Pediatric Fellowship Course Seminar Series

2013Health Sciences Library Resources: Searching Questions and Identifying

ResearchFebruary 12, 2013

Janice Lester, MLSReference and Education Librarian - LIJ Health Sciences

Library

Jennifer L. Boxen, MLS, AHIPEducation and Liaison Librarian - Hofstra North Shore-LIJ

School of Medicine

Objectives/Outcomes for this session:

1. Identify a clear structured searchable question.

2. Execute an appropriate search strategy to search the literature for specific types of study designs based on the topic.

3. Determine appropriate resources to answer background or foreground information questions.

Session Timeline:

7:30-7:45 Learning Objectives and Creating a Searchable Question

7:45-8:00 Finding Library Resources and

the Evidence Pyramid

8:00 – 8:30 Effective PubMed Search Techniques and Q&A

Tips When SearchingBe MethodicalLeave yourself adequate time to

search…But don’t spend too much time doing it

Document where you searched, what you searched and when you searched

Store all of your citations in the same place(EndnoteWeb or Zotero)

Seek help when you need it, but do not wait until the last minute

Steps of Basic Research1. Create an answerable research

question2. Break your research question

into searchable components 3. Choose the database4. Generate subject headings and

synonyms5. Execute your search and select

limiters6. Analyze your search results,

modifying and re-executing searches as needed

7. Review the Articles

Background Versus Foreground:Experience Determines Need

Background Questions- About conditions

Foreground Questions- About choices

Create an Answerable Research Question based on your Hypothesis

Know the difference between background and foreground questions.

Background Questions:

Ask for general knowledge about a condition or thing.

Have two essential components:

A question root (who, what, when, etc.) with a verb

A disorder, test, treatment, or other aspect of healthcare

Foreground Questions:Ask for specific knowledge to

inform clinical decisions or actions.

Usually have 3 or 4 essential components ◦ Patient and/or problem ◦ Intervention ◦ Comparative intervention

(not always needed) ◦ Clinical outcome

Break your research question into searchable components

PICOPatient/Population – Includes age, race,

sex, geographyProblem – Current health concern Intervention – Exposure of interestComparison – Alternate exposure (if any)Outcome – What is the desired outcome?

Does this mean that every research question can/should be answered this

way?

Creating the QuestionCreate one sentence (elevator statement)

that epitomizes what concept you are looking to search which includes PICO elements.

Does giving Prophylactic Acetaminophen to infants (age 2 months) after immunizations decrease sleep duration compared to a placebo?

What are our PICO Elements?

Bad QuestionWhat is the best way to treat children with asthma?Best how? Fewer asthma attacks? Less severe

attacks? A treatment that is cheaper? Fewer side effects? Fewer adverse reactions with other medications?

What type of asthma? Chronic? Exercised Induced? How are we diagnosing asthma and who is doing it?Best compared to what? What type of treatment are

we considering?Is there a more specific age group that we are

addressing? Are we measuring any of these concerns? If so, how?

Good Question

Among young children with acute asthma exacerbation, is a single dose of IM dexamethasone comparable to five days of oral prednisolone for resolution of asthma symptoms?

More detail is usually better; not always possible, but generally better.

Selecting and Searching the Databases

Wait a minute… Where did you say the databases are again?

Accessing EMIL from Healthport

Access through LIJMC/CCMC

Finding e-textbooks on EMIL by subject – eg. Pediatrics

Finding journals and books by subject

Finding journals – e-Journal portal 360 Link-search by Pubmed or DOI

Pediatric Care Online

ACP Journal Club summarizes the best new evidence for internal medicine from over 130 clinical journals accessible through ‘E-journal portal’ on EMIL

Example of a Point of Care - Evidence Based Clinical Database

Available RemotelyTutorials AvailableCan be accessed through mobile

devices

Locating the Pubmed and OvidSP

PubMed is Useful For the Following:

When searching for the newest information (pre-pub)

Very obscure cases [bot fly]Special types of articles (case

reports, trials)Articles that have been corrected

or retractedLimiting by specific factors, like

females and ageSeeking Higher Levels of Evidence

and specific search filters

Tutorials

Access PubMed Through HealthportNot the Internet

Back to Our Clinical Questions…

Find all of the searchable elements of the sentence. Then look up the appropriate subject headings and two synonyms (usually). If you can’t find the exact subject heading, do the best you can. If you need more than 2 synonyms, add them.

Our QuestionWill providing patient education reduce

admissions for head trauma caused by parental abuse or neglect of children?

