Do they really expect a systematic review?
Creating your protocol
Developing your protocol
What sources should you search?
Developing the search strategy
Running and recording your search strategies
Managing your search results
What to do next with your search results
Writing and reporting your review
Sources of help
Search more than one source. You should be as exhaustive as you can given the scope, time and resources that you have available. In general a systematic review in the area of health and medicine should search MEDLINE and EMBASE plus another specific database related to the review subject. Don’t just search Google Scholar. It can work well when finding single relevant papers but it is not ideal for literature searching for a review because:
the search interface is quite limited
it isn’t clear what you are searching which makes it difficult to report your searches
different users may get different search results depending on their historical use of Google
saving or downloading all results in one go is not possible
Bibliographic databases are the best resources to use as they allow structured searches using both text and subject indexing, and allow you to save your search strategies and search results easily. Selection of databases and resources will depend on your topic -- be sure to refer to your Subject Guide and the E-resources Guide.
Remember there is a difference between databases (eg MEDLINE, CINAHL Complete) and the search interfaces that host them (eg Ovid, EBSCO). Different databases may be available to search via the same search interface (for instance you can search MEDLINE, EMBASE and PsycINFO databases via Ovid) and then there is no need to search them via other interfaces.
Ideally your systematic review should cover published and unpublished literature in order to avoid publication bias. The unpublished literature is sometimes referred to as grey literature (see the links below). If time and resources allow you should include such literature in your search.