Perform Power Calculations Research Articles

Prophylactic antibiotics for excision of premalignant vulvar lesions: A pilot randomized controlled trial

No prospective data have been described to inform guidelines on antibiotic prophylaxis for partial vulvectomies. Thus, we conducted a single-center, pilot, double-blind randomized controlled trial to assess the effectiveness of prophylactic antibiotics to prevent wound complications after partial vulvectomies. Patients were randomly assigned 1:1 to preoperative antibiotics or no preoperative antibiotics. The primary outcome of 30-day postoperative wound complications occurred in 31 (62 %) of all patients, with no differences between groups. The most common wound complications were superficial separation (54.2 % antibiotic prophylaxis vs. 65.3 % no prophylaxis, p = 0.37) and surgical site infection (0 % antibiotic prophylaxis vs 7.7 % no prophylaxis, p = 0.49). However, this study was limited by differences in patient characteristics between the groups. This study provides data to perform power calculations for a trial examining the effect of preoperative antibiotics on surgical site infection.

Moderators of treatment effect of non-steroidal anti-inflammatory drugs for patients with (sub) acute low back pain: Protocol for a systematic review with individual participant data meta-analysis

A Cochrane review found that non-steroidal anti-inflammatory drugs (NSAIDs) are slightly more effective than placebo on acute and subacute low back pain (LBP) outcomes (pain intensity, disability, and global improvement). Our objectives are: (1) to assess the overall treatment effect of NSAIDs in adults with acute and subacute LBP; (2) to identify the moderation of baseline patients’ characteristics on treatment effect. We will conduct a systematic search of RCTs on effectiveness of NSAIDs compared with placebo in adults with non-chronic LBP in Medline ALL, Embase, Cochrane Central Register of Controlled Trials*. We will screen the records after January 2020, and include eligible RCTs before January 2020 screened by the Cochrane review mentioned above. Our primary outcomes are pain intensity, disability, and health-related quality of life, secondary outcomes are adverse events. Our IPD dataset will consist of the information on each eligible trial characteristics and included variables according to a predefined coding scheme. We will assess risk-of-bias of included RCTs with the Cochrane Risk Of Bias (RoB)-2 assessment tool. We will perform power calculations with closed-form solutions and prioritize a one-stage approach for IPD-MA. For reporting the results, we will adhere to the PRISMA-IPD statement.

Simulation-based sample size calculations of marginal proportional means models for recurrent events with competing risks.

In randomised controlled trials, the outcome of interest could be recurrent events, such as hospitalisations for heart failure. If mortality rates are non-negligible, both recurrent events and competing terminal events need to be addressed when formulating the estimand and statistical analysis is no longer trivial. In order to design future trials with primary recurrent event endpoints with competing risks, it is necessary to be able to perform power calculations to determine sample sizes. This paper introduces a simulation-based approach for power estimation based on a proportional means model for recurrent events and a proportional hazards model for terminal events. The simulation procedure is presented along with a discussion of what the user needs to specify to use the approach. The method is flexible and based on marginal quantities which are easy to specify. However, the method introduces a lack of a certain type of dependence. This is explored in a sensitivity analysis which suggests that the power is robust in spite of that. Data from a randomised controlled trial, LEADER, is used as the basis for generating data for a future trial. Finally, potential power gains of recurrent event methods as opposed to first event methods are discussed.

Percutaneous Transluminal Angioplasty for Infrapopliteal Chronic Limb-Threatening Ischemia: A Systematic Review and Meta-analysis of Primary Patency and Binary Restenosis Rates.

