Search

Objective To evaluate how often sample size calculations and methods of statistical analysis are pre-specified or changed in randomised trials. Design Retrospective cohort study. Data source Protocols and journal publications of published randomised parallel group trials initially approved in 1994-5 by the scientific-ethics committees for Copenhagen and Frederiksberg, Denmark (n=70). Main outcome measure Proportion of protocols and publications that did not provide key information about sample size calculations and statistical methods; proportion of trials with discrepancies between information presented in the protocol and the publication. Results Only 11/62 trials described existing sample size calculations fully and consistently in both the protocol and the publication. The method of handling protocol deviations was described in 37 protocols and 43 publications. The method of handling missing data was described in 16 protocols and 49 publications. 39/49 protocols and 42/43 publications reported the statistical test used to analyse primary outcome measures. Unacknowledged discrepancies between protocols and publications were found for sample size calculations (18/34 trials), methods of handling protocol deviations (19/43) and missing data (39/49), primary outcome analyses (25/42), subgroup analyses (25/25), and adjusted analyses (23/28). Interim analyses were described in 13 protocols but mentioned in only five corresponding publications. Conclusion When reported in publications, sample size calculations and statistical methods were often explicitly discrepant with the protocol or not pre-specified. Such amendments were rarely acknowledged in the trial publication. The reliability of trial reports cannot be assessed without having access to the full protocols. [ABSTRACT FROM AUTHOR]

The article provides an extension to the Consolidated Standards of Reporting Trials (CONSORT) guidance with guidance for reporting pragmatic trials. The authors note that randomised controlled trials minimise the risk of bias when conducted properly. The need to clearly report trial results is addressed. The applicability of the trial results to practice is addressed. A table listing the differences between explanatory and pragmatic attitudes is included. A checklist of items for reporting pragmatic trials is presented including background, methods, participants, interventions, and objectives of the trials.

Objective To examine whether the association of inadequate or unclear allocation concealment and lack of blinding with biased estimates of intervention effects varies with the nature of the intervention or outcome. Design Combined analysis of data from three meta- epidemiological studies based on collections of meta- analyses. Data sources 146 meta-analyses including 1346 trials examining a wide range of interventions and outcomes. Main outcome measures Ratios of odds ratios quantifying the degree of bias associated with inadequate or unclear allocation concealment, and lack of blinding, for trials with different types of intervention and outcome. A ratio of odds ratios <1 implies that inadequately concealed or non- blinded trials exaggerate intervention effect estimates. Results In trials with subjective outcomes effect estimates were exaggerated when there was inadequate or unclear allocation concealment (ratio of odds ratios 0.69 (95% Cl 0.59 to 0.82)) or lack of blinding (0.75 (0.61 to 0.93)). In contrast, there was little evidence of bias in trials with objective outcomes: ratios of odds ratios 0.91 (0.80 to 1.03) for inadequate or unclear allocation concealment and 1.01 (0.92 to 1.10) for lackof blinding. There was little evidence for a difference between trials of drug and non- drug interventions. Except fortrials with all cause mortality as the outcome, the magnitude of bias varied between meta-analyses. Conclusions The average bias associated with defects in the conduct of randomised trials varies with the type of outcome. Systematic reviewers should routinely assess the risk of bias in the results of trials, and should report meta-analyses restricted to trials at low risk of bias either as the primary analysis or in conjunction with less restrictive analyses. INSET: WHAT IS ALREADY KNOWN ON THIS TOPIC. [ABSTRACT FROM AUTHOR]

The article argues that medical ethics for peer review outweigh the privacy and commercial interests of researchers developing drugs. The lawsuits brought against the pharmaceutical company Merck & Co. concerning side effects of their cholesterol drug Vioxx have highlighted the importance of medical research integrity. It is recommended that a protocol be developed which allows external peer review of medical research.

Presents a study comparing how allocation concealment is described in publications of randomized clinical trials and corresponding protocols. Aim of the study to estimate how often trial publications with unclear allocation concealment have adequate concealment according to the protocol; Review of how the study was conducted and of statistical data gathered by the study; Conclusion that most randomized clinical trials have unclear allocation concealment on the basis of trial publication alone and that trials also have unclear allocation concealment according to their protocol.

