Efficient Study Design Research Articles

Parametric Statistical Inference for Comparing Means and Variances.

PurposeThe purpose of this tutorial is to provide visual scientists with various approaches for comparing two or more groups of data using parametric statistical tests, which require that the distribution of data within each group is normal (Gaussian). Non-parametric tests are used for inference when the sample data are not normally distributed or the sample is too small to assess its true distribution.MethodsMethods are reviewed using retinal thickness, as measured by optical coherence tomography (OCT), as an example for comparing two or more group means. The following parametric statistical approaches are presented for different situations: two-sample t-test, Analysis of Variance (ANOVA), paired t-test, and the analysis of repeated measures data using a linear mixed-effects model approach.ResultsAnalyzing differences between means using various approaches is demonstrated, and follow-up procedures to analyze pairwise differences between means when there are more than two comparison groups are discussed. The assumption of equal variance between groups and methods to test for equal variances are examined. Examples of repeated measures analysis for right and left eyes on subjects, across spatial segments within the same eye (e.g. quadrants of each retina), and over time are given.ConclusionsThis tutorial outlines parametric inference tests for comparing means of two or more groups and discusses how to interpret the output from statistical software packages. Critical assumptions made by the tests and ways of checking these assumptions are discussed. Efficient study designs increase the likelihood of detecting differences between groups if such differences exist. Situations commonly encountered by vision scientists involve repeated measures from the same subject over time, measurements on both right and left eyes from the same subject, and measurements from different locations within the same eye. Repeated measurements are usually correlated, and the statistical analysis needs to account for the correlation. Doing this the right way helps to ensure rigor so that the results can be repeated and validated.

Cortical atrophy patterns of incident MCI subtypes in the Mayo Clinic Study of Aging.

IntroductionWe examined differences in cortical thickness in empirically derived mild cognitive impairment (MCI) subtypes in the Mayo Clinic Study of Aging.MethodsWe compared cortical thickness of four incident MCI subtypes (n = 192) to 1257 cognitive unimpaired individuals.ResultsThe subtle cognitive impairment cluster had atrophy in the entorhinal and parahippocampal cortex. The amnestic, dysnomic, and dysexecutive clusters also demonstrated entorhinal cortex atrophy as well as thinning in temporal, parietal, and frontal isocortex in somewhat different patterns.DiscussionWe found patterns of atrophy in each of the incident MCI clusters that corresponded to their patterns of cognitive impairment. The identification of MCI subtypes based on cognitive and structural features may allow for more efficient trial and study designs. Given individuals in the subtle cognitive impairment cluster have less structural changes and cognitive decline and may represent the earliest group, this could be a unique group to target with early interventions.

Historical Review of Combined Experimental and Computational Approaches for Investigating Annulus Fibrosus Mechanics.

Intervertebral disc research has sought to develop a deeper understanding of spine biomechanics, the complex relationship between disc health and back pain, and the mechanisms of spinal injury and repair. To do so, many researchers have focused on characterizing tissue-level properties of the disc, where the roles of tissue subcomponents can be more systematically investigated. Unfortunately, experimental challenges often limit the ability to measure important disc tissue- and subtissue-level behaviors, including fiber-matrix interactions, transient nutrient and electrolyte transport, and damage propagation. Numerous theoretical and numerical modeling frameworks have been introduced to explain, complement, guide, and optimize experimental research efforts. The synergy of experimental and computational work has significantly advanced the field, and these two aspects have continued to develop independently and jointly. Meanwhile, the relationship between experimental and computational work has become increasingly complex and interdependent. This has made it difficult to interpret and compare results between experimental and computational studies, as well as between solely computational studies. This paper seeks to explore issues of model translatability, robustness, and efficient study design, and to propose and motivate potential future directions for experimental, computational, and combined tissue-level investigations of the intervertebral disc.

Definition of Clinical Outcomes in Pediatric Anesthesia Research: It Is Like the Tower of Babel!

Definition of Clinical Outcomes in Pediatric Anesthesia Research: It Is Like the Tower of Babel!

Neuroanatomical spread of amyloid β and tau in Alzheimer's disease: implications for primary prevention.

