Search PubMed⌕ Search

Biomedical subjects

Douglas G Altman

Publications and source records attributed to Douglas G Altman.

At least 19 recordsLinked to original sources
PubMed2003

Analysis and reporting of factorial trials: a systematic review.

CONTEXT: Although factorial trials have become common, standards for the analysis and reporting of such trials have not been established and, despite concerns about the possibility of unrecognized interactions between therapies in factorial trials, the magnitude of this potential problem is unknown. OBJECTIVE: To examine the rationale, methods, and analysis of randomized factorial trials. DATA SOURCES AND STUDY SELECTION: We searched MEDLINE, EMBASE, and the Cochrane Controlled Trials Register using the terms factorial, interaction, 2 x 2, 2 by 2, and incremental to identify factorial randomized trials published from January 2000 to July 2002. To identify trials missed by the electronic search, we performed a hand search of English-language trials in a defined topic area (using the term myocardial ischemia [exp]) listed in MEDLINE (1966-2002), EMBASE (1980-2002), and the Cochrane Controlled Trials Register, as well as all trials in any topic area published in December 2000, excluding trials reporting only continuous surrogate end points. The final set of 33 eligible publications described 29 unique trials. DATA EXTRACTION: Two investigators independently identified factorial trials, generated a list of items affecting validity of results, and abstracted these items from each trial. DATA SYNTHESIS: The sensitivity of electronic searching for identifying factorial trials was 76%. Our 3-pronged search strategy identified 44 factorial trials with clinically important binary outcomes: 36 (82%) were done for reasons of efficiency (testing 2 interventions in the same patient population), and 8 (18%) were done to assess the incremental benefits of combining the 2 treatments. All but 1 of the trials reported treatment effects by comparing all patients who received treatment A (ie, those receiving either A alone or both A and B) vs all those not receiving treatment A (ie, those receiving either B alone or neither A nor B). Twenty-nine of the 44 trials (66%) reported the data from each of the treatment groups separately; 26 trials (59%) reported testing for interactions between the treatments. Only 2 of 31 (6%) comparisons demonstrated a statistically significant interaction between the 2 treatments. CONCLUSIONS: Accurate interpretation of factorial trials depends on the transparent reporting of data for each treatment cell. Despite concerns about unrecognized interactions, our findings suggest that investigators are appropriately restricting their use of the factorial design to those situations in which 2 (or more) treatments do not have the potential for substantive interaction.

Data Interpretation, Statistical↗
PubMed2003

Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses.

OBJECTIVE: To determine the validity of adjusted indirect comparisons by using data from published meta-analyses of randomised 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 (kappa 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.

Data Interpretation, Statistical↗
PubMed2003

The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials.

To comprehend the results of a randomised controlled trial (RCT), readers must understand its design, conduct, analysis, and interpretation. That goal can be achieved only through total transparency from authors. Despite several decades of educational efforts, the reporting of RCTs needs improvement. Investigators and editors developed the original CONSORT (Consolidated Standards of Reporting Trials) statement to help authors improve reporting by use of a checklist and flow diagram. The revised CONSORT statement presented here incorporates new evidence and addresses some criticisms of the original statement. The checklist items pertain to the content of the Title, Abstract, Introduction, Methods, Results, and Discussion. The revised checklist includes 22 items selected because empirical evidence indicates that not reporting this information is associated with biased estimates of treatment effect, or because the information is essential to judge the reliability or relevance of the findings. We intended the flow diagram to depict the passage of participants through an RCT. The revised flow diagram depicts information from four stages of a trial (enrollment, intervention allocation, follow- up, and analysis). The diagram explicitly shows the number of participants, for each intervention group, included in the primary data analysis. Inclusion of these numbers allows the reader to judge whether the authors have done an intention- to-treat analysis. In sum, the CONSORT statement is intended to improve the reporting of an RCT, enabling readers to understand a trial's conduct and to assess the validity of its results.

Humans↗
PubMed2003

Developing a prognostic model in the presence of missing data: an ovarian cancer case study.

When developing prognostic models in medicine, covariate data are often missing and the standard response is to exclude those individuals whose data are incomplete from the analyses. This practice leads to a reduction in the statistical power, and may lead to biased results. We wished to develop a prognostic model for overall survival from 1,189 primary cases (842 deaths) of epithelial ovarian cancer. A complete case analysis restricted the sample size to 518 (380 deaths). After applying a multiple imputation (MI) framework we included three real values for each one imputed, and constructed a model composed of more statistically significant prognostic factors and with increased predictive ability. Missing values can be imputed in cases where the reason for the data being missing is known, particularly where it can be explained by available data. This will increase the power of an analysis and may produce models that are more statistically reliable and applicable within clinical practice.

Adolescent↗
PubMed2002

Meta-analysis combining parallel and cross-over clinical trials. I: Continuous outcomes.

