A good scientific review used to contain a discussion of results and ideas. That assumes a relatively high level of the review's author(s). For the early stage researchers, it is an opportunity to learn a lot about their upcoming scientific field, and a quite significant challenge.
A review is not just a collection of article conclusions and actual data, but it is a critical analysis of them. It is important to start working on it, for example, with a checklist of articles, studied subject matters, methodologies, methods, materials, propertied, results, and ideas.
Then, during the work on the review - to change / correct priorities of listings, subordination of importance, etc. A historical approach as to how this field evolved from its previous roots, what it has become by today, and what it is striving to achieve in the future would be highly desirable.
Any tips how to start writing for a scientific review paper?
You can also consider the following:
Start with a journal(s) that you want to publish in mind.
Gather & comply to the scientific review paper submission criteria e.g. format, no. of pages, font size, line spacing, figure & table, references etc. (note: submission criteria can be vary from journal to journal).
Plan out what are the section headings you want to include e.g. background / introduction (including research problem, research objectives, research questions), literature review etc.
Perform rigorous literature review - you can use the following Excel (see RG link below) to record what literature you'd reviewed, analyzed & synthesized etc. (the filled Excel can help you to write the literature review section later).
Fill up all the sections of your scientific review paper - perform numerous rounds to ensure completeness & flows of thoughts.
Get others e.g. peers, colleagues etc. to review your paper & do the needed improvement before you submit your paper.