I am positively surprised that the number of institutions considering this study design as acceptable for biomedical/clinical PhD programs is nearing half in the European survey that you cited.

In the Philippines, the prevailing academic view on systematic reviews (SRs) is much more conservative than the estimates in the aforementioned European survey. This is perhaps compared to the total number of biomedical/clinical/life science researchers in the country, most, if not all, of those who do SRs are in the medical-surgical field - even if it can be argued that the benefits of doing an SR permeate across these fields and beyond. Nevertheless, those who do SRs in medicine (1) form a minority of clinical researchers in the country, and (2) mostly do it as a research requirement for their postgraduate medical programs (residencies and subspecialty fellowships) rather than as a culminating endeavor for academic graduate programs. Furthermore, the country has yet to have a doctorate-level clinical epidemiology program, which I think will be completely amenable to SRs as dissertation requirement (albeit perhaps the more complex versions such as SR-network meta-analysis or SR-individual patient data meta-analysis).

One of the biggest issues raised against SR as a study design for doctoral dissertation work is, which was also raised in your cited survey, is whether working on this framework alone allows one to generate novel knowledge deemed academically adequate. Thorough understanding of conducting and interpreting SRs will perhaps do little to change the view of those who generally disagree. I believe that this is because SRs, by definition, utilizes secondary data with all the limitations and biases contained therein. At this point, I think that SRs will be more acceptable as doctoral-level work if it is a component/part of a larger work (e.g., by attempting to address a question in a way that is more comprehensive and innovative than previous studies, then attempt to analyze the doctoral candidate-generated primary data with those from the previous studies).

Ultimately, I think that there is a rightful, although gradual, increase on the acceptance of this design for the academic level of interest. The culture and standard by which doctoral work is judged should be moved at least partially from individual to team focus (most relevant research nowadays emerges from collaboration rather than from isolation anyway) and from novelty to relevance.

Though I am no longer a PhD student, I conducted a systematic map as part of my PhD thesis.

I had not originally considered using a systematic review as a research method, but as I conduct SR as part of my research work at UCL regularly, I understand the value that SRs can add. SRs are often conducted at masters level but are not so widely utilised for PhD research. SRs are particularly useful when undertaking research in new/emergent areas of research or where the research landscape is dense or convoluted.

The issue for PhD students is that there is an element of stigma attached to conducting SRs, in the context of demonstrating a student's advanced research skills. However, I feel that this is particularly in just, in that SRs are a rigorous method of gathering data and understanding research landscapes. SRs have evolved considerably, moving away from purely being used to understand intervention effectiveness, e.g. for clinical trials of pharmaceuticals. Now we see rather various applications of SRs, being used to understanding complex policy issues, patient experiences, and quality of components of interventions,to name but a few. I believe that when new areas of research are being examined SRs have a place at the high table due to their ability to map evidence, identify research gaps, combine findings to understand effectiveness of many aspects of peer reviewed research.

I used the SR map to look for evidence of children's views about medicines research and development. The use of a systematic map provided insight into whether children had been asked their views about medicines, and whether children had insight that could inform medicines R&D. This negated the need to conduct focus groups with children on this topic. Indeed, the cumulative number of children views elicited from the SR map, was far greater than I could have achieved for my research project.

It must be noted that the systematic map was one of a number of research methods that I employed and so was welcomed by the examiners who recognised its value from the offset. Had I employed only a SR for my research, then this might not have been the case.

In summary,I would argue that SR or systematic maps or gap maps are highly valuable. SRs help PhD researchers to have a solid understanding of their chosen subject matter and ensuring that their research explores areas where research evidence is lacking.

Thank the both of you for these very interesting insights from different parts of the world. That is really interesting. I think there is most likely some uneven distribution on how SRs are valued (that is not meant in a geographical sense). The source cited above was only able to look on universities that offer a PhD graduation program. So in Germany for instance only 4 out of 36 medical faculties were assessed. Some of the ones that were not assessed officially disagree with SRs while others are not so open the definition of 'original publication'. Often they are part of a cumulative dissertation with in most cases the necessity to graduate with more than one publication. Therefore there is a good way to implement SRs as a part of a bigger project.

Aedrian Abrilla: from what you mention I see similar arguments against SRs which I have difficulties with, considering that SRs form the hightes level of evidence. So arguments in the direction that no new knowledge or data is generated sounds somehow weird, conceptually.

Gillian Stokes: I like very much how you bring up the benefit for an individual project of performing any sort of evidence synthesis and even the more complex ones. Especially the value it adds to a broader research project as a foundation to better ground a theory or put findings into perspective. Very much appreciated. You are bringing up the point of even more sophisticated SR methods. Were systematic maps also eligible without problems?

Hope to get some more insights here.