Lessons from meta‐analyses of randomized clinical trials for analysis of distributed networks of observational databases

Networks of constellations of longitudinal observational databases, often electronic medical records or transactional insurance claims or both, are increasingly being used for studying the effects of medicinal products in real‐world use. Such databases are frequently configured as distributed networks. That is, patient‐level data are kept behind firewalls and not communicated outside of the data vendor other than in aggregate form. Instead, data are standardized across the network, and queries of the network are executed locally by data partners, and summary results provided to a central research partner(s) for amalgamation, aggregation, and summarization. Such networks can be huge covering years of data on upwards of 100 million patients. Examples of such networks include the FDA Sentinel Network, ASPEN, CNODES, and EU‐ADR. As this is a new emerging field, we note in this paper the conceptual similarities and differences between the analysis of distributed networks and the now well‐established field of meta‐analysis of randomized clinical trials (RCTs). We recommend, wherever appropriate, to apply learnings from meta‐analysis to help guide the development of distributed network analyses of longitudinal observational databases.     

Estimating the variance for heterogeneity in arm‐based network meta‐analysis

Network meta‐analysis can be implemented by using arm‐based or contrast‐based models. Here we focus on arm‐based models and fit them using generalized linear mixed model procedures. Full maximum likelihood (ML) estimation leads to biased trial‐by‐treatment interaction variance estimates for heterogeneity. Thus, our objective is to investigate alternative approaches to variance estimation that reduce bias compared with full ML. Specifically, we use penalized quasi‐likelihood/pseudo‐likelihood and hierarchical (h) likelihood approaches. In addition, we consider a novel model modification that yields estimators akin to the residual maximum likelihood estimator for linear mixed models. The proposed methods are compared by simulation, and 2 real datasets are used for illustration. Simulations show that penalized quasi‐likelihood/pseudo‐likelihood and h‐likelihood reduce bias and yield satisfactory coverage rates. Sum‐to‐zero restriction and baseline contrasts for random trial‐by‐treatment interaction effects, as well as a residual ML‐like adjustment, also reduce bias compared with an unconstrained model when ML is used, but coverage rates are not quite as good. Penalized quasi‐likelihood/pseudo‐likelihood and h‐likelihood are therefore recommended.     

Adaptive and repeated cumulative meta‐analyses of safety data during a new drug development process

During a new drug development process, it is desirable to timely detect potential safety signals. For this purpose, repeated meta‐analyses may be performed sequentially on accumulating safety data. Moreover, if the amount of safety data from the originally planned program is not enough to ensure adequate power to test a specific hypothesis (e.g., the noninferiority hypothesis of an event of interest), the total sample size may be increased by adding new studies to the program. Without appropriate adjustment, it is well known that the type I error rate will be inflated because of repeated analyses and sample size adjustment. In this paper, we discuss potential issues associated with adaptive and repeated cumulative meta‐analyses of safety data conducted during a drug development process. We consider both frequentist and Bayesian approaches. A new drug development example is used to demonstrate the application of the methods. Copyright © 2015 John Wiley & Sons, Ltd.     

Bayesian robustness in meta‐analysis for studies with zero responses

Statistical meta‐analysis is mostly carried out with the help of the random effect normal model, including the case of discrete random variables. We argue that the normal approximation is not always able to adequately capture the underlying uncertainty of the original discrete data. Furthermore, when we examine the influence of the prior distributions considered, in the presence of rare events, the results from this approximation can be very poor. In order to assess the robustness of the quantities of interest in meta‐analysis with respect to the choice of priors, this paper proposes an alternative Bayesian model for binomial random variables with several zero responses. Particular attention is paid to the coherence between the prior distributions of the study model parameters and the meta‐parameter. Thus, our method introduces a simple way to examine the sensitivity of these quantities to the structure dependence selected for study. For illustrative purposes, an example with real data is analysed, using the proposed Bayesian meta‐analysis model for binomial sparse data. Copyright © 2016 John Wiley & Sons, Ltd.     

