Systematic reviews are considered top of the evidence pyramid, sadly, it also depends on how good is that specific systematic review.

Overall cochrane and campbell reviews are considered good quality.

you can assess the quality of with tools like AMSTAR.

https://www.bmj.com/content/358/bmj.j4008

Hope this helps.

The Milbank Quarterly (SCI, IF 4.375) - The Mass Production of Redundant, Misleading, and Conflicted Systematic Reviews and Meta-analyses: Volume 94, Issue 3 (pages 485–514) DOI: 10.1111/1468-0009.12210 Published in 2016

Retrieved from https://www.milbank.org/quarterly/articles/mass-production-redundant-misleading-conflicted-systematic-reviews-meta-analyses/

Sys reviews are better then nothing which large scale randomize trials are not available.

it’s important to pay attention to the methods of the review. Have more studies been published since the systematic review was written? Reading a systematic review carefully will help you with determining what research you need to do next.

They can be intentionally or unintentionally misused, just like RCTs. The trust under question would be in the methodology not in the one using it or in a particular result. The methodology is not perfect!

Thank you all for actively participating in this discussion and providing valuable inputs. To my understanding, systematic review (if done properly) may be a good tool to arrive at a conclusion, when we do not have anything else to look at. Its value in terms of pure research is doubtful. The very premise with which a systematic review is initiated (most of the times to publish a paper) defeats the fundamental core of research- being unbiased.

I sense publications drive systematic reviews. The most influential publishing house get cited in systematic reviews, more often than not, thereby further increasing their influence in what I loosely term as " The politics of systematic reviews"

Looking forward further opinions on this.

I agree with the others that say SRs are only as good as the methods used to perform the review. Just like primary literature, they need to be critically appraised. As a librarian, I always look at the search strategies and whether or not a professional searcher, such as a librarian, was on the team. A poor search strategy already biases the results by possibly missing relevant articles, and I see a lot of poor searches in published SRs. In terms of publication bias, a good SR will also search unpublished literature, conference proceedings, trial registries etc., as well as handsearch reference lists of relevant papers and journals so as not to miss anything. If done well, a good SR can synthesize a body of evidence to draw a conclusion that can help drive evidence-based practice. I agree that Cochrane and Campbell are tops and offer excellent resources to help SR authors. There are many other helpful guides and tools available.The IOM Standards for Systematic Reviews still offer a clear pathway for comparative effectiveness research of therapeutic or surgical interventions: http://www.nationalacademies.org/hmd/Reports/2011/Finding-What-Works-in-Health-Care-Standards-for-Systematic-Reviews/Standards.aspx

Thanks Elizabeth for your valuable inputs, in particular the perspective of a librarian. Can we think of an alternative search strategy which reduces "Bias"?

https://m.youtube.com/watch?v=Uok-7NPFn4k

I assume you have access to a search engine. If you put in the search box "assessing the quality of systematic reviews," you will see that there is a literature on this subject.

The Milbank Quarterly (SCI, IF 4.375) - The Mass Production of Redundant, Misleading, and Conflicted Systematic Reviews and Meta-analyses: Volume 94, Issue 3 (pages 485–514) DOI: 10.1111/1468-0009.12210 Published in 2016

Retrieved from https://www.milbank.org/quarterly/articles/mass-production-redundant-misleading-conflicted-systematic-reviews-meta-analyses/

Thanks Stephen, I do have access to search engines, but i wanted opinion from researchers who have been taught time and again that Systematic reviews (with meta analysis) are considered highest on evidence. I have and will continue to have doubts on this.

Thanks Claire for sharing this article. I think it clearly proves as to where should we all be i terms of considering best for evidence.

In times when we publish on account of pure joy for publishing to getting promoted, a fully detailed original article provides better evidence than a systematic review.

And add this to systematic reviews and you have the perfect recipe:

Lundh A, Barbateskovic M, Hróbjartsson A, Gøtzsche PC (2010) Conflicts of Interest at Medical Journals: The Influence of Industry-Supported Randomised Trials on Journal Impact Factors and Revenue – Cohort Study. PLoS Med 7(10): e1000354.

I am confused by what you are looking for. I doubt you believe that no evidence is better than some evidence. Once one has evidence, then it is possible to ask: How good is this study? What are my concerns about it? Is there a way to get assurance about my concerns? - Ideally these are questions that would arise with good peer-reviewing.

Sometimes, persons wanting answers to specific questions that do not appear addressed in a published article will communicate directly with the authors (this is particularly an issue for persons trying to replicate a study) since no publication contains all the details.

You seem particularly concerned about conflict of interest. That is a very important concern, and as we all know, despite the fact that submissions to good journals require disclosure of potential conflicts, there are people in seemingly excellent institutions who conceal their conflicts. Sometimes that affects either what they publish, e.g., hiding results they didn’t want to see; sometimes it affects how they interpret and “advertise” their results; and sometimes their cheating on revealing their conflicts has deeper roots in that the research itself is bogus. But, let’s assume that there are enough other studies on that subject to do a systematic review or meta-analysis. Then the bogus study should be an outlier.

Overall, I still do not see what you are driving at. Opinion from others, including me, will only tell you that the quality of systematic reviews, the quality of the underlying studies, and the quality of the journals in which both were published (the reviews and underlying studies), all vary. Nonetheless, there are good journals, good underlying studies by good investigators, and good systematic reviews from good analysts.

Hi Sunil, am I right in thinking that your query is that SRs are inherently more focused on efficacy whilst in clinical practice, clinical/ cost Effectiveness are far more important?

Thanks Stephen and Thanks Arockia.

I think Stephen has summed it appropriately that the quality of systematic reviews, the quality of the underlying studies, and the quality of the journals in which both were published (the reviews and underlying studies), all vary.

I am worried as I am observing that the proportion of systematic reviews now published as part of total number of publications is increasing substantially and with the sole aim of publishing an article and not to advance an evidence. A small percentage of these reviews are registered, therefore rendering them liable for bias.

Lastly, i feel that a poor quality systematic review has a higher potential of causing harm in comparison to a single original article.