Greetings sir. I would recommend searching the keywords in the title, and abstract, and finding relevance in the full text, meaning if the title and abstract had relevant concepts in regard to your search strategy, research question, and inclusion criteria, at the third step evaluate for relevancy in the full text. it is important to consider the procedure for conducting a systematic review is more like a guideline and every step must be reported to the slightest detail, therefore taking the approach of Title screening ---> abstract screening ----> full-text screening is definitely adjustable with a systematic review approach. on the other hand regarding the keywords, if your keywords are difficult to find maybe they aren't the most relevant ones to your research question after all, if possible selecting new keywords and conducting and executing a new search process is preferable if that's the case. I understand that you have reasons for choosing those keywords but what will be their use if they cannot identify relevant studies to include and help in conducting a detailed work? moreover, since SLRs have very detailed methodologies, journal reviewers who are definitely more experienced than many of us, will notice in your methodology section and compare it to your text if your keywords and search strategy are well executed and whether details such as keywords or the starting and end date for your search process are logically selected, so starting a new search with filters would be a good guarantee. last but not least, remember one of the aims of the filtration process is to exclude as many irrelevant studies in the search process as possible to provide the most precise results, so don't shy about excluding those that are irrelevant but make sure to keep the count of the exact number of the studies you excluded because of the lack of relevance since you'll need to report that precisely in the methodology. Hope that helps

I would recommend you will utilize machine learning platforms as mySRL, ASreview or Codima because selection article work is least tough. Is important that your word string it found all article on topic. Personally for scopus I use TITLE-ABS-KEY. For pubmed and WoS only use ALL+ string.

Dear Mr,

Another tip : you can search systematic review at your subject (or not) : you have to add keywords such as review*, "state of the art" or survey* to your search strategy. And then read the section on the methods used. This will help you develop your own search strategy !

On the specific databases, have a look at the guides they offer, it helps.

Otherwise, the default fields (title, abstract, keywords) are perfectly adequate in my view.

Best regards