had a large bias when there was a low prevalence of the fully observed variable in the interaction. SIA performed poorly when the fully observed variable in the interaction had a continuous underlying form. Conclusion: SMCFCS and SIA give consistent estimation for logistic regression models with an interaction term when data are missing at random, and either can be used in most analyses. SMCFCS performed better than SIA when the fully observed variable in the interaction had an underlying continuous form. Researchers should be cautious when using SMCFCS when there is a low prevalence of the fully observed variable in the interaction., Comment: 26 pages, 9 figures, 4 tables

Effective decision making from randomised controlled clinical trials relies on robust interpretation of the numerical results. However, the language we use to describe clinical trials can cause confusion both in trial design and in comparing results across trials. ACceptability Curve Estimation using Probability Above Threshold (ACCEPT) aids comparison between trials (even where of different designs) by harmonising reporting of results, acknowledging different interpretations of the results may be valid in different situations, and moving the focus from comparison to a pre-specified value to interpretation of the trial data. ACCEPT can be applied to historical trials or incorporated into statistical analysis plans for future analyses. An online tool enables ACCEPT on up to three trials simultaneously.

Background. Designing trials to reduce treatment duration is important in several therapeutic areas, including TB and antibiotics. We recently proposed a new randomised trial design to overcome some of the limitations of standard two-arm non-inferiority trials. This DURATIONS design involves randomising patients to a number of duration arms, and modelling the so-called duration-response curve. This article investigates the operating characteristics (type-1 and type-2 errors) of different statistical methods of drawing inference from the estimated curve. Methods. Our first estimation target is the shortest duration non-inferior to the control (maximum) duration within a specific risk difference margin. We compare different methods of estimating this quantity, including using model confidence bands, the delta method and bootstrap. We then explore the generalisability of results to estimation targets which focus on absolute event rates, risk ratio and gradient of the curve. Results. We show through simulations that, in most scenarios and for most of the estimation targets, using the bootstrap to estimate variability around the target duration leads to good results for DURATIONS design-appropriate quantities analogous to power and type-1 error. Using model confidence bands is not recommended, while the delta method leads to inflated type-1 error in some scenarios, particularly when the optimal duration is very close to one of the randomised durations. Conclusions. Using the bootstrap to estimate the optimal duration in a DURATIONS design has good operating characteristics in a wide range of scenarios, and can be used with confidence by researchers wishing to design a DURATIONS trial to reduce treatment duration. Uncertainty around several different targets can be estimated with this bootstrap approach., Comment: 4 figures, 1 table + additional material

Background. Non-inferiority (NI) trials are increasingly used to evaluate new treatments expected to have secondary advantages over standard of care, but similar efficacy on the primary outcome. When designing a NI trial with a binary primary outcome, the choice of effect measure for the NI margin has an important effect on sample size calculations; furthermore, if the control event risk observed is markedly different from that assumed, the trial can quickly lose power or the results become difficult to interpret. Methods. We propose a new way of designing NI trials to overcome the issues raised by unexpected control event risks by specifying a NI frontier, i.e. a curve defining the most appropriate non-inferiority margin for each possible value of control event risk. We propose a fixed arcsine difference frontier, the power-stabilising transformation for binary outcomes. We propose and compare three ways of designing a trial using this frontier. Results. Testing and reporting on the arcsine scale leads to results which are challenging to interpret clinically. Working on the arcsine scale generally requires a larger sample size compared to the risk difference scale. Therefore, working on the risk difference scale, modifying the margin after observing the control event risk, might be preferable, as it requires a smaller sample size. However, this approach tends to slightly inflate type I error rate; a solution is to use a lower significance level for testing. When working on the risk ratio scale, the same approach leads to power levels above the nominal one, maintaining type I error under control. Conclusions. Our proposed methods of designing NI trials using power-stabilising frontiers make trial design more resilient to unexpected values of the control event risk, at the only cost of requiring larger sample sizes when the goal is to report results on the risk difference scale., Comment: ~4000 words + 5 figures + 1 table + additional material (2 appendices, 1 table, 4 figures)

Background: trials to identify the minimal effective treatment duration are needed in different therapeutic areas, including bacterial infections, TB and Hepatitis--C. However, standard non-inferiority designs have several limitations, including arbitrariness of non-inferiority margins, choice of research arms and very large sample sizes. Methods: we recast the problem of finding an appropriate non-inferior treatment duration in terms of modelling the entire duration-response curve within a pre-specified range. We propose a multi-arm randomised trial design, allocating patients to different treatment durations. We use fractional polynomials and spline-based methods to flexibly model the duration-response curve. We compare different methods in terms of a scaled version of the area between true and estimated prediction curves. We evaluate sensitivity to key design parameters, including sample size, number and position of arms. Results: a total sample size of $\sim 500$ patients divided into a moderate number of equidistant arms (5-7) is sufficient to estimate the duration-response curve within a $5\%$ error margin in $95\%$ of the simulations. Fractional polynomials provide similar or better results than spline-based methods in most scenarios. Conclusions: our proposed practical randomised trial design is an alternative to standard non-inferiority designs, avoiding many of their limitations, and yet being fairly robust to different possible duration-response curves. The trial outcome is the whole duration-response curve, which could be used by clinicians and policy makers to make informed decisions, facilitating a move away from a forced binary hypothesis testing paradigm.

We present and compare multiple imputation methods for multilevel continuous and binary data where variables are systematically and sporadically missing. The methods are compared from a theoretical point of view and through an extensive simulation study motivated by a real dataset comprising multiple studies. Simulations are reproducible. The comparisons show why these multiple imputation methods are the most appropriate to handle missing values in a multilevel setting and why their relative performances can vary according to the missing data pattern, the multilevel structure and the type of missing variables. This study shows that valid inferences can only be obtained if the dataset gathers a large number of clusters. In addition, it highlights that heteroscedastic MI methods provide more accurate inferences than homoscedastic methods, which should be reserved for data with few individuals per cluster. Finally, the method of Quartagno and Carpenter (2016a) appears generally accurate for binary variables, the method of Resche-Rigon and White (2016) with large clusters, and the approach of Jolani et al. (2015) with small clusters.