This is very interesting article and I enjoyed reading it!

I only have some minor comments to make. Design, methodology and data analysis are all sound.

Term use – clinician and nurses – is this the same or deliberately different?
Background: I agree with your statement of the difficulties of researching patients EOLC in ICU. But what is the importance here? The focus in on families. Perhaps add a sentence and link these two?
Consumer? Patient? Client would be another term. They are not value neutral. Please explain what the difference is and thus the impact on quality of care or just adjust language. I am thinking here specifically re the power issues/ differences inherent in these terms.

Nice to see a definition of what family means in this context!
Design: sound and transparent
Inclusion/ exclusion criteria:
Organ donors – this needs an explanatory sentence why they were included.
Withdrawing treatment in these circumstances is perhaps different for families in the sense that they experience a period of ambiguous loss which is absent in other cases.
I am also wondering what you would have done or did do with non-English studies? Was the research team multi-lingual? Perhaps clarify this point. It is an interesting one because we usually set English as language despite many research teams being perfectly able to review studies in other languages.

Data abstraction and synthesis: Was there discussion between reviewers/ analyst around coding/ findings? How were differences dealt with? Perhaps add a sentence to this effect.
What was your epistemological stance towards data? Please add that one too.
As for the second step: again, who did this work? One or two? And how did they negotiate disagreement. This info just needs adding here.
Results:
Preparing the family - Where in any of the studies children and young people considered? Or were they all around adult family members?

This is a nicely written result section. I do not get the link to the conceptual representation (section 3.2) . There is a link missing. ‘Preparedness’ pops up here but isn’t introduced. Do you really need this? It could be integrated into the discussion section which, to me, would make more sense.

No further comments. Issues raised were addressed and this is a well written and very interesting paper!