Percutaneous transluminal angioplasty (PTA) remains the mainstay of endovascular therapy for infrapopliteal chronic limb-threatening ischemia (CLTI), but outcomes have not been well characterized using high-quality data. The aim of this meta-analysis was to provide an updated benchmark for rates of primary patency and binary restenosis after PTA using prospectively collected, predominantly core-lab adjudicated randomized controlled trial (RCT) data. MEDLINE, EMBASE, Cochrane Central, and ClinicalTrials.gov were queried for RCTs published through November 2022 using PTA as a control arm and including patients with infrapopliteal CLTI. Studies were excluded if >25% of patients had intermittent claudication, other vessels were included, or primary patency or binary restenosis were not outcomes. Outcomes were analyzed using random effects models. This analysis was publicly registered (PROSPERO ID#394543). No funding was utilized. Seventeen RCTs were included (1048 patients, 1279 lesions). Pooled primary patency rates using data from 6 RCTs were 68% at 6 months (95% confidence interval [CI]=45%-84%) and 66% at 12 months (95% CI=51%-79%). Pooled binary restenosis rates using data from 11 RCTs were 54% at 6 months (95% CI=33%-73%) and 60% at 9 to 12 months (95% CI=39%-78%). Significant heterogeneity was present in all outcomes (I2>50%, p<0.0001). Publication bias was not observed (Egger's p>0.1). This meta-analysis provides estimates for binary restenosis and primary patency following PTA utilizing prospectively collected, predominantly core-lab adjudicated data. Results demonstrate 1-year primary patency rates that are 10% to 20% higher than what has been historically used in power calculations. These new estimates will help facilitate more accurate power analysis for future RCTs. Rates of primary patency and binary restenosis after percutaneous transluminal angioplasty (PTA) have not been well-described using high-quality data, and investigators have been utilizing estimates of 40% to 50% and 45% to 65%, respectively, when performing power calculations for trials. This meta-analysis demonstrates using high-quality, prospectively collected, and predominantly core-lab adjudicated randomized controlled trial data that actual rates of primary patency are closer to 60% up to 1 year following PTA and provides the first meta-analysis estimate of binary restenosis rates up to 1 year after PTA. These estimates will help facilitate more accurate power calculations for future RCTs in this space.

Causal associations between erectile dysfunction and high blood pressure, negative psychology: a Mendelian randomization study

Erectile dysfunction (ED) has been closely associated with both high blood pressure (HBP) and psychological traits, but the causal relationship between them remains unclear. Herein, we aimed to identify the causal risk factors for ED. We conducted univariable and multivariable Mendelian randomization (MR) analyses using genetic variants associated with metabolic syndrome and psychology traits at the genome-wide significance (p &lt; 5 × 10−8) level obtained from corresponding genome-wide association studies. We used summary-level statistical data for ED from the European Bioinformatics Institute (EBI) database of complete Genome-Wide Association Studies (GWAS) summary data. We also conducted reverse causality and performed power calculations for MR. Our results showed that HBP was associated with increased odds of ED (odds ratio (OR) = 1.66 (95% confidence interval (CI), 1.13–2.45), a p-value for the inverse variance-weighted method (PIVW ) = 1.06 × 10−2, Power = 100%), as were myocardial infarction (OR = 1.09 (95% CI, 1.02–1.17), PIVW = 1.18 × 10−2, Power = 56%) and ischemic stroke (OR = 1.21 (95% CI, 1.02–1.43), PIVW = 2.87 × 10−2, Power = 10%). In terms of psychological traits, irritable mood (OR = 1.86 (95% CI, 1.14–3.02), PIVW = 1.30 × 10−2, Power = 96%) and neuroticism (OR = 1.36 (95% CI, 1.04–1.79), PIVW = 2.66 × 10−2, Power = 80%) were associated with increased odds of ED. Mendelian randomization pleiotropy residual sum and outlier (MR-PRESSO) showed no evidence of pleiotropic bias, and sensitivity analyses confirmed the robustness of our results. We have established a causal link between HBP and ED, and we have also found evidence suggesting a causal relationship between irritable mood and ED.

Analysis of non-human primate models for evaluating prion disease therapeutic efficacy.

