The first step is to always see if you can find a Meta-Analysis, Systematic Review or Evidence-Based Guideline on your topic. To do this, there are two main places to search:
Because of the nature of these resources, you will not do a full PICO-style search in them. Their search features are much more simplistic, so you would do a standard keyword search to see what comes up in the results. You may not find anything and that is ok. There may be systematic reviews or guidelines from other sources that will appear in your search results when you run a PICO search.
When you create a PICO question, you are actually thinking about how discrete topics relate to each other. Each discrete topic: your P, your I, your C and your O, will probably all have different ways of being expressed when an author writes about them - synonyms. When you use databases or search engines, they only retrieve what you type in, nothing more and nothing related. It is up to you to think of how an author might express your topics in their writing in order to expand the possibility of catching all the relevant articles.
It's helpful to create a chart to stay organized. Place your PICO elements at the top of a column, then write the possible synonyms in the row(s) below.
In infants born prematurely, compared to those born at full term, what is the subsequent lifetime prevalence of sensory deafness?
|P: infants||I: premature||C: full term||O: sensorial deafness|
Now that you have created your synonyms, you need to know a little bit about how databases and search engines work in order to ensure you create a search string that retrieves the types of results you want.
Using our example from above and the operators and other search rules we've learned, we can create a PICO search string that we can type into a one line search box.
(infant OR infants OR newborn OR baby OR babies) AND (premature OR "pre term" OR pre-term OR premie) AND ("full term" OR full-term OR "39 weeks" OR "40 weeks") AND ("sensorial deafness" OR deaf OR "hearing loss")
Notice the rules we've followed: