Searching the evidence is not an easy task especially when the amount of papers published is literally overwhelming. It is hard to know where to start and even Google is not helping when the evidence sought is pretty much specific.
This is a simple guide of how you can perform an effective search to identify potential articles for a particular topic in health care context.
Think of your question:
Not every question that comes through your head is scientific. You need to know exactly what you are looking for and to do that consider these question:
- What patients/ population you are concerned with?
- What intervention/indicator you want to examine?
- What outcome/tool you want to measure?
- Are comparing this with something else?
These questions are usually called PICO ( Population, Intervention, Comparison and Outcome). It is used to formulate research questions and trust me formulating sound research questions is not an easy task thus I will be dedicating one separate post for it.
Choose your keywords:
According to dictionary.com (2013) a keyword is defined as 'a word that serves as a key, as to the meaning of another word, a sentence, passage, or the like'. However, technically a keyword is 'a word used to classify or organize digital content, or to facilitate an online search for information'.
Okay sounds complicated I know. To say it simply, search engines can not answer your research question directly and if you do search for the question as it is in PICO format you are more likely to get no results. this is due to the system which search engines operate with. According to Portney and Watkins (2009) search engines are 'information retrieval systems that search the internet and electronic data bases for websites, files or documents based on key words or phrases'. This means that search engines use keywords to classify the content of the papers. Consequently, in order to obtain the article related to your question you need to select appropriate keywords to attract the desired literature.
choose an appropriate database:
Databases are 'Organized systems that allow search access for specific content or information' (Portney and Watkins 2009). Think of it as bank of information that stores specific data only. Some of these databases requires you to pay subscription fess to get full access tot he articles while others do not require subscription and free text is offered. Abstracts are usually free to access and they are powerful tool to get a quick insight about the paper's content. It is important to know that databases are equipped with search engines to retrieve the abstracts but not every search engine has access to these databases. There are many database available online such as Medline, CINHAL or PEDro..etc. and each one is offering unique content in specialized field or discipline so make sure to do some search before selecting your databases.
Use Boolean Commands:
Boolean commands (A.K.A Boolean logic) are powerful simple commands that most search engines use to narrow down the search results. Say you are investigating the relationship between lung cancer and smoking and you want articles that study this relationship only. You can get the articles by entering (lung cancer AND smoking) if you notice that the articles you got also include another factor for example heart disease and you want to exclude that. To do so you need to enter ( lung cancer AND smoking NOT heart disease). There are other Boolean commands such as:
- OR: used in case you are given the search engine choice between two variables (lung cancer AND smoking OR heart disease).
- + (Plus sign): it is used when two keywords must appear in the same abstract.
- - (Minus Sign): Used to exclude abstracts that have two keywords together.
After you use these steps you should read/scan the abstracts that the search process yielded and from their choose the potential abstracts that might hold an answer for your question.
Okay now you know some basic so you can go ahead and try apply these steps and I would recommend you start with PubMed.
If you like this post please subscript and if there is anything I missed I would be pleased to know it using the comments section.