Prion disease is a fatal neurodegenerative disease caused by the conformational corruption of the prion protein (PrP), encoded by the prion protein gene (PRNP). While no disease-modifying therapy is currently available, genetic and pharmacological proofs of concept support development of therapies that lower PrP levels in the brain. In light of proposals for clinical testing of such drugs in presymptomatic individuals at risk for genetic prion disease, extensive nonclinical data are likely to be required, with extra attention paid to choice of animal models. Uniquely, the entire prion disease process can be faithfully modeled through transmission of human prions to non-human primates (NHPs), raising the question of whether NHP models should be used to assess therapeutic efficacy. Here we systematically aggregate data from N = 883 prion-inoculated animals spanning six decades of research studies. Using this dataset, we assess prion strain, route of administration, endpoint, and passage number to characterize the relationship of tested models to currently prevalent human subtypes of prion disease. We analyze the incubation times observed across diverse models and perform power calculations to assess the practicability of testing prion disease therapeutic efficacy in NHPs. We find that while some models may theoretically be able to support therapeutic efficacy studies, pilot studies would be required to confirm incubation time and attack rate before pivotal studies could be designed, cumulatively requiring several years. The models with the shortest and most tightly distributed incubation times are those with smaller brains and weaker homology to humans. Our findings indicate that it would be challenging to conduct efficacy studies in NHPs in a paradigm that honors the potential advantages of NHPs over other available models, on a timeframe that would not risk unduly delaying patient access to promising drug candidates.

Technical Performance Evaluation of Olink Proximity Extension Assay for Blood-Based Biomarker Discovery in Longitudinal Studies of Alzheimer's Disease.

The core Alzheimer's disease (AD) cerebrospinal fluid (CSF) biomarkers; amyloid-β (Aß), total tau (t-tau), and phosphorylated tau (p-tau181), are strong indicators of the presence of AD pathology, but do not correlate well with disease progression, and can be difficult to implement in longitudinal studies where repeat biofluid sampling is required. As a result, blood-based biomarkers are increasingly being sought as alternatives. In this study, we aimed to evaluate a promising blood biomarker discovery technology, Olink Proximity Extension Assays for technical reproducibility characteristics in order to highlight the advantages and disadvantages of using this technology in biomarker discovery in AD. We evaluated the performance of five Olink Proteomic multiplex proximity extension assays (PEA) in plasma samples. Three technical control samples included on each plate allowed calculation of technical variability. Biotemporal stability was measured in three sequential annual samples from 54 individuals with and without AD. Coefficients of variation (CVs), analysis of variance (ANOVA), and variance component analyses were used to quantify technical and individual variation over time. We show that overall, Olink assays are technically robust, with the largest experimental variation stemming from biological differences between individuals for most analytes. As a powerful illustration of one of the potential pitfalls of using a multi-plexed technology for discovery, we performed power calculations using the baseline samples to demonstrate the size of study required to overcome the need for multiple test correction with this technology. We show that the power of moderate effect size proteins was strongly reduced, and as a result investigators should strongly consider pooling resources to perform larger studies using this multiplexed technique where possible.

PowerBacGWAS: a computational pipeline to perform power calculations for bacterial genome-wide association studies

Genome-wide association studies (GWAS) are increasingly being applied to investigate the genetic basis of bacterial traits. However, approaches to perform power calculations for bacterial GWAS are limited. Here we implemented two alternative approaches to conduct power calculations using existing collections of bacterial genomes. First, a sub-sampling approach was undertaken to reduce the allele frequency and effect size of a known and detectable genotype-phenotype relationship by modifying phenotype labels. Second, a phenotype-simulation approach was conducted to simulate phenotypes from existing genetic variants. We implemented both approaches into a computational pipeline (PowerBacGWAS) that supports power calculations for burden testing, pan-genome and variant GWAS; and applied it to collections of Enterococcus faecium, Klebsiella pneumoniae and Mycobacterium tuberculosis. We used this pipeline to determine sample sizes required to detect causal variants of different minor allele frequencies (MAF), effect sizes and phenotype heritability, and studied the effect of homoplasy and population diversity on the power to detect causal variants. Our pipeline and user documentation are made available and can be applied to other bacterial populations. PowerBacGWAS can be used to determine sample sizes required to find statistically significant associations, or the associations detectable with a given sample size. We recommend to perform power calculations using existing genomes of the bacterial species and population of study.

The Power of Microbiome Studies: Some Considerations on Which Alpha and Beta Metrics to Use and How to Report Results.