A note regarding meta‐analysis of sequential trials with stopping for efficacy

It is shown that fixed‐effect meta‐analyses of naïve treatment estimates from sequentially run trials with the possibility of stopping for efficacy based on a single interim look are unbiassed (or at the very least consistent, depending on the point of view) provided that the trials are weighted by information provided. A simple proof of this is given. An argument is given suggesting that this also applies in the case of multiple looks. The implications for this are discussed. Copyright © 2014 John Wiley & Sons, Ltd.     

An investigation into the two‐stage meta‐analytic copula modelling approach for evaluating time‐to‐event surrogate endpoints which comprise of one or more events of interest

Clinical trials of experimental treatments must be designed with primary endpoints that directly measure clinical benefit for patients. In many disease areas, the recognised gold standard primary endpoint can take many years to mature, leading to challenges in the conduct and quality of clinical studies. There is increasing interest in using shorter‐term surrogate endpoints as substitutes for costly long‐term clinical trial endpoints; such surrogates need to be selected according to biological plausibility, as well as the ability to reliably predict the unobserved treatment effect on the long‐term endpoint. A number of statistical methods to evaluate this prediction have been proposed; this paper uses a simulation study to explore one such method in the context of time‐to‐event surrogates for a time‐to‐event true endpoint. This two‐stage meta‐analytic copula method has been extensively studied for time‐to‐event surrogate endpoints with one event of interest, but thus far has not been explored for the assessment of surrogates which have multiple events of interest, such as those incorporating information directly from the true clinical endpoint. We assess the sensitivity of the method to various factors including strength of association between endpoints, the quantity of data available, and the effect of censoring. In particular, we consider scenarios where there exist very little data on which to assess surrogacy. Results show that the two‐stage meta‐analytic copula method performs well under certain circumstances and could be considered useful in practice, but demonstrates limitations that may prevent universal use.     

Decision‐making in early clinical drug development

This paper illustrates an approach to setting the decision framework for a study in early clinical drug development. It shows how the criteria for a go and a stop decision are calculated based on pre‐specified target and lower reference values. The framework can lead to a three‐outcome approach by including a consider zone; this could enable smaller studies to be performed in early development, with other information either external to or within the study used to reach a go or stop decision. In this way, Phase I/II trials can be geared towards providing actionable decision‐making rather than the traditional focus on statistical significance. The example provided illustrates how the decision criteria were calculated for a Phase II study, including an interim analysis, and how the operating characteristics were assessed to ensure the decision criteria were robust. Copyright © 2016 John Wiley & Sons, Ltd.     

Decision‐making in drug development using a composite definition of success

Evidence‐based quantitative methodologies have been proposed to inform decision‐making in drug development, such as metrics to make go/no‐go decisions or predictions of success, identified with statistical significance of future clinical trials. While these methodologies appropriately address some critical questions on the potential of a drug, they either consider the past evidence without predicting the outcome of the future trials or focus only on efficacy, failing to account for the multifaceted aspects of a successful drug development. As quantitative benefit‐risk assessments could enhance decision‐making, we propose a more comprehensive approach using a composite definition of success based not only on the statistical significance of the treatment effect on the primary endpoint but also on its clinical relevance and on a favorable benefit‐risk balance in the next pivotal studies. For one drug, we can thus study several development strategies before starting the pivotal trials by comparing their predictive probability of success. The predictions are based on the available evidence from the previous trials, to which new hypotheses on the future development could be added. The resulting predictive probability of composite success provides a useful summary to support the discussions of the decision‐makers. We present a fictive, but realistic, example in major depressive disorder inspired by a real decision‐making case.     

A review and re‐interpretation of a group‐sequential approach to sample size re‐estimation in two‐stage trials

In this paper, we review the adaptive design methodology of Li et al. (Biostatistics 3 :277–287) for two‐stage trials with mid‐trial sample size adjustment. We argue that it is closer in principle to a group sequential design, in spite of its obvious adaptive element. Several extensions are proposed that aim to make it even more attractive and transparent alternative to a standard (fixed sample size) trial for funding bodies to consider. These enable a cap to be put on the maximum sample size and for the trial data to be analysed using standard methods at its conclusion. The regulatory view of trials incorporating unblinded sample size re‐estimation is also discussed. © 2014 The Authors. Pharmaceutical Statistics published by John Wiley & Sons, Ltd.     