With recent advances in our understanding of the continuous pathophysiological changes that begin many years prior to symptom onset, it is now apparent that Alzheimer’s disease cannot be adequately described by discrete clinical stages, but should also incorporate the continuum of biological changes that precede and underlie the clinical representation of the disease. By jointly considering longitudinal changes of all available biomarkers and clinical assessments, variation within individuals can be integrated into a single continuous measure of disease progression and used to identify the earliest pathophysiological changes. Disease time, a measure of disease severity, was estimated using a Bayesian latent time joint mixed-effects model applied to an array of imaging, biomarker and neuropsychological data. Trajectories of regional amyloid β and tau PET uptake were estimated as a function of disease time. Regions with early signs of elevated amyloid β uptake were used to form an early, focal composite and compared to a commonly used global composite, in a separate validation sample. Disease time was estimated in 279 participants (183 cognitively unimpaired individuals, 61 mild cognitive impairment and 35 Alzheimer’s disease dementia patients) with available amyloid β and tau PET data. Amyloid β PET uptake levels in the posterior cingulate and precuneus start high and immediately increase with small increases of disease time. Early elevation in tau PET uptake was found in the inferior temporal lobe, amygdala, banks of the superior temporal sulcus, entorhinal cortex, middle temporal lobe, inferior parietal lobe and the fusiform gyrus. In a separate validation sample of 188 cognitively unimpaired individuals, the early, focal amyloid β PET composite showed a 120% increase in the accumulation rate of amyloid β compared to the global composite (P < 0.001), resulting in a 60% increase in the power to detect a treatment effect in a primary prevention trial design. Ordering participants on a continuous disease time scale facilitates the inspection of the earliest signs of amyloid β and tau pathology. To detect early changes in amyloid β pathology, focusing on the earliest sites of amyloid β accumulation results in more powerful and efficient study designs in early Alzheimer’s disease. Targeted composites could be used to re-examine the thresholds for amyloid β-related study inclusion, especially as the field shifts to focus on primary and secondary prevention. Clinical trials of anti-amyloid β treatments may benefit from the use of focal composites when estimating drug effects on amyloid β and tau changes in populations with minimal amyloid β and tau pathology and limited expected short-term accumulation.

How Often Do Orthopaedic Matched Case-Control Studies Use Matched Methods? A Review of Methodological Quality.

Case-control studies are a common method of analyzing associations between clinical outcomes and potential risk factors. Matching cases to controls based on known confounding variables can decrease bias and allow investigators to assess the association of interest with increased precision. However, the analysis of matched data generally requires matched statistical methods, and failure to use these methods can lead to imprecise or biased results. The appropriate use of matched statistical methods in orthopaedic case-control studies has not been documented. (1) What proportion of matched orthopaedic case-control studies use the appropriate matched statistical analyses? (2) What study factors are associated with the use of appropriate matched statistical tests? All matched case-control studies published in the top 10 orthopaedic journals according to impact factor from 2007 to 2016 were identified by literature review. Studies using appropriate statistical techniques were identified by two independent evaluators; discrepancies were settled by a third evaluator, all with advanced training in biostatistics. The number of studies using appropriate matched statistical methods was compared with the number of studies reviewed. Logistic regression was used to identify key study factors (including journal, publication year, rank according to impact factor, number of matching factors, number of controls per case, and the inclusion of a biostatistician coauthor) associated with the use of appropriate statistical methods. Three hundred nineteen articles that were initially classified as case-control studies were screened, yielding 83 matched case-control studies. One hundred two of the excluded articles were cohort or cross-sectional studies that were misclassified as case-control studies. The median number of matching factors was 3.0 (range, 1-10) and the median number of controls per case was 1.0 (range, 0.5-6.0). Thirty studies (36%) had a statistician coauthor. Thirty of the 83 included studies (36%) used appropriately matched methods throughout, 11 (13%) used matched methods for multivariable but not univariable analyses, and 42 (51%) used only unmatched methods, which we considered inappropriate. After controlling for the number of controls per case and publication year, we found that the inclusion of a statistician coauthor (70% versus 38%; odds ratio, 3.6; 95% confidence interval, 1.4-20.3; p = 0.01) and journal were associated with the use of appropriate methods. Although matched case-control studies can be statistically more efficient study designs, in that they are capable of generating more precise effect size estimates than other kinds of retrospective research, most orthopaedic case-control studies use inappropriate statistical methods in their analyses. Additionally, the high degree of study misclassification indicates a need to more rigorously define differences among case-control, cohort, and cross-sectional study designs. Failing to use matched statistical tests may lead to imprecise and/or biased effect estimates, which may lead to a tendency to overestimate or underestimate associations between possible risk factors and clinically relevant outcomes. Orthopaedic researchers should be cognizant of the risks and benefits of matching and should consult individuals with biostatistical expertise as needed to ensure that their statistical methods are appropriate and methodologically rigorous.