BackgroundSince sequencing techniques have become less expensive, larger sample sizes are applicable for microbiota studies. The aim of this study is to show how, and to what extent, different diversity metrics and different compositions of the microbiota influence the needed sample size to observe dissimilar groups. Empirical 16S rRNA amplicon sequence data obtained from animal experiments, observational human data, and simulated data were used to perform retrospective power calculations. A wide variation of alpha diversity and beta diversity metrics were used to compare the different microbiota datasets and the effect on the sample size.ResultsOur data showed that beta diversity metrics are the most sensitive to observe differences as compared with alpha diversity metrics. The structure of the data influenced which alpha metrics are the most sensitive. Regarding beta diversity, the Bray–Curtis metric is in general the most sensitive to observe differences between groups, resulting in lower sample size and potential publication bias.ConclusionWe recommend performing power calculations and to use multiple diversity metrics as an outcome measure. To improve microbiota studies, awareness needs to be raised on the sensitivity and bias for microbiota research outcomes created by the used metrics rather than biological differences. We have seen that different alpha and beta diversity metrics lead to different study power: because of this, one could be naturally tempted to try all possible metrics until one or more are found that give a statistically significant test result, i.e., p-value < α. This way of proceeding is one of the many forms of the so-called p-value hacking. To this end, in our opinion, the only way to protect ourselves from (the temptation of) p-hacking would be to publish a statistical plan before experiments are initiated, describing the outcomes of interest and the corresponding statistical analyses to be performed.

Understanding Mixed-Effects Models Through Data Simulation

Experimental designs that sample both subjects and stimuli from a larger population need to account for random effects of both subjects and stimuli using mixed-effects models. However, much of this research is analyzed using analysis of variance on aggregated responses because researchers are not confident specifying and interpreting mixed-effects models. This Tutorial explains how to simulate data with random-effects structure and analyze the data using linear mixed-effects regression (with the lme4 R package), with a focus on interpreting the output in light of the simulated parameters. Data simulation not only can enhance understanding of how these models work, but also enables researchers to perform power calculations for complex designs. All materials associated with this article can be accessed at https://osf.io/3cz2e/ .

Anodal tDCS Over the Left Prefrontal Cortex Does Not Cause Clinically Significant Changes in Circulating Metabolites.

BackgroundTranscranial direct current stimulation (tDCS), a putative treatment for depression, has been proposed to affect peripheral metabolism. Metabolic products from brain tissue may also cross the blood–brain barrier, reflecting the conditions in the brain. However, there are no previous data regarding the effect of tDCS on circulating metabolites.ObjectiveTo determine whether five daily sessions of tDCS modulate peripheral metabolites in healthy adult men.MethodsThis double-blind, randomized controlled trial involved 79 healthy males (aged 20–40 years) divided into two groups, one receiving tDCS (2 mA) and the other sham stimulated. The anode was placed over the left dorsolateral prefrontal cortex and the cathode over the corresponding contralateral area. Venous blood samples were obtained before and after the first stimulation session, and after the fifth stimulation session. Serum levels of 102 metabolites were determined by mass spectrometry. The results were analysed with generalised estimating equations corrected for the family-wise error rate. In addition, we performed power calculations estimating sample sizes necessary for future research.ResultsTDCS-related variation in serum metabolite levels was extremely small and statistically non-significant. Power calculations indicated that for the observed variation to be deemed significant, samples sizes of up to 11,000 subjects per group would be required, depending on the metabolite of interest.ConclusionOur study found that five sessions of tDCS induced no major effects on peripheral metabolites among healthy men. These observations support the view of tDCS as a safe treatment that does not induce significant changes in the measured peripheral metabolites in healthy male subjects.

In vivo diffusion-weighted MRS using semi-LASER in the human brain at 3T: Methodological aspects and clinical feasibility.

