Have you ever meta-analysis you didn't like?
Article Abstract:
Meta-analysis is the process of combining data from several studies to determine the effectiveness of specific treatments. The data obtained from the studies must be compatible to be used for meta-analysis. A recent investigation used meta-analysis to examine the effectiveness of steroids in treating chronic obstructive pulmonary disease. This lung disorder involves a disease process that results in the decreased ability of the lungs to perform their function of ventilation. The use of meta-analysis in the study of therapeutic effectiveness is discussed. Meta-analysis differs from a review in several ways; it tends to: (1) focus on a single clinical question, which may concern treatment, cause, or accuracy of a diagnostic test; (2) combine results from several studies to give a more precise estimate; and (3) amplify an important main or subgroup effect that may be too small to be measured accurately in individual trials. The process of meta-analysis involves: (1) evaluating the quality of the information provided by separate studies; (2) quantitatively assessing the studies to ensure similarity between trials in order to combine their results; (3) providing a sense of what is being combined, possibly through graphical presentations; (4) and statistical pooling. The summary of a meta-analysis may not be useful to the clinician who wants to focus on a single trial. However, the meta-analysis summary is useful in explaining differences between clinical trials; revealing weaknesses and strengths of a specific body of research; and providing clinicians with an objective evaluation of the research literature. (Consumer Summary produced by Reliance Medical Information, Inc.)
Publication Name: Annals of Internal Medicine
Subject: Health
ISSN: 0003-4819
Year: 1991
User Contributions:
Comment about this article or add new information about this topic:
The P Value Fallacy
Article Abstract:
The Bayes factor is a better statistic for analyzing medical research than the P value. The P value indicates the likelihood that an experimental result simply occurred by chance, rather than as a result of the experimental drug or procedure. The Bayes value separates the strength of the evidence from the long-term aspects of the intervention, and is a more useful for evidence-based statistical analysis.
Publication Name: Annals of Internal Medicine
Subject: Health
ISSN: 0003-4819
Year: 1999
User Contributions:
Comment about this article or add new information about this topic:
The Bayes Factor
Article Abstract:
The Bayes factor in medical statistics indicates the likelihood that an experimental effect is significant, and the probability that conclusions drawn from the experiment are correct. Compared to the P value, which only describes the influence of chance on the outcome of an experiment, the Bayes factor contributes to the analysis of the experimental data.
Publication Name: Annals of Internal Medicine
Subject: Health
ISSN: 0003-4819
Year: 1999
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: Posterior mediastinal mass in a black patient with back pain. Clinical trials partnership seeks to boost research in developing countries
- Abstracts: Optimal time in the menstrual cycle for serum progesterone measurement to diagnose luteal phase defects. Accuracy of gestational age estimation by means of fetal crown-rump length measurement
- Abstracts: Treatment of recurrent herpes simplex labialis with oral acyclovir. Early application of topical 15% Idoxuridine in dimethyl sulfoxide shortens the course of herpes simplex labialis: a multicenter placebo-controlled trial
- Abstracts: Subcutaneous interferon alpha in the treatment of refractory condylomata. First-void urine testing for Chlamydia trachomatis by polymerase chain reaction in asymptomatic women
- Abstracts: Detrusor instability - current management. Symptoms analysis for the diagnosis of genuine stress incontinence