PICOPATIENT = ChildrenPROBLEM = Parental abuseINTERVENTION = Patient Education COMPARATIVE = No Action (In other Situations

Watchful Waiting/Placebo)

OUTCOME = Reduce Head Trauma Admissions

How to find Subject Headings and Synonyms

MeSH DatabasePearl Growing

◦Abstract view of PubMed results or other articles

Synonym generation◦Plurals◦Hyphenation◦Different Spellings (British)◦Narrow MeSH Headings

In a PubMed General Search

2000-2013:

Nosebleed – 2,432 Results

Nose Bleed – 2,427 results

Nosebleeds – 81 resultsEpistaxis (MeSH) –

2,413 results

Nosebleed

Nosebleeds

Nose Bleed

Epistaxis

Synonyms – Why you should use them

2476 Result

s

BUT

Generate Subject Headings and Synonyms

Will providing patient education reduce admissions for head trauma caused by parental abuse or neglect of children?

PICOChildren = This is a limiter so we do not necessarily need a

Subject term- It depends on the database.

Parental abuse = Child Abuse OR child neglect OR infant abuse OR child maltreatment

Patient Education = Patient Education as Topic OR parent education OR education of patient OR patient education

Head Trauma= Craniocerebral Trauma OR head injuries OR head trauma OR head injury OR shaken baby syndrome

PubMedRegular PubMed can be searched two ways:

1. Creating one large search.2. Doing smaller searches and adding them together

later.

(Child Abuse OR child neglect OR infant abuse OR child maltreatment) AND (Patient Education as Topic OR parent education OR education of patient OR patient education) AND (Craniocerebral Trauma OR head injuries OR head trauma OR head injury OR shaken baby syndrome)

PubMed Clinical Queries can also be searched both ways, but the second way tends to be a little more difficult.

Database Filter and Study Design

Before you search the PICO elements of your clinical question, it’s important to know:

What TYPE of question are you asking?

What is the best STUDY DESIGN to search for to find evidence to answer your clinical question?

What Kind Of Question is This?As a physician, you would like to know the

prevalence of headaches three and 12 months after mild, moderate or severe traumatic brain injury (TBI) in children ages 5 to 17. Which of the following clinical query filters would you want to apply to receive the most relevant and specific results?

Etiology TherapyPrognosisDiagnosisClinical prediction guidelines

The prognosis filter in Pubmed Clinical Queries automatically applies a specific search strategy (prognos*[Title/Abstract] OR (first[Title/Abstract] AND episode[Title/Abstract]) OR cohort[Title/Abstract]) to the search terms entered in order to retrieve article and study types that best address the concept of prognosis. In this case we are interested in following the course of the disease (TBI) to see what the outcome and frequency of headaches will be in these patients.

This is different from etiology/harm how?

Execute Your Search

Execute your search and select limiters if needed

Select the appropriate filter Advanced search building and limiters May depend on Database.

Analyze Your Search Results

This may be the point where you discover that you either have too many results or too few (or off topic results).

Too Many - Apply more or stricter limiters, look for higher levels of evidence or make search terms more specific.

Too Few (or off topic) – Remove any limiters, execute a general PubMed search, add search terms or generalize search terms (think drug class as opposed to specific drug)

Read the Abstracts/Review Articles

At this point you should be able to discover some basic information about the results by looking at the titles and abstracts. Look to see what type of results you have, and whether it suits your purpose.

Does this mean that you have found everything ever written because you searched in Pubmed?

Of course not.

In most cases, if you are doing searching that is clinical or research based you will not be doing yourself a disservice simply searching Pubmed. If, on the other hand, you are writing a literature review, systematic review, or meta-analysis you may also need to search grey literature, which includes conference proceedings, institutional publications, white papers, unpublished trials, and foreign language journals.

New ResourcesVisual DXWeb Of Knowledge

Visual DX Differential Builder-by lesion

Web of Knowledge

Journal Citation Reports will

organize journals by impact factor

Journal Citation Reports

Web of Scienc

e

Cited 18 times

Web of Science

Web of Science- Citation Map

Web of Science- Citation map – Backward and Forward

Perceptions can be Deceiving

References

Straus SE. Evidence-based medicine: How to practice and teach EBM. Edinburgh: Elsevier/Churchill Livingstone; 2005.

Guyatt G. JAMA's users' guides to the medical literature: A manual for evidence-based clinical practice. New York: McGraw-Hill Medical; 2008.

Moyer V. Weighing the evidence: PICO questions: What are they, and why bother? AAP Grand Rounds 2008 Jan; 19(1): 2.