Skip to Main Content

Problem-based Learning & Finding the Evidence: Evidence on Prognosis

Prognosis: PICO Question

Designing your PICO question

Putting your search query in the PICO format is really just about structuring your clinical search and providing reminders about what criteria are important when evaluating the articles.  

In this case, I didn't fill out all of the slots of the PICO because I didn't have the information and it really wasn't that necessary - I can get good results with this particular search with just the 2 items, P & O.  So I left it as:

   P:  3 yo African-american female, possible neuroblastoma (stage, when known would be good to add)

   O:  Prognosis

 

Prognosis: PubMed Search

Performing your PubMed search

This search is a fairly simple one and you might be tempted to use just a couple of key words.  However, you have *lots*of sorting to do if you don't include enough words and a limit.  With just a little bit of planning, you can get a great (small) set of articles.

So the obvious ones to use are prognosis and the disease in question - say neuroblastoma. 

By the Numbers:  prognosis neuroblastoma  = over 5,000

 

Because a well structured PICO gives you additional things to consider --the patient's age

1)  Step 1:  Apply the filter for patient's age group


How would you search this concept?  Is child or pediatric better?  Maybe children would be better?  Should you use all of them?  PubMed has a tool for that -- Limits for age.  Although it is hidden, check along the left side of the results page to "Show additional filters" and add the Ages group.  See image below.  You can now select Child

 

      By the Numbers:  prognosis neuroblastom  with Filter for Child  =  over 3,000

 

Better, but still too many to scan through.  Again, think about the info listed in the PICO, what else could help  --  the patients ethnicity.  You have learned about many different diseases that affect groups of people differently, so it can be helpful to include that information on some searches.

2)  Step 2:  Include the patient's ethnicity

   Ethnicity can be a little tricky, especially with a key word search.  In this case, what is the right term? Think like the author of the articles -- how would they have described ethnicity?  Could it be Black?  African Americans? Something else?  I would suggest using at least these 2 in a combined group of (black OR african americans).

      By the Numbers:  prognosis neuroblastoma (black OR african americans)    =  15

And this set includes a short list of good articles done within the last few years.