What makes a biostatistician?

Biostatisticians play an important role in medical research. They are co-responsible for an appropriate and efficient study design, they are involved in monitoring the study conduct, they plan and perform the data analysis, and they are involved in interpreting and publishing the results. However, how are the biostatisticians prepared for their tasks and responsibilities? Graduate programs in biostatistics are being offered, but some practicing biostatisticians completed their studies in a mathematical or epidemiological program, or obtained their degree in subject-specific fields (such as medicine or biology). Therefore, the expertise and the competencies can vary widely between the individual biostatisticians, also depending on the application field. In this article, focusing on European and US practices, we discuss the required professional expertise for the main areas of applications in the medical field as well as the necessary soft skill competencies of a biostatistician.

Designing cancer immunotherapy trials with random treatment time-lag effect.

In some clinical settings such as the cancer immunotherapy trials, a treatment time-lag effect may be present and the lag duration possibly vary from subject to subject. An efficient study design and analysis procedure should not only take into account the time-lag effect but also consider the individual heterogeneity in the lag duration. In this paper, we present a Generalized Piecewise Weighted Logrank (GPW-Logrank) test, designed to account for the random time-lag effect while maximizing the study power with respect to the weights. Based on the proposed test, both analytic and numeric approaches are developed for the sample size and power calculation. Asymptotic properties are derived and finite sample efficiency is evaluated in simulations. Compared with the standard practice ignoring the delayed effect, the proposed design and analysis procedures are substantially more efficient when a random lag is expected; further, compared with the existing methods by Xu et al considering the fixed time-lag effect, the proposed approaches are significantly more robust when the lag model is misspecified. An R package (DelayedEffect.Design) is developed for implementation.

Bayesian adaptive trials for rare cardiovascular conditions.

Escalating costs of cardiovascular trials are limiting medical innovations, prompting the development of more efficient and flexible study designs. The Bayesian paradigm offers a framework conducive to adaptive trial methodologies and is well suited for the study of small populations. Bayesian adaptive trials provide a statistical structure for combining prior information with accumulating data to compute probabilities of unknown quantities of interest. Adaptive design features are useful in modifying randomization schemes, adjusting sample sizes and providing continuous surveillance to guide decisions on dropping study arms or premature trial interruption. Advantages include greater efficiency, minimization of risks, inclusion of knowledge as it is generated, cost savings and more intuitive interpretability. Extensive high-level computations are facilitated by an expanding armamentarium of available tools.

The Socio-Moral Image Database (SMID): A novel stimulus set for the study of social, moral and affective processes.

A major obstacle for the design of rigorous, reproducible studies in moral psychology is the lack of suitable stimulus sets. Here, we present the Socio-Moral Image Database (SMID), the largest standardized moral stimulus set assembled to date, containing 2,941 freely available photographic images, representing a wide range of morally (and affectively) positive, negative and neutral content. The SMID was validated with over 820,525 individual judgments from 2,716 participants, with normative ratings currently available for all images on affective valence and arousal, moral wrongness, and relevance to each of the five moral values posited by Moral Foundations Theory. We present a thorough analysis of the SMID regarding (1) inter-rater consensus, (2) rating precision, and (3) breadth and variability of moral content. Additionally, we provide recommendations for use aimed at efficient study design and reproducibility, and outline planned extensions to the database. We anticipate that the SMID will serve as a useful resource for psychological, neuroscientific and computational (e.g., natural language processing or computer vision) investigations of social, moral and affective processes. The SMID images, along with associated normative data and additional resources are available at https://osf.io/2rqad/.

Pharmacoepidemiology in Pharmacogenetics.