Diffusion-weighted (DW-) MRS investigates non-invasively microstructural properties of tissue by probing metabolite diffusion in vivo. Despite the growing interest in DW-MRS for clinical applications, little has been published on the reproducibility of this technique. In this study, we explored the optimization of a single-voxel DW-semi-LASER sequence for clinical applications at 3T, and evaluated the reproducibility of the method under different experimental conditions. DW-MRS measurements were carried out in 10 healthy participants and repeated across three sessions. Metabolite apparent diffusion coefficients (ADCs) were calculated from mono-exponential fits (ADCexp ) up to b=3300s/mm2 , and from the diffusional kurtosis approach (ADCK ) up to b=7300s/mm2 . The inter-subject variabilities of ADCs of N-acetylaspartate + N-acetylaspartylglutamate (tNAA), creatine + phosphocreatine, choline containing compounds, and myo-inositol were calculated in the posterior cingulate cortex (PCC) and in the corona radiata (CR). We explored the effect of physiological motion on the DW-MRS signal and the importance of cardiac gating and peak thresholding to account for signal amplitude fluctuations. Additionally, we investigated the dependence of the intra-subject variability on the acquisition scheme using a bootstrapping resampling method. Coefficients of variation were lower in PCC than CR, likely due to the different sensitivities to motion artifacts of the two regions. Finally, we computed coefficients of repeatability for ADCexp and performed power calculations needed for designing clinical studies. The power calculation for ADCexp of tNAA showed that in the PCC seven subjects per group are sufficient to detect a difference of 5% between two groups with an acquisition time of 4min, suggesting that ADCexp of tNAA is a suitable marker for disease-related intracellular alteration even in small case-control studies. In the CR, further work is needed to evaluate the voxel size and location that minimize the motion artifacts and variability of the ADC measurements.

Determining sample size adequacy for animal model studies in nutrition research: limits and ethical challenges of ordinary power calculation procedures

Animal models are widely used in the field of nutrition research. Scientifically and ethically sound experiments need an adequate number of experimental units. The use of 5-10 units is common, but such sample sizes can be justified for large effect sizes only. We reviewed animal model studies recently published in selected journals in the field of nutrition sciences. We performed a simulation study aimed at determining the adequate sample size for normality assessment. We then performed power calculations for a number of statistical tests commonly found in rodent model studies in nutrition research. Among the selected papers, sample sizes ranged from 6-18 units per group. None of them justified the sample size. However, such sample sizes do not allow for normality testing, thus, graphical approaches should be used. Parametric approaches result in higher statistical power when compared to their non-parametric counterparts. Repeated measures analysis should always be preferred, when possible.

Guidance for DNA methylation studies: statistical insights from the Illumina EPIC array

BackgroundThere has been a steady increase in the number of studies aiming to identify DNA methylation differences associated with complex phenotypes. Many of the challenges of epigenetic epidemiology regarding study design and interpretation have been discussed in detail, however there are analytical concerns that are outstanding and require further exploration. In this study we seek to address three analytical issues. First, we quantify the multiple testing burden and propose a standard statistical significance threshold for identifying DNA methylation sites that are associated with an outcome. Second, we establish whether linear regression, the chosen statistical tool for the majority of studies, is appropriate and whether it is biased by the underlying distribution of DNA methylation data. Finally, we assess the sample size required for adequately powered DNA methylation association studies.ResultsWe quantified DNA methylation in the Understanding Society cohort (n = 1175), a large population based study, using the Illumina EPIC array to assess the statistical properties of DNA methylation association analyses. By simulating null DNA methylation studies, we generated the distribution of p-values expected by chance and calculated the 5% family-wise error for EPIC array studies to be 9 × 10− 8. Next, we tested whether the assumptions of linear regression are violated by DNA methylation data and found that the majority of sites do not satisfy the assumption of normal residuals. Nevertheless, we found no evidence that this bias influences analyses by increasing the likelihood of affected sites to be false positives. Finally, we performed power calculations for EPIC based DNA methylation studies, demonstrating that existing studies with data on ~ 1000 samples are adequately powered to detect small differences at the majority of sites.ConclusionWe propose that a significance threshold of P < 9 × 10− 8 adequately controls the false positive rate for EPIC array DNA methylation studies. Moreover, our results indicate that linear regression is a valid statistical methodology for DNA methylation studies, despite the fact that the data do not always satisfy the assumptions of this test. These findings have implications for epidemiological-based studies of DNA methylation and provide a framework for the interpretation of findings from current and future studies.

Beyond "Treatment Versus Control": How Bayesian Analysis Makes Factorial Experiments Feasible in Education Research.

