Timeline for Is it possible to have exactly identical output of random effects and fixed effect models in a network meta-analysis?
Current License: CC BY-SA 4.0
17 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 21, 2019 at 11:39 | vote | accept | Prabhakar | ||
| Jul 18, 2019 at 5:18 | answer | added | Björn | timeline score: 1 | |
| Jul 18, 2019 at 3:11 | history | edited | Prabhakar | CC BY-SA 4.0 |
deleted 2 characters in body
|
| Jul 17, 2019 at 11:11 | comment | added | user10619 | I do not know how tau square is estimated to be zero. Its formula suggests that it should not be zero. Genrally, we shoud expect certain differences in the effect-sizes. Default setting could be zero. | |
| Jul 17, 2019 at 11:02 | comment | added | user10619 | We develop measures of the impact of heterogeneity on a meta-analysis, from mathematical criteria, that are independent of the number of studies and the treatment effect metric. We derive and propose three suitable statistics: H is the square root of the chi2 heterogeneity statistic divided by its degrees of freedom; R is the ratio of the standard error of the underlying mean from a random effects meta-analysis to the standard error of a fixed effect meta-analytic estimate, and I2 is a transformation of (H). This suggests that it is not a standard tool for testing the heterogeneity. | |
| Jul 17, 2019 at 9:05 | comment | added | Prabhakar | @SubhashC.Davar Method to calculate I^2 is published in literature and used commonly in evidence synthesis (check Higgins and Thompson [2002]). The netmeta package I mentioned in my question uses the same method to quantify statistical heterogeneity. Effect size was mean diff.; however, I don't think the class of effect size can explain my question. | |
| Jul 17, 2019 at 6:50 | comment | added | user10619 | I^2 = 0. How did u measure ? The formula ? | |
| Jul 16, 2019 at 16:21 | comment | added | user10619 | exactly identical output of random effects and fixed effect models , which output are you talking of ? what are your effectsizes ? | |
| Jul 16, 2019 at 2:07 | history | edited | Prabhakar | CC BY-SA 4.0 |
Provided more details about my study especially highlighting the low level of heterogeneity and limited number of studies
|
| Jul 15, 2019 at 16:16 | review | Suggested edits | |||
| Jul 15, 2019 at 16:56 | |||||
| Jul 15, 2019 at 13:10 | comment | added | Giuseppe Biondi-Zoccai | I don't recommend to exclude studies based on their findings, but only given their design features (eg randomized trials included, non-randomized trials excluded). A star-shaped network is a network where most treatments have been compared to a single agent (eg placebo). | |
| Jul 15, 2019 at 10:51 | comment | added | Prabhakar | @Joe_74 thank you for response! The number of studies is 10 with a total of 32 treatment arms. I understand getting exactly same results is not an unusual finding. Could you please also comment on my approach of excluding studies from analysis to manage heterogeneity, is this approach reasonable? Could you please also elaborate more on the "star shaped network" you mentioned? | |
| Jul 14, 2019 at 14:25 | comment | added | user10619 | Without a summary of results and a brief picture of your study, it is difficult to answer your concerns. | |
| Jul 14, 2019 at 11:29 | comment | added | Giuseppe Biondi-Zoccai | If heterogeneity is very low and the number of studies limited I think it is perfectly OK. Indeed, report both results and highlight this finding. It could also depend on the fact that your evidence network is mostly star-shaped (amounting mostly to an adjusted indirect comparison...). | |
| Jul 13, 2019 at 18:03 | comment | added | Prabhakar | ndeed, heterogeneity is very low (tau^2 = 0; I^2 = 0%). Actually, I removed two of the studies from this analysis that were resulting in very high heterogeneity (I^2 >95%). I did pairwise meta-analyses to identify those studies and removed them from the main NMA, kept them for a sensitivity analysis, which I haven't done yet. Could you please suggest if this approach is reasonable? Could you please also elaborate more on consequences of having low heterogeneity? | |
| Jul 13, 2019 at 12:30 | comment | added | mdewey | Low heterogeneity? | |
| Jul 13, 2019 at 7:38 | history | asked | Prabhakar | CC BY-SA 4.0 |