Epidemiologic methods provide rigorous means by which to study the interplay between genetic factors and drug response. In this chapter, we describe the differences between experimental and observational study designs, and illustrate how to implement the highly efficient case-control study design. We discuss analytic approaches to evaluating gene-drug interactions within typical study designs and review sources of bias that must be assessed and accounted for in epidemiologic analyses.

An efficient study design to test parent-of-origin effects in family trios.

Increasing evidence has shown that genes may cause prenatal, neonatal, and pediatric diseases depending on their parental origins. Statistical models that incorporate parent-of-origin effects (POEs) can improve the power of detecting disease-associated genes and help explain the missing heritability of diseases. In many studies, children have been sequenced for genome-wide association testing. But it may become unaffordable to sequence their parents and evaluate POEs. Motivated by the reality, we proposed a budget-friendly study design of sequencing children and only genotyping their parents through single nucleotide polymorphism array. We developed a powerful likelihood-based method, which takes into account both sequence reads and linkage disequilibrium to infer the parental origins of children's alleles and estimate their POEs on the outcome. We evaluated the performance of our proposed method and compared it with an existing method using only genotypes, through extensive simulations. Our method showed higher power than the genotype-based method. When either the mean read depth or the pair-end length was reasonably large, our method achieved ideal power. When single parents' genotypes were unavailable or parental genotypes at the testing locus were not typed, both methods lost power compared with when complete data were available; but the power loss from our method was smaller than the genotype-based method. We also extended our method to accommodate mixed genotype, low-, and high-coverage sequence data from children and their parents. At presence of sequence errors, low-coverage parental sequence data may lead to lower power than parental genotype data.

A randomized trial of selenium and vitamin E for primary prevention of non-melanoma skin cancer: Trial results and experience from a low-resource setting.

1511 Background: Selenium and vitamin E supplementation previously have shown some evidence of a beneficial effect in the secondary prevention of non-melanoma skin cancer (NMSC). More than 100 million people worldwide are at increased risk of NMSC and other cancers due to arsenic exposure from drinking water and diet – a risk that persists for several decades even after the exposure has terminated. Here, we report on the design, methods and main results of the Bangladesh Vitamin E and Selenium Trial (BEST) – a population-based chemoprevention study conducted among 7,000 adults with visible arsenic toxicity. Methods: BEST is a 2 × 2 full factorial, double-blind, randomized placebo controlled trial of 7,000 adults having manifest arsenical skin lesions evaluating the efficacy of 6-year supplementation with alpha-tocopherol (100 mg daily) and L-selenomethionine (200 μg daily) for the prevention of NMSC incidence. Results: Excellent compliance was maintained through the course of the trial, based on data from pill counts, self-reported compliance, and bioadherence. Among participants on treatment through the end of the 6-year intervention period, &gt; 85% were adherent to at least 80% of study supplements. More than 500 new NMSC cases were ascertained using a three-tiered biopsy protocol. No significant beneficial effects were observed on NMSC incidence during the study period for selenium or vitamin E. Among more than 500 observed deaths (including 182 cancer-related deaths), there were also no significant treatment effects on total mortality, cancer-related mortality or arsenical cancer-related mortality. Conclusions: This large population-based trial does not support any beneficial effect of vitamin E or selenium supplementation for the primary prevention of NMSC or mortality in an arsenic-exposed population. With the rapidly increasing burden of preventable cancers in low- and middle-income countries, efficient and feasible chemoprevention study designs and approaches, such as employed in Bangladesh, may prove impactful in conceiving large scale cancer chemoprevention trials in low-resource settings. Clinical trial information: NCT00392561.

Sample size considerations for trials using cerebral white matter hyperintensity progression as an intermediate outcome at 1\xa0year after mild stroke: results of a prospective cohort study