Researchers often wish to test a large set of related interventions or approaches to implementation. A factorial experiment accomplishes this by examining not only basic treatment-control comparisons but also the effects of multiple implementation "factors" such as different dosages or implementation strategies and the interactions between these factor levels. However, traditional methods of statistical inference may require prohibitively large sample sizes to perform complex factorial experiments. We present a Bayesian approach to factorial design. Through the use of hierarchical priors and partial pooling, we show how Bayesian analysis substantially increases the precision of estimates in complex experiments with many factors and factor levels, while controlling the risk of false positives from multiple comparisons. Using an experiment we performed for the U.S. Department of Education as a motivating example, we perform power calculations for both classical and Bayesian methods. We repeatedly simulate factorial experiments with a variety of sample sizes and numbers of treatment arms to estimate the minimum detectable effect (MDE) for each combination. The Bayesian approach yields substantially lower MDEs when compared with classical methods for complex factorial experiments. For example, to test 72 treatment arms (five factors with two or three levels each), a classical experiment requires nearly twice the sample size as a Bayesian experiment to obtain a given MDE. Bayesian methods are a valuable tool for researchers interested in studying complex interventions. They make factorial experiments with many treatment arms vastly more feasible.

MendelProb: probability and sample size calculations for Mendelian studies of exome and whole genome sequence data.

For the design of genetic studies, it is necessary to perform power calculations. Although for Mendelian traits the power of detecting linkage for pedigree(s) can be determined, it is also of great interest to determine the probability of identifying multiple pedigrees or unrelated cases with variants in the same gene. For many diseases, due to extreme locus heterogeneity this probability can be small. If only one family is observed segregating a variant classified as likely pathogenic or of unknown significance, the gene cannot be implicated in disease etiology. The probability of identifying several disease families or cases is dependent on the gene-specific disease prevalence and the sample size. The observation of multiple disease families or cases with variants in the same gene as well as evidence of pathogenicity from other sources, e.g. in silico prediction, expression and functional studies, can aid in implicating a gene in disease etiology. MendelProb can determine the probability of detecting a minimum number of families or cases with variants in the same gene. It can also calculate the probability of detecting genes with variants in different data types, e.g. identifying a variant in at least one family that can establish linkage and more the two additional families regardless of their size. Additionally, for a specified probability MendelProb can determine the number of probands which need to be screened to detect a minimum number of individuals with variants within the same gene. A single Mendelian disease family is not sufficient to implicate a gene in disease etiology. It is necessary to observe multiple families or cases with potentially pathogenic variants in the same gene. MendelProb, an R library, was developed to determine the probability of observing multiple families and cases with variants within a gene and to also establish the numbers of probands to screen to detect multiple observations of variants within a gene. https://github.com/statgenetics/mendelprob.

Implementing near real-time vaccine safety surveillance using the Clinical Practice Research Datalink (CPRD).

Near real-time vaccine safety surveillance (NRTVSS) using electronic health records is increasingly used to rapidly detect vaccine safety signals. NRTVSS has not been fully implemented in the UK. We assessed the feasibility of implementing this surveillance using the UK Clinical Practice Research Datalink (CPRD). We selected seasonal influenza vaccine/Guillain-Barré Syndrome (GBS) as an example of a rare outcome and measles-mumps-rubella (MMR) vaccine/febrile seizures as a positive control. For influenza/GBS we implemented a system for the 2013/2014 and 2014/2015 influenza seasons; for MMR/seizures the surveillance period was July 2014-June 2015. We used the continuous Poisson-based maximized sequential probability ratio test (PMaxSPRT), comparing observed-to-expected events, for both pairs. We calculated an age-sex-adjusted rate using 5years of historic data and used this rate to calculate the expected number of events in pre-specified post-vaccination risk-window (GBS: 0-42days, seizures: 6-21days). For MMR/seizures we also implemented the system using the Binominal-based maximized sequential probability ratio test (BMaxSPRT). For this, we compared seizures in the risk-window (6-21days) to a control window (0-5 and 22-32days). Delays in recording outcomes influence the data available, so we adjusted the expected number of events using a historical distribution of delays in recording GBS/febrile seizures. Analyses were run using data up to each CPRD monthly release. We also performed power calculations for detecting increases in relative risk (RR) from 1.5 to 10. For influenza/GBS we implemented a system in both seasons with no signal. Power to detect a signal was >80% for RR≥4. For MMR/seizures we were able to identify a signal with PMaxSPRT but not with BMaxSPRT. Power≥80% for RR≥2.5 for both tests. CPRD is a potential data source to implement NRTVSS to exclude large increases in the risk of rare outcomes after seasonal influenza and lower increases in risk for more frequent outcomes.

