Incorporating data received after a sequential trial has stopped into the final analysis: implementation and comparison of methods

In a sequential clinical trial, accrual of data on patients often continues after the stopping criterion for the study has been met. This is termed “overrunning.” Overrunning occurs mainly when the primary response from each patient is measured after some extended observation period. The objective of this article is to compare two methods of allowing for overrunning. In particular, simulation studies are reported that assess the two procedures in terms of how well they maintain the intended type I error rate. The effect on power resulting from the incorporation of “overrunning data” using the two procedures is evaluated.

Meta-analysis of individual patient data from randomised trials: a review of methods used in practice

Background: Meta-analyses based on individual patient data (IPD) are regarded as the gold standard for systematic reviews. However, the methods used for analysing and presenting results from IPD meta-analyses have received little discussion. Methods We review 44 IPD meta-analyses published during the years 1999–2001. We summarize whether they obtained all the data they sought, what types of approaches were used in the analysis, including assumptions of common or random effects, and how they examined the effects of covariates. Results: Twenty-four out of 44 analyses focused on time-to-event outcomes, and most analyses (28) estimated treatment effects within each trial and then combined the results assuming a common treatment effect across trials. Three analyses failed to stratify by trial, analysing the data is if they came from a single mega-trial. Only nine analyses used random effects methods. Covariate-treatment interactions were generally investigated by subgrouping patients. Seven of the meta-analyses included data from less than 80% of the randomized patients sought, but did not address the resulting potential biases. Conclusions: Although IPD meta-analyses have many advantages in assessing the effects of health care, there are several aspects that could be further developed to make fuller use of the potential of these time-consuming projects. In particular, IPD could be used to more fully investigate the influence of covariates on heterogeneity of treatment effects, both within and between trials. The impact of heterogeneity, or use of random effects, are seldom discussed. There is thus considerable scope for enhancing the methods of analysis and presentation of IPD meta-analysis.

Comparative costs of the commercial process in different construction procurement methods

The commercial process in construction projects is an expensive and highly variable overhead. Collaborative working practices carry many benefits, which are widely disseminated, but little information is available about their costs. Transaction Cost Economics is a theoretical framework that seeks explanations for why there are firms and how the boundaries of firms are defined through the “make-or-buy” decision. However, it is not a framework that offers explanations for the relative costs of procuring construction projects in different ways. The idea that different methods of procurement will have characteristically different costs is tested by way of a survey. The relevance of transaction cost economics to the study of commercial costs in procurement is doubtful. The survey shows that collaborative working methods cost neither more nor less than traditional methods. But the benefits of collaboration mean that there is a great deal of enthusiasm for collaboration rather than competition.

Fecal water as a non-invasive biomarker in nutritional intervention: comparison of preparation methods and refinement of different endpoints

The assessment of cellular effects by the aqueous phase of human feces (fecal water, FW) is a useful biomarker approach to study cancer risks and protective activities of food. In order to refine and develop the biomarker, different protocols of preparing FW were compared. Fecal waters were prepared by 3 methods: (A) direct centrifugation; (B) extraction of feces in PBS before centrifugation; and (C) centrifugation of lyophilized and reconstituted feces. Genotoxicity was determined in colon cells using the Comet assay. Selected samples were investigated for additional parameters related to carcinogenesis. Two of 7 FWs obtained by methods A and B were similarly genotoxic. Method B, however, yielded higher volumes of FW, allowing sterile filtration for long-term culture experiments. Four of 7 samples were non-genotoxic when prepared according to all 3 methods. FW from lyophilized feces and from fresh samples were equally genotoxic. FWs modulated cytotoxicity, paracellular permeability, and invasion, independent of their genotoxicity. All 3 methods of FW preparation can be used to assess genotoxicity. The higher volumes of FWobtained by preparation method B greatly enhance the perspectives of measuring different types of biological parameters and using these to disclose activities related to cancer development.

Methods for comparing and constraining models of dendritic spines

Physiological evidence using Infrared Video Microscopy during the uncaging of glutamate has proven the existence of excitable calcium ion channels in spine heads, highlighting the need for reliable models of spines. In this study we compare the three main methods of simulating excitable spines: Baer & Rinzel's Continuum (B&R) model, Coombes' Spike-Diffuse-Spike (SDS) model and paired cable and ion channel equations (Cable model). Tests are done to determine how well the models approximate each other in terms of speed and heights of travelling waves. Significant quantitative differences are found between the models: travelling waves in the SDS model in particular are found to travel at much lower speeds and sometimes much higher voltages than in the Cable or B&R models. Meanwhile qualitative differences are found between the B&R and SDS models over realistic parameter ranges. The cause of these differences is investigated and potential solutions proposed.

Limitations of small x resummation methods from F2data

We discuss several methods of calculating the DIS structure functions F2(x,Q2) based on BFKL-type small x resummations. Taking into account new HERA data ranging down to small xand low Q2, the pure leading order BFKL-based approach is excluded. Other methods based on high energy factorization are closer to conventional renormalization group equations. Despite several difficulties and ambiguities in combining the renormalization group equations with small x resummed terms, we find that a fit to the current data is hardly feasible, since the data in the low Q2 region are not as steep as the BFKL formalism predicts. Thus we conclude that deviations from the (successful) renormalization group approach towards summing up logarithms in 1/x are disfavoured by experiment.

A comparative review of dimension reduction methods in approximate Bayesian computation

Approximate Bayesian computation (ABC) methods make use of comparisons between simulated and observed summary statistics to overcome the problem of computationally intractable likelihood functions. As the practical implementation of ABC requires computations based on vectors of summary statistics, rather than full data sets, a central question is how to derive low-dimensional summary statistics from the observed data with minimal loss of information. In this article we provide a comprehensive review and comparison of the performance of the principal methods of dimension reduction proposed in the ABC literature. The methods are split into three nonmutually exclusive classes consisting of best subset selection methods, projection techniques and regularization. In addition, we introduce two new methods of dimension reduction. The first is a best subset selection method based on Akaike and Bayesian information criteria, and the second uses ridge regression as a regularization procedure. We illustrate the performance of these dimension reduction techniques through the analysis of three challenging models and data sets.

Portmanteau model diagnostics and tests for nonlinearity: a comparative Monte Carlo study of two alternative methods

This paper employs an extensive Monte Carlo study to test the size and power of the BDS and close return methods of testing for departures from independent and identical distribution. It is found that the finite sample properties of the BDS test are far superior and that the close return method cannot be recommended as a model diagnostic. Neither test can be reliably used for very small samples, while the close return test has low power even at large sample sizes