Marking criterion Comments Mark
Background, rationale, objectives (20%) This study is a systematic review investigating the effectiveness of telephone-based assistance interventions on stress, burden, and depression of carers of people with dementia. You draw on relevant data, including the GBD data, to justify your focus on dementia. You also show evidence of widely reading and using a variety of sources.
You claim that ‘Dementia is not only devasting for who suffers from this impairment, but also for those who look after dementia person’. This may be true for some, but always try to be more careful and measured when making claims like this. Language is important, and the suggestion that dementia is always ‘devastating’ and associated with suffering can be stigmatising, and even detrimental to how people with dementia are treated. The language of ‘burden’ should also be used with care for the same reasons.
Given that you discuss the experience of carers, it would have been helpful to see you drawing on relevant qualitative research detailing the actual experiences and perspectives of carers (and also people with dementia) in their own words. This would have added depth and nuance.
Overall, at a structural level the Introduction section includes the expected sub-sections including a description of the condition of interest, some descriptive epidemiology for a sense of scale and impact, description of the challenges faced by carers of people with dementia (CPWD), in particular an attempt to disambiguate the 3 health outcomes of interest (depression, stress, carer burden), a discussion of the 3 interventions of interest (telephone-based counselling, psychosocial support &CBT), an objective, and some discussion of previous systematic reviews in the area. The difficulty when evaluating this work concerns the quality of the writing which makes the content difficult or impossible to understand. Throughout the introduction, the arguments that you attempt to build are frequently lost due to poor grammatical structure or sentences that are unclear. To give just one area as an example, although you have clearly understood the need to justify the conceptual distinction between the 3 health outcomes, you have not provided clear, consensual definitions of each outcomes (with citations) or a clear explanation of how there are different from each other. Similarly, I cannot tell whether there is sufficient distinction between the 3 types of interventions that you focus on, or how your systematic review is carving out a distinct research question compared to the SRs that have gone before.
In your rationale, you state that while other reviews examined the effectiveness of telephone-based interventions, these have included a wider range of non-dementia specific care givers. To justify your focus only on dementia caregivers you say that ‘caregiving for dementia patient is entirely different and distressing (Finkel, 2007)’. This is a really general and potentially problematic statement, that needs further unpacking and explaining, i.e. why is it different and why always distressing?
8.0
Methods (30%) A similar pattern emerges for the Methods section. The overall structure (i.e. the sub-sections you have include and the order of these sub-sections) is good but, in many places, the quality of writing makes the content difficult to understand. There are some elements in the Methods sections that suggest a lack of consistency in the work or a lack of understanding of the concepts you are discussing.
For example: In the Introduction you describe 3 types of intervention but in the Methods under Types of Intervention you appear to list 4 types (i.e. counselling, education about dementia, support services and cognitive behavioural therapy) – what are “support services” and why didn’t explain this in the Introduction? The Exclusion Criteria you present are simply the opposite of your Inclusion criteria which renders the Exclusion criteria redundant – exclusion criteria should identify specific sub-groups from within the population identified by the inclusion criteria (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6044655/) . The search strategy was not as carefully constructed as it could be, e.g. the strategy seems to assume dementia is equivalent to memory loss or cognitive impairment, but these terms do not mean the same thing. Why have you included the search term “female” – are you excluding male caregivers? The section under Evidence Synthesis was not sufficiently clear. It would have also been helpful to see more justification as to how you decided what your exclusion criteria would be. Also, unless I’ve missed it, there seems to be no mention of the timeframe for included papers, i.e. from what year to what year? More detail about the tools being used to measure depression and burden within the reviewed studies would have been helpful. These can vary enormously, again leading to findings that are not comparable.
There are positives in your work. You are attempting to follow existing guidelines to help you conduct and report the SR (PRISMA), and you are using a range of tools (CASP checklists, Cochrane RoB Tool, guidelines for grading evidence syntheses, methods for conducting narrative syntheses). The ambition here is impressive but the execution, specifically the quality of writing, undermines the impact of the work. A little more justification as to why certain choices were made would have been helpful.
11.0
Results (20%) The same pattern is seen here as in the preceding sections. The overarching structure is good – the key elements of a Results section for a SR are present and (mostly) in the appropriate order. However, the content of the various sections (i.e. the writing) again demonstrates problems with the clarity of the sentences throughout most of the Results. The strength of this section is in the way you have systematically examined the evidence you identified, using separate tables to present a detail analysis of various aspects of the studies (e.g. study characteristics, intervention description, findings) and you have attempt to break down your analysis separately by the three different intervention types you specified earlier, and separately by the 3 outcomes of interest. This approach is to be commended. The limitations in the section are about the implementation of these ideas, with the lack of clarity n the written again undermining the impact of your approach.
There were some specific issues that I would have expected you to recognise and comment on. E.g. you failed to notice that 3 studies (Topfer et al, 2018; Meichsner et al, 2019; Wilz et al, 2018) are simply 3 reports from the same study – this should not be triple-counted in your analysis as you have done. Some of the discussion of the various biases were not sufficiently clear, e.g. under Selection Bias, I think you are describing allocation bias but it’s unclear – the point is, nothing should be unclear. You state that Topfer et al (2018) used ‘utilisation of resource’ as a marker of both Stress and Bruden, while Meichsner et al (2019) used a QoL scale to measure stress. This blurring of health outcomes and their and measurement is a fundamental issue that I would have expected you to offer more critical evaluation of the primary studies.
The studies being reviewed seem to cover quite different kinds of intervention including, telephone-based support/ counselling; telephone-based cognitive behavioural therapy, and telephone-based psychosocial intervention. I would like to have seen some discussion of whether these interventions are really comparable, given their very different histories, assumptions and approaches. Alternatively, the authors of the primary studies might be using terminology very loosely and there may be a lot of overlap between interventions labelled as support vs counselling vs education vs CBT – it’s your role as review author to explore and unpack these issues and make the case to your reader.
On a positive note you present some interesting findings in this section, and with improved clarity at the sentence level this would have been a really strong section.
11.0
Discussion (20%) A nice summary table of the findings but this should have been presented in the Results, not the Discussion. Again, there is a sensible structure and some really interesting critical evaluation of the identified studies and analysis here (e.g. highlighting the failure of randomisation in one study; the focus on the nature of the intervention (dose, frequency, content etc.). You discus existing reviews and make a case for how your review is distinct – this could have been more convincing. There is still the issue of the quality of writing which undermines the clarity of your ideas. There are also a few specific issues with the interpretation of the findings, e.g. under Implications for Practice you highlight the need for “further clarification” on the efficacy of telephone interventions for carers of PwD and state there is “no evidence” for the optimal format of the intervention, but despite these apparently serious concerns you say it is “reasonable” to promote telephone based intervention in practice – this seems to be a questionable interpretation! You make some attempts to link your findings into wider literature, although this is quite brief.
9.0
Structure, presentation and referencing (10%) The structure of the work was good; the quality of writing was frequently below what is expected at Masters level, and there were frequent formatting errors in the use of citations/ references (and some references in the body of the text do not appear in the reference lists).
4.0