SurvivalGWAS_Power: a user friendly tool for power calculations in pharmacogenetic studies with "time to event" outcomes.

BackgroundPower calculators are currently available for the design of genetic association studies of binary phenotypes and quantitative traits, but not for “time to event” outcomes, which are of particular relevance in pharmacogenetics. With the rapid emergence of pharmacogenetic association studies of single nucleotide polymorphisms (SNPs), and the complexity of clinical outcomes they consider, there is a need for software to perform power calculations of time to event data over a range of design scenarios and analytical methodologies.ResultsWe have developed the user friendly software tool SurvivalGWAS_Power to perform power calculations for time to event outcomes over a range of study designs and different analytical approaches. The software calculates the power to detect SNP association with a time to event outcome over a range of study design scenarios. The software enables analyses under a Cox proportional hazards model or Weibull regression model, and can account for treatment and SNP-treatment interaction effects. Simulated data sets can also be generated by SurvivalGWAS_Power to enable analyses with methods that are not currently supported by the power calculator, thereby increasing the flexibility of the software.ConclusionsSurvivalGWAS_Power addresses the need for flexible and user-friendly software for power calculations for genetic association studies of time to event outcomes, with particular design features of relevance in pharmacogenetics.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-1407-9) contains supplementary material, which is available to authorized users.

Treatment Trials for Neonatal Seizures: The Effect of Design on Sample Size.

Neonatal seizures are common in the neonatal intensive care unit. Clinicians treat these seizures with several anti-epileptic drugs (AEDs) to reduce seizures in a neonate. Current AEDs exhibit sub-optimal efficacy and several randomized control trials (RCT) of novel AEDs are planned. The aim of this study was to measure the influence of trial design on the required sample size of a RCT. We used seizure time courses from 41 term neonates with hypoxic ischaemic encephalopathy to build seizure treatment trial simulations. We used five outcome measures, three AED protocols, eight treatment delays from seizure onset (Td) and four levels of trial AED efficacy to simulate different RCTs. We performed power calculations for each RCT design and analysed the resultant sample size. We also assessed the rate of false positives, or placebo effect, in typical uncontrolled studies. We found that the false positive rate ranged from 5 to 85% of patients depending on RCT design. For controlled trials, the choice of outcome measure had the largest effect on sample size with median differences of 30.7 fold (IQR: 13.7–40.0) across a range of AED protocols, Td and trial AED efficacy (p<0.001). RCTs that compared the trial AED with positive controls required sample sizes with a median fold increase of 3.2 (IQR: 1.9–11.9; p<0.001). Delays in AED administration from seizure onset also increased the required sample size 2.1 fold (IQR: 1.7–2.9; p<0.001). Subgroup analysis showed that RCTs in neonates treated with hypothermia required a median fold increase in sample size of 2.6 (IQR: 2.4–3.0) compared to trials in normothermic neonates (p<0.001). These results show that RCT design has a profound influence on the required sample size. Trials that use a control group, appropriate outcome measure, and control for differences in Td between groups in analysis will be valid and minimise sample size.

Power calculation for causal inference in social science: sample size and minimum detectable effect determination

ABSTRACTThis paper presents the statistical concepts used in power calculations for experimental design. It provides detailed definitions of parameters used to perform power calculations, useful rules of thumb and different approaches that can be used when performing power calculations. The authors draw from real-world examples to calculate statistical power for individual and cluster randomised controlled trials. This paper provides formulae for sample size determination and minimum detectable effect (MDE) associated with a given statistical power. The paper is accompanied by the sample size and MDE calculator©, a free online tool that allows users to work with the formulae presented in Section 4.