Objective To examine the extent and nature of outcome reporting bias in a broad cohort of published randomized trials. Design Retrospective review of publications and follow up survey of authors. Cohort All journal articles of randomized trials indexed in PubMed whose primary publication appeared in December 2000. Main outcome measures Prevalence of incompletely reported outcomes per trial; reasons for not reporting outcomes; association between completeness of reporting and statistical significance. Results 519 trials with 553 publications and 10 557 outcomes were identified. Survey responders (response rate 69%) provided information on unreported outcomes but were often unreliable--for 32% of those who denied the existence of such outcomes there was evidence to the contrary in their publications. On average, over 20% of the outcomes measured in a parallel group trial were incompletely reported. Within a trial, such outcomes had a higher odds of being statistically non-significant compared with fully reported outcomes (odds ratio 2.0 (95% confidence interval 1.6 to 2.7) for efficacy outcomes; 1.9 (1.1 to 3.5) for harm outcomes). The most commonly reported reasons for omitting efficacy outcomes included space constraints, lack of clinical importance, and lack of statistical significance. Conclusions Incomplete reporting of outcomes within published articles of randomized trials is common and is associated with statistical non-significance. The medical literature therefore represents a selective and biased subset of study outcomes, and trial protocols should be made publicly available. [ABSTRACT FROM AUTHOR]

Offers a look at the effective reporting of randomized controlled trials. Guidelines in the Consolidated Standards of Reporting Trials (CONSORT) statement; Discussion of methodological issues in cluster randomized trials; Quality of reporting of cluster trials; Extension of CONSORT statement to cluster trials; Example of titles and abstracts; Discussion of the types of diagrams.

Systematic reviews and meta-analyses can provide convincing and reliable evidence relevant to many aspects of medicine and health care. In an attempt to establish whether studies are consistent, reports of meta-analyses commonly present a statistical test of heterogeneity. The usual test statistic is computed by summing the squared deviations of each study's estimate from the overall meta-analytic estimate, weighting each study's contribution in the same manner as in the meta-analysis.

Presents letters to the editor in relation to medicine. Responses about screening for depression in primary care; Comments about cardiovascular mortality after radiotherapy for breast cancer; Views on the idea of monitoring the marketing of infant formula feeds.

Abstract Objective To determine the validity of adjusted indirect comparisons by using data from published meta-analyses of randornised trials. Design Direct comparison of different interventions in randomised trials and adjusted indirect comparison in which two interventions were compared through their relative effect versus a common comparator. The discrepancy between the direct and adjusted indirect comparison was measured by the difference between the two estimates. Data sources Database of abstracts of reviews of effectiveness (1994-8), the Cochrane database of systematic reviews, Medline, and references of retrieved articles. Results 44 published meta-analyses (from 28 systematic reviews) provided sufficient data. In most cases, results of adjusted indirect comparisons were not significantly different from those of direct comparisons. A significant discrepancy (P < 0.05) was observed in three of the 44 comparisons between the direct and the adjusted indirect estimates. There was a moderate agreement between the statistical conclusions from the direct and adjusted indirect comparisons (κ 0.51). The direction of discrepancy between the two estimates was inconsistent. Conclusions Adjusted indirect comparisons usually but not always agree with the results of head to head randomised trials. When there is no or insufficient direct evidence from randomised trials, the adjusted indirect comparison may provide useful or supplementary information on the relative efficacy of competing interventions. The validity of the adjusted indirect comparisons depends on the internal validity and similarity of the included trials. [ABSTRACT FROM AUTHOR]

Presents letters to the editor about medical topics. Number of trust doctors in Great Britain; Peer review of statistics in medical research; Decline in the number of cases of polio in Nigeria; Misleading editorial on epidemiology of HIV in China; Trends in HIV, gonorrhea, and syphilis; Intermediate care; Selection of medical students; Acid, burns and feeding tubes; Ideals of Hippocrates.

Discusses systematic reviews of prognostic studies. Difficulty of searching literature for prognostic studies; Assessment of methodological quality of design and analysis; Meta-analysis; Problems of publication bias.

Discusses the concept of study quality and the methods used to assess quality of randomised controlled trials. Validity of the findings as a dimension of quality; Bias in clinical trials, such as selection bias and performance bias; Applicability of the results of a study to other populations and settings.

Focuses on a study that showed how to obtain a number needed to treat for controlled trials and studies where the primary outcome is the time to an event. Methods and examples in the study; Results and discussion; Conclusion.

Focuses on the number needed to treat (NNT), a useful way of reporting the results of randomized controlled trials. Calculation of the NNT; Description of the confidence interval for the NNT when the difference between the two treatments is not significant; Construction of sensible confidence intervals for the NNT.

Studies the efficacy of an educational visit to help obstetricians and midwives select and use evidence from Cochrane database containing systematic reviews. Methodology; Data analysis; Discussion on the results of the study; Conclusions.

