Bmc Med Res Methodol
-
Bmc Med Res Methodol · Jan 2005
Comparative StudySystematic reviews of epidemiology in diabetes: finding the evidence.
Methodological research to support searching for those doing systematic reviews of epidemiological studies is a relatively neglected area. Our aim was to determine how many databases it is necessary to search to ensure a comprehensive coverage of the literature in diabetes epidemiology, with the aim of examining the efficiency of searching in support of systematic reviews of the epidemiology of diabetes ⋯ Our results show that when searching for articles on diabetes epidemiology, MEDLINE and Embase would suffice for English language papers, with LILACS giving some additional non-English articles from Latin America. Although a MEDLINE-only search will retrieve the vast majority of the relevant literature, Embase and LILACs should also be searched to ensure the search is comprehensive. Searching for meeting abstracts is recommended to alert reviewers to unpublished work. The low rate of full publication of meeting abstracts has the danger of producing bias in reviews. Our findings on scattering show that the core literature in this field is concentrated in just six journals.
-
Bmc Med Res Methodol · Jan 2005
Parametric versus non-parametric statistics in the analysis of randomized trials with non-normally distributed data.
It has generally been argued that parametric statistics should not be applied to data with non-normal distributions. Empirical research has demonstrated that Mann-Whitney generally has greater power than the t-test unless data are sampled from the normal. In the case of randomized trials, we are typically interested in how an endpoint, such as blood pressure or pain, changes following treatment. Such trials should be analyzed using ANCOVA, rather than t-test. The objectives of this study were: a) to compare the relative power of Mann-Whitney and ANCOVA; b) to determine whether ANCOVA provides an unbiased estimate for the difference between groups; c) to investigate the distribution of change scores between repeat assessments of a non-normally distributed variable. ⋯ ANCOVA is the preferred method of analyzing randomized trials with baseline and post-treatment measures. In certain extreme cases, ANCOVA is less powerful than Mann-Whitney. Notably, in these cases, the estimate of treatment effect provided by ANCOVA is of questionable interpretability.
-
The counterfactual or potential outcome model has become increasingly standard for causal inference in epidemiological and medical studies. ⋯ Counterfactuals are the basis of causal inference in medicine and epidemiology. Nevertheless, the estimation of counterfactual differences pose several difficulties, primarily in observational studies. These problems, however, reflect fundamental barriers only when learning from observations, and this does not invalidate the counterfactual concept.
-
Bmc Med Res Methodol · Jan 2005
Randomized Controlled TrialA randomised controlled trial to assess the effectiveness of offering study results as an incentive to increase response rates to postal questionnaires [ISRCTN26118436].
Postal questionnaires are widely used to collect outcome data on participants. However, a poor response to questionnaires will reduce the statistical power of the study and may introduce bias. A meta analysis of ten trials offering study results, largely in the fields of education and marketing, was shown to be ineffective, with the odds ratio for response with offering research findings is 0.92 (95% CI 0.75 to 1.11). However uncertainty still exists as it is uncertain whether results from such trials can be extrapolated to that of a health care setting. The aim of this study was to assess whether offering participants study results increases the response rates to postal questionnaires. ⋯ Offering study results to women living in the community aged over 70 does not increase response rates to postal questionnaires. Although researchers have an ethical obligation to offer participants study results, since 10% of women did not wish to receive the results, investigators should give participants the option to opt out of receiving the study's results.
-
Bmc Med Res Methodol · Jan 2005
Comparative StudyAdaptive designs based on the truncated product method.
Adaptive designs are becoming increasingly important in clinical research. One approach subdivides the study into several (two or more) stages and combines the p-values of the different stages using Fisher's combination test. ⋯ It is recommended to apply the TPM rather than Fisher's combination test whenever an early termination due to insufficient effects is not suitable within the adaptive design.