Integrated Likelihood Inference in Small Sample Meta‐analysis for Continuous Outcomes

This paper proposes the use of the integrated likelihood for inference on the mean effect in small‐sample meta‐analysis for continuous outcomes. The method eliminates the nuisance parameters given by variance components through integration with respect to a suitable weight function, with no need to estimate them. The integrated likelihood approach takes into proper account the estimation uncertainty of within‐study variances, thus providing confidence intervals with empirical coverage closer to nominal levels than standard likelihood methods. The improvement is remarkable when either (i) the number of studies is small to moderate or (ii) the small sample size of the studies does not allow to consider the within‐study variances as known, as common in applications. Moreover, the use of the integrated likelihood avoids numerical pitfalls related to the estimation of variance components which can affect alternative likelihood approaches. The proposed methodology is illustrated via simulation and applied to a meta‐analysis study in nutritional science.     

Equal‐tailed confidence intervals for comparison of rates

Several methods are available for generating confidence intervals for rate difference, rate ratio, or odds ratio, when comparing two independent binomial proportions or Poisson (exposure‐adjusted) incidence rates. Most methods have some degree of systematic bias in one‐sided coverage, so that a nominal 95% two‐sided interval cannot be assumed to have tail probabilities of 2.5% at each end, and any associated hypothesis test is at risk of inflated type I error rate. Skewness‐corrected asymptotic score methods have been shown to have superior equal‐tailed coverage properties for the binomial case. This paper completes this class of methods by introducing novel skewness corrections for the Poisson case and for odds ratio, with and without stratification. Graphical methods are used to compare the performance of these intervals against selected alternatives. The skewness‐corrected methods perform favourably in all situations—including those with small sample sizes or rare events—and the skewness correction should be considered essential for analysis of rate ratios. The stratified method is found to have excellent coverage properties for a fixed effects analysis. In addition, another new stratified score method is proposed, based on the t‐distribution, which is suitable for use in either a fixed effects or random effects analysis. By using a novel weighting scheme, this approach improves on conventional and modern meta‐analysis methods with weights that rely on crude estimation of stratum variances. In summary, this paper describes methods that are found to be robust for a wide range of applications in the analysis of rates.     

Control‐based imputation for sensitivity analyses in informative censoring for recurrent event data

In clinical trials, missing data commonly arise through nonadherence to the randomized treatment or to study procedure. For trials in which recurrent event endpoints are of interests, conventional analyses using the proportional intensity model or the count model assume that the data are missing at random, which cannot be tested using the observed data alone. Thus, sensitivity analyses are recommended. We implement the control‐based multiple imputation as sensitivity analyses for the recurrent event data. We model the recurrent event using a piecewise exponential proportional intensity model with frailty and sample the parameters from the posterior distribution. We impute the number of events after dropped out and correct the variance estimation using a bootstrap procedure. We apply the method to an application of sitagliptin study.     

Investigation of mixed model repeated measures analyses and non‐linear random coefficient models in the context of long‐term efficacy data

The longitudinal data from 2 published clinical trials in adult subjects with upper limb spasticity (a randomized placebo‐controlled study [NCT01313299] and its long‐term open‐label extension [NCT01313312]) were combined. Their study designs involved repeat intramuscular injections of abobotulinumtoxinA (Dysport®), and efficacy endpoints were collected accordingly. With the objective of characterizing the pattern of response across cycles, Mixed Model Repeated Measures analyses and Non‐Linear Random Coefficient (NLRC) analyses were performed and their results compared. The Mixed Model Repeated Measures analyses, commonly used in the context of repeated measures with missing dependent data, did not involve any parametric shape for the curve of changes over time. Based on clinical expectations, the NLRC included a negative exponential function of the number of treatment cycles, with its asymptote and rate included as random coefficients in the model. Our analysis focused on 2 specific efficacy parameters reflecting complementary aspects of efficacy in the study population. A simulation study based on a similar study design was also performed to further assess the performance of each method under different patterns of response over time. This highlighted a gain of precision with the NLRC model, and most importantly the need for its assumptions to be verified to avoid potentially biased estimates. These analyses describe a typical situation and the conditions under which non‐linear mixed modeling can provide additional insights on the behavior of efficacy parameters over time. Indeed, the resulting estimates from the negative exponential NLRC can help determine the expected maximal effect and the treatment duration required to reach it.     