Analyzes serial measurements in medical research. Use of simple statistical techniques; Inadequacy of data; Types of time dependency.

Presents several letters on medical practice. Sensitivity of glucose tolerance test; Acceptability of measuring glycated hemoglobin concentration; Problems of obtaining informed consent from patient entering randomized trials.

Presents several letters on medical education in Great Britain. Effect of health prevention messages to the public; Epidemiology of endometriosis; Increase of patients knowledge of secondary contraception.

Comments on several topics on medical sciences. Disagreement on the growth and clotting factors in adult life by Richard de Courcy-Wheeler and Charles Wolfe; Argumentation over the ratio of placental weight to birth weight as a marker of fetal growth; Dosage of oxybutynin for the unstable bladder.

This article presents an analysis of guidelines for reporting observational studies in epidemiology written by an international group of methodologists, researchers and medical journal editors in the STROBE study. Medical research relies on observational studies. They are especially useful in detecting late adverse reactions to treatments but poor reporting of this research puts the results in question. The STROBE checklist is presented to help researchers put their information into a uniform system that will be easily understood by all. Separate versions of the checklist for cohort, case-control and cross-sectional studies are available at www.strobe-statement.org.

Abstract Objectives: To evaluate the efficacy of low and medium dose aspirin therapy after coronary surgery by using an indirect comparison meta-analysis. Data sources: Systematic literature search of Medline, Embase, Cochrane controlled trials register, and trial register sites on the internet. Study selection: Outcome was evaluated by angiography and reported as graft occlusion and rate of events in patients. Trials that did not include aspirin as the sole therapy or did not have a placebo control arm were excluded. Articles were assessed for eligibility and quality and grouped according to dosage. The estimated difference in effect of low and medium dose aspirin on graft occlusion was obtained by combining the estimated log relative risks of low dose with placebo and medium dose with placebo. Results: For graft occlusion, the medium dose trials yielded a relative risk reduction of 45% compared with 26% for the low dose trials. The greater effect in the medium dose trials is summarised by a relative risk ratio of 0.74 (95% confidence interval 0.52 to 1.06; P = 0.10) for graft occlusion and 0.81 (0.57 to 1.16; P = 0.25) for events in patients. Conclusions: Medium dose aspirin may more successfully reduce graft occlusion than low dose regimens within the first year after coronary surgery. [ABSTRACT FROM AUTHOR]

Opinion. Comments on the clinical acceptability of prognostic models. Reasons why doctors reject the publication of prognostic models; Clinical credibility of the model; Evidence of accuracy; Evidence of generality; Evidence of clinical effectiveness.

Presents a study to determine the extent to which leading medical journals had incorporated the Consolidated Standards of Reporting Trials (CONSORT) recommendations into their instructions for authors. Description of the recommendations; Assertion that journals should be more explicit in their expectations of authors and ensure the accuracy of their instructions to authors; Indication that the recommendations have helped to improve the quality of reporting of trials.

Presents statistical notes on likelihood ratios as opposed to risk factors as indicators of a patient's disease. Belief that the use of likelihood ratios is superior for diagnostic accuracy; How the ratio is the probability of the specific test result in people who do have the disease to the probability in people who do not; What the ratio numbers mean; Example of a table which tracked the value of a history of smoking in diagnosing obstructive airway disease; Belief that likelihood ratios can been used to help adapt the results of a study to specific patients; Role of the Baynes theorem; Suggestion that likelihood rations are more useful in clinical practice.

Discusses scales and indexes of quality to measure Web sites with medical information, among other things. Way measurements can be defined in terms of a fundamental unit; Quantities which are difficult to measure and evaluate, such as cardiac stroke volume; Discussion of questionnaire scales in the social sciences; Need for a scale to be objective.

Evaluates the use of baseline and follow up measurements in randomised trials. Statistical comparisons of baseline and follow up scores for control and treatment groups; Use of change scores; Perceived advantages of analysis of covariance, including a greater statistical power to detect a treatment effect.

Discusses the importance of concealing treatment allocation until a patient has entered a randomised clinical trial. Attempts to eliminate bias; How to ensure that allocation sequence will be concealed, including using external help; Difference between randomised and double blind trials.

Considers how to generate a random allocation sequence. Definition of simple randomization; Equal allocation; When random allocation sequence is used.

Highlights the Royal Statistical Society meeting on the excess of cases of cancer around the nuclear installation in Great Britain. Excess of childhood leukemia; No causal pathway being identified; Development of leukemia due to the existence of clusters of cases in excess of the number expected by chance.