BackgroundWhite matter hyperintensities (WMHs) are commonly seen on in brain imaging and are associated with stroke and cognitive decline. Therefore, they may provide a relevant intermediate outcome in clinical trials. WMH can be measured as a volume or visually on the Fazekas scale. We investigated predictors of WMH progression and design of efficient studies using WMH volume and Fazekas score as an intermediate outcome.MethodsWe prospectively recruited 264 patients with mild ischaemic stroke and measured WMH volume, Fazekas score, age and cardiovascular risk factors at baseline and 1 year. We modelled predictors of WMH burden at 1 year and used the results in sample size calculations for hypothetical randomised controlled trials with different analysis plans and lengths of follow-up.ResultsFollow-up WMH volume was predicted by baseline WMH: a 0.73-ml (95% CI 0.65–0.80, p < 0.0001) increase per 1-ml baseline volume increment, and a 2.93-ml increase (95% CI 1.76–4.10, p < 0.0001) per point on the Fazekas scale. Using a mean difference of 1 ml in WMH volume between treatment groups, 80% power and 5% alpha, adjusting for all predictors and 2-year follow-up produced the smallest sample size (n = 642). Other study designs produced samples sizes from 2054 to 21,270. Sample size calculations using Fazekas score as an outcome with the same power and alpha, as well as an OR corresponding to a 1-ml difference, were sensitive to assumptions and ranged from 2504 to 18,886.ConclusionsBaseline WMH volume and Fazekas score predicted follow-up WMH volume. Study size was smallest using volumes and longer-term follow-up, but this must be balanced against resources required to measure volumes versus Fazekas scores, bias due to dropout and scanner drift. Samples sizes based on Fazekas scores may be best estimated with simulation studies.

The Drosophila Gene Expression Tool (DGET) for expression analyses

BackgroundNext-generation sequencing technologies have greatly increased our ability to identify gene expression levels, including at specific developmental stages and in specific tissues. Gene expression data can help researchers understand the diverse functions of genes and gene networks, as well as help in the design of specific and efficient functional studies, such as by helping researchers choose the most appropriate tissue for a study of a group of genes, or conversely, by limiting a long list of gene candidates to the subset that are normally expressed at a given stage or in a given tissue.ResultsWe report DGET, a Drosophila Gene Expression Tool (www.flyrnai.org/tools/dget/web/), which stores and facilitates search of RNA-Seq based expression profiles available from the modENCODE consortium and other public data sets. Using DGET, researchers are able to look up gene expression profiles, filter results based on threshold expression values, and compare expression data across different developmental stages, tissues and treatments. In addition, at DGET a researcher can analyze tissue or stage-specific enrichment for an inputted list of genes (e.g., ‘hits’ from a screen) and search for additional genes with similar expression patterns. We performed a number of analyses to demonstrate the quality and robustness of the resource. In particular, we show that evolutionary conserved genes expressed at high or moderate levels in both fly and human tend to be expressed in similar tissues. Using DGET, we compared whole tissue profile and sub-region/cell-type specific datasets and estimated a potential source of false positives in one dataset. We also demonstrated the usefulness of DGET for synexpression studies by querying genes with expression profile similar to the mesodermal master regulator Twist.ConclusionAltogether, DGET provides a flexible tool for expression data retrieval and analysis with short or long lists of Drosophila genes, which can help scientists to design stage- or tissue-specific in vivo studies and do other subsequent analyses.

Defining Cancer Subtypes With Distinctive Etiologic Profiles: An Application to the Epidemiology of Melanoma

ABSTRACTWe showcase a novel analytic strategy to identify subtypes of cancer that possess distinctive causal factors, that is, subtypes that are “etiologically” distinct. The method involves the integrated analysis of two types of study design: an incident series of cases with double primary cancers with detailed information on tumor characteristics that can be used to define the subtypes; a case-series of incident cases with information on known risk factors that can be used to investigate the specific risk factors that distinguish the subtypes. The methods are applied to a rich melanoma dataset with detailed information on pathologic tumor factors, and comprehensive information on known genetic and environmental risk factors for melanoma. Identification of the optimal subtyping solution is accomplished using a novel clustering analysis that seeks to maximize a measure that characterizes the distinctiveness of the distributions of risk factors across the subtypes and that is a function of the correlations of tumor factors in the case-specific tumor pairs. This analysis is challenged by the presence of extensive missing data. If successful, studies of this nature offer the opportunity for efficient study design to identify unknown risk factors whose effects are concentrated in defined subtypes. Supplementary materials for this article are available online.