Among clinical trials assessing a given treatment, often parallel and cross-over designs are used together. In the first paper of a series of three, we explore two methods to pool continuous outcomes in a meta-analysis combining parallel and cross-over trial designs: the weighted mean difference (WMD) and the standardized weighted mean difference (SWMD). The combined design meta-analytic formulae are based on a weighted average of the two design treatment estimates. A random effects model can be implemented. Both WMD and SWMD can be used, the choice of the method is determined by the type of outcomes obtained in the trials. Compared to the number of included subjects, the relative weight of the cross-over design is large in combined-design meta-analysis. Differences in the weight estimation between WMD and SWMD can also accentuate the relative weight of cross-over trials, which must be considered a case of design-specific bias.

Blood Pressure↗
PubMed2002

Meta-analysis combining parallel and cross-over clinical trials. II: Binary outcomes.

We examine different methods to pool binary outcomes used both in parallel and cross-over trials. Odds ratio (OR) estimators obtained from joint conditional probabilities in cross-over trials, such as the Mantel-Haenszel and Peto methods, are compared to an OR estimator using marginal results of cross-over trials. When there is correlation between the outcomes in the two cross-over periods, joint conditional ORs differ from marginal ORs and cannot be combined with OR estimates from parallel trials. The marginal OR estimate is independent of the between-period correlation and it includes a correction for cross-over correlation in the variance estimate. As its computation is similar in cross-over and parallel trials, it is the method of choice to pool results from parallel and cross-over trials in a combined design meta-analysis.

Albuminuria↗
PubMed2002

Meta-analysis combining parallel and cross-over clinical trials. III: The issue of carry-over.

In meta-analysis combining results from parallel and cross-over trials, there is a risk of bias originating from the carry-over effect in cross-over trials. When pooling treatment effects estimated from parallel trials and two-period two-treatment cross-over trials, meta-analytic estimators of treatment effect can be obtained from the combination of parallel trial results either with cross-over trial results based on data of the first period only or with cross-over trial results analysed with data from both periods. Taking data from the first cross-over period protects against carry-over but gives less efficient treatment estimators and may lead to selection bias. This study evaluates in terms of variance reduction and mean square error the cost of calculating meta-analysis estimates with data from the first period instead of data from the two cross-over periods. If the information on cross-over sequence is available, we recommend performing two combined design meta-analyses, one using the first cross-over period data and one based on data from both cross-over periods. To investigate simultaneously the statistical significance of these two estimators as well as the carry-over at meta-analysis level, a method based on a multivariate analysis of the meta-analytic treatment effect and carry-over estimates is proposed.

Blood Pressure↗
PubMed2002

Statistical methods for assessing the influence of study characteristics on treatment effects in 'meta-epidemiological' research.

Biases in systematic reviews and meta-analyses may be examined in 'meta-epidemiological' studies, in which the influence of trial characteristics such as measures of study quality on treatment effect estimates is explored. Published studies to date have analysed data from collections of meta-analyses with binary outcomes, using logistic regression models that assume that there is no between- or within-meta-analysis heterogeneity. Using data from a study of publication bias (39 meta-analyses, 394 published and 88 unpublished trials) and language bias (29 meta-analyses, 297 English language trials and 52 non-English language trials), we compare results from logistic regression models, with and without robust standard errors to allow for clustering on meta-analysis, with results using a 'meta-meta-analytic' approach that can allow for between- and within-meta-analysis heterogeneity. We also consider how to allow for the confounding effects of different trial characteristics. We show that both within- and between meta-analysis heterogeneity may be of importance in the analysis of meta-epidemiological studies, and that confounding exists between the effects of publication status and trial quality.

Bias↗
PubMed2002

Poor-quality medical research: what can journals do?

The aim of medical research is to advance scientific knowledge and hence--directly or indirectly--lead to improvements in the treatment and prevention of disease. Each research project should continue systematically from previous research and feed into future research. Each project should contribute beneficially to a slowly evolving body of research. A study should not mislead; otherwise it could adversely affect clinical practice and future research. In 1994 I observed that research papers commonly contain methodological errors, report results selectively, and draw unjustified conclusions. Here I revisit the topic and suggest how journal editors can help.

Peer Review, Research↗
PubMed2002

How statistical expertise is used in medical research.

CONTEXT: Investigation of the nature and frequency of statistician involvement in medical research and its relation to the final editorial decision. METHODS: Authors of original research articles who submitted to BMJ and Annals of Internal Medicine from May through August 2001 were sent a short questionnaire at the time of manuscript submission. Authors were asked if they received assistance from a person with statistical expertise, the nature of any such contribution, and reasons why, if no statistical input was received. RESULTS: The response rate was 75% (704/943); methodological input was reported for 514 (73%) of these papers. In 435 papers (85%), such input was provided by biostatisticians or epidemiologists and, if deemed significant, was typically associated with authorship. A total of 33 of 122 methodologists (27%) whose main contribution started at the analysis stage received neither acknowledgment nor authorship. Research without methodological assistance was more likely to be rejected without review (71% vs 57%; chi(2) = 10.6; P =.001) and possibly less likely to be accepted for publication (7% vs 11%; chi(2) = 2.37; P =.12). CONCLUSIONS: Statistical input to medical research is widely recommended but inconsistently obtained. Individuals providing such expertise are often not involved until the analysis of data and many go unrecognized by either authorship or acknowledgment.

Authorship↗