Focuses on the scandal of poor medical research. Publication of misleading medical journals; Use of inappropriate designs, unrepresented samples, small samples, incorrect methods of analysis and faulty interpretation; Reasons to the occurrence of errors.

Focuses on the importance of randomization for reducing medical research biases. Explanation concerning random allocation in a clinical trial; Suggestion to use method of allocation aiming to remove the problems of bias; Presentation of cases leading to the impossibility of using randomization.

Focuses on survival data and survival analysis in relation to health care in Great Britain. Why it is conventional to talk about these two factor irregardless of the positive or negative outcome; Discussion on the distinguishing features of survival data; Statistics on survival times for patients in a randomized trial conducted in Great Britain.

Focuses on generalising medical research with individual persons. Information on how treatments work in mild and severe cases; Insight on the correct use of generalisation; Highlights of various experiments.

Editorial. Comments that randomized clinical trials are the best way to compare the effectiveness of different interventions. Validity of cause and effect inferences; Potential of trials to affect patient care; Requirement for higher trial standards.

Considers the use of correlation coefficients to quantify measurement error. Common design for the investigation of measurement error; Problems with the use of the correlation coefficient; Preference for within subjects standard deviation as an index of measurement error.

Part XX. Discusses the use of one way analysis of variance (ANOVA) for comparing three or more groups of subjects. Basis of the ANOVA the same as in the t test method; Division of variability of individual data values into components corresponding to within and between group variation; Extension of the ANOVA to more complex data sets.

This article presents a letter to the editor in response to "Masking is better than blinding," by D. Morris, S. Fraser and R. Wormald in the April 14, 2007 issue.

The article discusses the issue of missing data in quantitative medical research. Where losses are expected, one should try to increase the sample size. Ways to handle missing data include eliminating from analysis variables which have missing values or eliminating individuals with incomplete data or estimation or imputation of the missing values. The greater the missing data instances, the higher the potential for biased results and the threat to the study's integrity. Causes of missing data include such circumstances as a postal questionnaire survey in which not all selected individuals respond or patients who are lost to follow-up before the end of the study.

The article discusses the advantages and disadvantages of dichotomising continuous variables in medical data analyses. While the statistical methodology of analysing two groups simplifies the medical research process, it can lead to false positive correlations. By dichotomising a variable at the median, researchers are essentially discarding one-third of the data of the study.

Focuses on the terms "standard error" and "standard deviation." Explanation of standard deviation, which is a measure of variability that estimates variation within a population from which the sample was drawn; Observation that 95% of the population is within two standard deviations, but a sample mean varies from sample to sample; Explanation of standard error, which is a measurement of uncertainty around the sample man.

Looks at the statistical aspects of randomized controlled clinical trials. Discussion of the process of minimization used in small trials; Contrasts between minimization and randomization; Reference to a diet therapy trial which employed minimization of variables; Use of randomization and minimization in determining the treatment of a trial participant; Validity of minimization as an alternative to randomization; Benefits of using software designed to aid the minimization process.

Presents a letter to the editor about blinding in relation to clinical studies.

Presents a description of the logrank test, the most popular method of comparing the survival of groups, which takes the whole follow up period into account. How the logrank test is most likely to detect a difference between groups when the risk of an event is consistently greater for one group than another; Assertion that because the logrank test is purely a test of significance it cannot provide an estimate of the size of the difference between the groups or a confidence interval.

Considers the concept of comparing two estimates of the same quantity derived from separate analyses, and of tests of interaction. Idea that the two estimates should be independent, and not obtained from he same individuals in a randomized trial; Comparison of the relative risks or odds ratios, which are analyzed on a log scale; Limitations to the power to detect interactions.

Comments on the treatment of warts and a systematic review for treatments for warts by Gibbs and colleagues in this issue. Idea that standard methods of analysis are needed to analyze results for trials of wart treatment; Design options for clinical trials of wart treatment; Comparison of treatments within patients with multiple lesions.

Discusses the relevance of allocation concealment, or blinding, in clinical trial practice. Importance of blinding in trials when response criteria is subjective; Difficulties involved in conducting double blind trials; Risk of bias associated with inadequate blinding; Use of blinding in other types of research.

Explains and contrasts the terms, parameter and variable as used in statistics. Attributes of variables; Why statistical methods are called parametric; Concerns about the misuse of the terms.

Explores the importance of randomizing treatment allocation in controlled trials for the publication of studies in Great Britain. Difficulties in contemporary controls; Purpose of randomization; Definition of random in medical aspect.

Estimates the probability of patients' survival using the Kaplan-Meier method. Calculation of survival; Assumptions; Need for formal methods for testing hypotheses about survival in two or more groups.