China
Africa
Explore chapters and articles related to this topic
Fixed Effects and Random Effects in Meta-Analysis
Published in Ding-Geng (Din) Chen, Karl E. Peace, Applied Meta-Analysis with R and Stata, 2021
Ding-Geng (Din) Chen, Karl E. Peace
Publication bias is sometimes referred to as selection bias. In meta-analysis, the studies selected to be included are vital to the inferential conclusion. Publication bias could arise when only positive studies (those that demonstrate statistical significance or if not statistically significant do not reflect qualitative interaction) of a drug are published. Therefore even though all published studies of a drug for the treatment of some disease may be selected for a meta-analysis, the resulting inferential results may be biased (may overestimate the efficacy of the drug). The bias may be particularly significant when meta-analyses are conducted or are sponsored by a group with a vested interest in the results.
Introduction to Systematic Review and Meta-Analysis
Published in Christopher H. Schmid, Theo Stijnen, Ian R. White, Handbook of Meta-Analysis, 2020
Christopher H. Schmid, Ian R. White, Theo Stijnen
The extent to which a meta-analysis captures the truth about treatment effects depends on how accurately the studies included represent the populations and settings for which inferences must be made. Research gaps represent studies that need to be done. Publication bias and reporting bias relate to studies that have been done but that have been incompletely reported. Publication bias refers to the incorrect estimation of a summary treatment effect from the loss of information resulting from studies that are not published because they had uninteresting, negative, or non-statistically significant findings. Failure to include such studies in a review leads to an overly optimistic view of treatment effects, biased toward positive results (Dickersin, 2005).
Essential knowledge and hot topics
Published in Shelly Griffiths, Surgical Interviews: The Survival Guide, 2019
Information bias describes incorrect recording of measurements. This can be due to inaccurate measurements (measurement bias), incorrect classification of a categorical variable (misclassification bias), under- or over-reporting of a particular variable (observer or assessment bias), over-reporting of ‘interesting’ findings (reporting bias) or a tendency to report positive and not negative findings (publication bias). One of the biggest challenges when performing systematic reviews is taking into account the effect of publication bias. Funnel plots can be used to check for the existence of publication bias in systematic reviews or meta-analyses. The basic assumption is made that the results from the largest studies will be near to the average and the smaller studies spread symmetrically on both sides. Any variation from this assumption (i.e. any asymmetry) can indicate publication bias. Steps to reduce publication bias have been made by encouraging trial registration whereby high-quality trial protocols are registered and published before the trial accrues data.
World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for treatment of anxiety, obsessive-compulsive and posttraumatic stress disorders – Version 3. Part I: Anxiety disorders
Published in The World Journal of Biological Psychiatry, 2023
Borwin Bandelow, Christer Allgulander, David S. Baldwin, Daniel Lucas da Conceição Costa, Damiaan Denys, Nesrin Dilbaz, Katharina Domschke, Elias Eriksson, Naomi A. Fineberg, Josef Hättenschwiler, Eric Hollander, Hisanobu Kaiya, Tatiana Karavaeva, Siegfried Kasper, Martin Katzman, Yong-Ku Kim, Takeshi Inoue, Leslie Lim, Vasilios Masdrakis, José M. Menchón, Euripedes C. Miguel, Hans-Jürgen Möller, Antonio E. Nardi, Stefano Pallanti, Giampaolo Perna, Dan Rujescu, Vladan Starcevic, Dan J. Stein, Shih-Jen Tsai, Michael Van Ameringen, Anna Vasileva, Zhen Wang, Joseph Zohar
Moreover, there are a number of reasons why the effects of psychotherapy may be overestimated in the available studies. In waitlist studies, the patients are not blinded to the treatment arm (Bandelow 2021). Patients treated in psychotherapy trials were significantly less severely ill than patients in drug trials. In the vast majority of ‘pure’ psychotherapy studies, patients were not excluded when they were on medication (Bandelow et al. 2015; Skapinakis et al. 2016). Thus, in some patients, the measured pre-post effect is in fact a combination of psychotherapy and medication in a substantial percentage of the participants. Only in half of the psychotherapy studies, intent-to-treat (ITT) analyses were used for missing data. Drug trials were significantly shorter than psychotherapy studies. The longer the trial, the more the treatment effects could be attributable to spontaneous remission. Sample sizes in drug trials were about five times larger than in psychotherapy studies, due to the lack of commercial funding in the latter. Small studies are more prone to publication bias, as journal editors may be reluctant to publish small studies showing no difference from a control group. Publication bias is an issue both in pharmacotherapy and psychotherapy studies (Driessen et al. 2015; Turner et al. 2008).
Rare forms of inflammatory myopathies - part II, localized forms
Published in Expert Review of Clinical Immunology, 2023
Claudio Galluzzo, Ilaria Chiapparoli, Ada Corrado, Francesco Paolo Cantatore, Carlo Salvarani, Nicolò Pipitone
Rare diseases present the clinician with many challenges, and localized myositis is no exception. First, there is generally a lack of adequate knowledge about them. This often translates in delayed diagnosis, misdiagnosis, or both. Localized forms of myositis, in particular, are easily prone to be misdiagnosed as tumors. Second, they often remain poorly characterized because the information about them derives from scattered case series and not infrequently single-case reports. Therefore, it is a cumbersome task to put together the various published date to generate a somewhat coherent picture of these forms. The risk of bias is especially high, since case reports or small case series may highlight some striking features at the expense of others, thus hampering a balanced view of the disease’s picture. The risk of bias is perhaps even higher when treatment is concerned, because cases are more likely to be published if they report successful therapies (‘publication bias’). Lastly, while systemic inflammatory myopathies are usually managed by Rheumatologists, it is less clear who should be in charge of managing localized myositis. A case in point is orbital myositis, which mostly present with ocular symptoms and signs, thus typically resulting in referrals to the Ophthalmologists. However, most Ophthalmologists would simply rule out a primary eye disease and declare orbital myositis not to fall under their realm of expertise. At the same time, the majority of Neurologists and Rheumatologists are just as unfamiliar with this condition.
Safety and efficacy of transcatheter aortic valve implantation in stenotic bicuspid aortic valve compared to tricuspid aortic valve: a systematic review and meta-analysis
Published in Expert Review of Cardiovascular Therapy, 2022
Mohamed Zghouzi, Heba Osman, Waqas Ullah, Abdul-Rahman Suleiman, Parveen Razvi, Mukhlis Abdalrazzak, Firas Rabbat, Mowaffak Alraiyes, Yasar Sattar, Rodrigo Bagur, Timir Paul, Andrija Matetic, Mamas A. Mamas, Nasser Lakkis, M. Chadi Alraies
The DerSimonian and Laird test on a random-effects model was used as a statistical analysis method to calculate odds ratios (OR) for dichotomous variables. Subgroup analysis stratified by follow-up duration (In-hospital, 30 days, and 1-year) was also performed. In order to perform sensitivity analysis, ‘leave-one-out’ strategy was performed to evaluate the influence of individual studies on pooled estimates. To assess variations in outcomes of the included studies, Higgins I-squared (I2) statistical model was used. Publication bias was illustrated using a funnel plots and calculated quantitatively using Egger’s tests. The methodological quality assessment of the included trials was done as well and performed using the Oxford scoring – Jadad scale and the risk of bias (Rob-2) tool. The Newcastle–Ottawa Scale was mainly used for the bias assessment of OCS. The ‘test for overall effect’ was reported as the z value with its 95% confidence interval (CI). The probability value of p < 0.05 was considered statistically significant. All statistical analyses were performed using STATA version 16.