Iterative Bias Correction of the Cross‐Validation Criterion

Abstract. The cross‐validation (CV) criterion is known to be asecond‐order unbiased estimator of the risk function measuring the discrepancy between the candidate model and the true model, as well as the generalized information criterion (GIC) and the extended information criterion (EIC). In the present article, we show that the 2kth‐order unbiased estimator can be obtained using a linear combination from the leave‐one‐out CV criterion to the leave‐k‐out CV criterion. The proposed scheme is unique in that a bias smaller than that of a jackknife method can be obtained without any analytic calculation, that is, it is not necessary to obtain the explicit form of several terms in an asymptotic expansion of the bias. Furthermore, the proposed criterion can be regarded as a finite correction of a bias‐corrected CV criterion by using scalar coefficients in a bias‐corrected EIC obtained by the bootstrap iteration.     

Subgroup analysis and interpretation for phase 3 confirmatory trials: White paper of the EFSPI/PSI working group on subgroup analysis

Subgroup by treatment interaction assessments are routinely performed when analysing clinical trials and are particularly important for phase 3 trials where the results may affect regulatory labelling. Interpretation of such interactions is particularly difficult, as on one hand the subgroup finding can be due to chance, but equally such analyses are known to have a low chance of detecting differential treatment effects across subgroup levels, so may overlook important differences in therapeutic efficacy. EMA have therefore issued draft guidance on the use of subgroup analyses in this setting. Although this guidance provided clear proposals on the importance of pre‐specification of likely subgroup effects and how to use this when interpreting trial results, it is less clear which analysis methods would be reasonable, and how to interpret apparent subgroup effects in terms of whether further evaluation or action is necessary. A PSI/EFSPI Working Group has therefore been investigating a focused set of analysis approaches to assess treatment effect heterogeneity across subgroups in confirmatory clinical trials that take account of the number of subgroups explored and also investigating the ability of each method to detect such subgroup heterogeneity. This evaluation has shown that the plotting of standardised effects, bias‐adjusted bootstrapping method and SIDES method all perform more favourably than traditional approaches such as investigating all subgroup‐by‐treatment interactions individually or applying a global test of interaction. Therefore, these approaches should be considered to aid interpretation and provide context for observed results from subgroup analyses conducted for phase 3 clinical trials.     

A uniform saddlepoint expansion for the null‐distribution of the wilcoxon‐mann‐whitney statistic

The authors give the exact coefficient of 1/N in a saddlepoint approximation to the Wilcoxon‐Mann‐Whitney null‐distribution. This saddlepoint approximation is obtained from an Edgeworth approximation to the exponentially tilted distribution. Moreover, the rate of convergence of the relative error is uniformly of order O (1/N) in a large deviation interval as defined in Feller (1971). The proposed method for computing the coefficient of 1/N can be used to obtain the exact coefficients of 1/Nⁱ, for any i. The exact formulas for the cumulant generating function and the cumulants, needed for these results, are those of van Dantzig (1947‐1950).     

Estimation of the need for child care in Canada

Finding adequate child care is a serious problem for many Canadian parents. The purpose of this case study is to estimate the need for child care in Canada, including ihc portion of this need that may be hidden. We utilize the Family History Survey conducted by Statistics Canada (1984) to explore patterns of both met and unmet child-care needs, based, in the latter case, on varying assumptions about the degree of parents' desire for child care.