PLEASANT: efficient study design for economic analysis

PLEASANT: efficient study design for economic analysis

Guidelines for Large-Scale Sequence-Based Complex Trait Association Studies: Lessons Learned from the NHLBI Exome Sequencing Project

Massively parallel whole-genome sequencing (WGS) data have ushered in a new era in human genetics. These data are now being used to understand the role of rare variants in complex traits and to advance the goals of precision medicine. The technological and computing advances that have enabled us to generate WGS data on thousands of individuals have also outpaced our ability to perform analyses in scientifically and statistically rigorous and thoughtful ways. The past several years have witnessed the application of whole-exome sequencing (WES) to complex traits and diseases. From our analysis of NHLBI Exome Sequencing Project (ESP) data, not only have a number of important disease and complex trait association findings emerged, but our collective experience offers some valuable lessons for WGS initiatives. These include caveats associated with generating automated pipelines for quality control and analysis of rare variants; the importance of studying minority populations; sample size requirements and efficient study designs for identifying rare-variant associations; and the significance of incidental findings in population-based genetic research. With the ESP as an example, we offer guidance and a framework on how to conduct a large-scale association study in the era of WGS.

Born in Bradford's Better Start: an experimental birth cohort study to evaluate the impact of early life interventions.

BackgroundEarly interventions are recognised as key to improving life chances for children and reducing inequalities in health and well-being, however there is a paucity of high quality research into the effectiveness of interventions to address childhood health and development outcomes. Planning and implementing standalone RCTs for multiple, individual interventions would be slow, cumbersome and expensive. This paper describes the protocol for an innovative experimental birth cohort: Born in Bradford’s Better Start (BiBBS) that will simultaneously evaluate the impact of multiple early life interventions using efficient study designs. Better Start Bradford (BSB) has been allocated £49 million from the Big Lottery Fund to implement 22 interventions to improve outcomes for children aged 0–3 in three key areas: social and emotional development; communication and language development; and nutrition and obesity. The interventions will be implemented in three deprived and ethnically diverse inner city areas of Bradford.MethodThe BiBBS study aims to recruit 5000 babies, their mothers and their mothers’ partners over 5 years from January 2016-December 2020. Demographic and socioeconomic information, physical and mental health, lifestyle factors and biological samples will be collected during pregnancy. Parents and children will be linked to their routine health and local authority (including education) data throughout the children’s lives. Their participation in BSB interventions will also be tracked. BiBBS will test interventions using the Trials within Cohorts (TwiCs) approach and other quasi-experimental designs where TwiCs are neither feasible nor ethical, to evaluate these early life interventions. The effects of single interventions, and the cumulative effects of stacked (multiple) interventions on health and social outcomes during the critical early years will be measured.DiscussionThe focus of the BiBBS cohort is on intervention impact rather than observation. As far as we are aware BiBBS is the world’s first such experimental birth cohort study. While some risk factors for adverse health and social outcomes are increasingly well described, the solutions to tackling them remain elusive. The novel design of BiBBS can contribute much needed evidence to inform policy makers and practitioners about effective approaches to improve health and well-being for future generations.Electronic supplementary materialThe online version of this article (doi:10.1186/s12889-016-3318-0) contains supplementary material, which is available to authorized users.

Next-generation sequencing for virus detection: covering all the bases.

BackgroundThe use of next-generation sequencing has become an established method for virus detection. Efficient study design for accurate detection relies on the optimal amount of data representing a significant portion of a virus genome.FindingsIn this study, genome coverage at different sequencing depths was determined for a number of viruses, viroids, hosts and sequencing library types, using both read-mapping and de novo assembly-based approaches. The results highlighted the strength of ribo-depleted RNA and sRNA in obtaining saturated genome coverage with the least amount of data, while even though the poly(A)-selected RNA yielded virus-derived reads, it was insufficient to cover the complete genome of a non-polyadenylated virus. The ribo-depleted RNA data also outperformed the sRNA data in terms of the percentage of coverage that could be obtained particularly with the de novo assembled contigs.ConclusionOur results suggest the use of ribo-depleted RNA in a de novo assembly-based approach for the detection of single-stranded RNA viruses. Furthermore, we suggest that sequencing one million reads will provide sufficient genome coverage specifically for closterovirus detection.Electronic supplementary materialThe online version of this article (doi:10.1186/s12985-016-0539-x) contains supplementary material, which is available to authorized users.