 |
Evidence-Based
Practice |
Planning a Search With the Help of MOLES
(Methodology Oriented Locaters for Evidence Searching)
Planning a search involves these steps:
- In Row Two of the table below, insert the four elements of your question across.
- Right below these terms, in the first four columns of Row 3 insert as many terms as you can find that clearly designate the same concept as those in the first four columns of Row 2.
- Determine your question type, and then determine the appropriate Methodology Oriented Locaters for Evidence Searching (MOLES) for your question. These have been called “methodologic filters” (Sackett, Richardson, Rosenberg, & Haynes, 1997, p. 60). and “terms to identify studies”( Gibbs, 1991, p.175). See the MOLES table for these terms.
- Insert the appropriate MOLES in the last row on the right of Row 3.
- When you conduct your search, connect each set of the major terms (the ones that most clearly designate the column’s concept) vertically like this: (Term 1 OR Term 2 OR Term 3…). Column 1 will be Set 1; Column 2 will be Set 2; Column 3 will be Set 3, and Column 4 will be Set 4; and Column 5 will be Set 5.
- Then, and this is where the art comes in, proceed from the set that most clearly designates your topic. You might want to search this set alone first without intersecting it with other sets. Then connect this set with other sets with AND, for example: (Set 1 AND Set 2). The search term AND will limit your search. You’ll want the intersect that locates the best evidence that also designates your topic. Don’t forget to use the MOLES as a set.
Search Planning Sheet (to add information to the search planning sheet and print it, open this Word document.) (Connect terms vertically with OR into sets, and combine sets with AND to limit your search. I find it useful to construct the MOLES set first, then dip into it with other sets using AND.)
| Column 1: Client Type and Problem | Column 2: What You Might Do | Column 3: Alternate Course of Action | Column 4: Intended Result | Column: MOLES Appropriate to Question Type (Effectiveness, Prevention, Risk, Assessment, Description | |
|
Row 2: Determine Your Question Type, Then Insert Elements of Your Question in Spaces on Right
|
Leave Blank | ||||
| Row
3: Insert Key Terms from Above, Synonyms, or Terms from Thesaurus
or Controlled Language Vertically Connecting Terms by OR |
In This box Insert appropriate MOLES for Your Question Type. | ||||
| This table follows Sackett, D. L., Richardson, W. S., Rosenberg, W., & Haynes, R. B. (1997). Evidence-based Medicine: How to practice and teach EBM. New York: Churchill Livingstone. Adapted with permission. | |||||
MOLES for Each Question Type (You may mark MOLES and copy them right into the Planning Sheet above after opening the Planning Sheet.)
| Effectiveness Questions | Prevention Questions | Risk/Prognosis Questions | Assessment Questions | Description Questions (With Qualitative Studies a Subset) | Syntheses of Studies (These work primarily with Effectiveness and Prevention Questions but may work with others.) |
| Random*
OR Controlled Clinical trial* OR Control group* OR Evaluation stud* OR Study design OR Statistical* Significan* OR Double-blind OR Placebo OR RCT n.b. if too many hits for random* insert random* assign* OR random* control* trial* |
(Random*
OR Controlled Clinical trial* OR Control Group* OR Evaluation Stud* OR Study Design OR Statistical* Significan* OR Double Blind OR Placebo) OR RCT) AND Prevent* |
(Risk
Assessment OR Predictive Validity OR Predictive Value OR Receiver Operat* OR ROC OR Sensitivity OR Specificity OR False positive* OR False negative* OR Prognos*) AND Predict* |
(inter-rater
OR Inter-observer OR True positive* OR Specificity OR False Positive* OR False negative* OR Sensitivity OR predict* OR Receiver operat* OR ROC) AND (assess* OR diagnos*) |
(Random*
Select* OR Survey OR Representative Sample) AND (Client satisfaction OR Patient satisfaction OR Needs assessment) n.b. for Qualitative Studies use these: Qualitative Study OR Qualitative Analysis OR Content Analysis OR In Depth Interview* OR In-Depth Interview* OR Participant Observation OR Focus Group* |
meta-anal*
OR meta anal* OR metaanal* OR Systematic Review* OR synthesis of studies OR Study Synthesis OR Metasynthesis OR Meta-synthesis meta synthesis |
| Suggestions: The MOLES appear in rough descending order of their utility; so you might start with those at the top and, if you find few references, add more MOLEs downwardly with the OR command to enlarge the MOLES set. Also, some of the columns at their bottom include another set connected by the AND command. These additional terms generally mark the topic for their respective question type. MOLES reflect my search experience, ideas from Gibbs (1991), and ideas in McKibbon, Eady and Marks’s (1999). PDQ Evidence-based principles and practice Hamilton: B.C. Decker, also suggestions from Aron Shlonsky, School of Social Work, University of Toronto, Canada. An Stanley McCracken of the University of Chicago. An earlier version of this list was published in Gibbs, L. (2003). Evidence-based practice for the helping professions. Pacific Grove, CA: Brooks/Cole Thomson Learning | |||||
*This symbol may vary with the database. It means all words with the same root (e.g. randomly, randomized, random). In MEDLINE this term is a $.
General
Suggestions for Searching
Always remember that the point of a search is to get an accurate answer.
One cannot get an accurate answer if one looks only for evidence that
confirms a given position. Resolve to look just as diligently for disconfirming
evidence as for confirming evidence.
As you examine what you have found, you might address the following questions:
- Did your search strategy convey the master concept and related concepts accurately? If not, revise your search terms.
- Did your search logic limit your search so severely that you net few useful sources? If so, try searching each set of concepts before connecting them by AND.
- Did your strategy miss a useful database? If this might be so, try looking at our web site’s database grid again, or try to gain access to databases that are not free on our web site.
- Do records retrieved apply to your clients by client type, action you might take, alternate course of action, and intended result? If not, specify these variables more accurately.
- Have you addressed all of these questions and still found nothing of use to guide your practice? Then sufficient evidence may simply not exist to guide you at this time.
- You can obtain full text of documents through a university’s inter-library loan department, or through fax from some databases.
Some Problems and Ways to Counter Them
| Problem | Suggested Countermeasure |
| Too Few References? | Select search terms that more accurately delineate the topic. |
| Select a better database. | |
| Examine the search history to see that terms were not combined with AND too soon. | |
| Too Many References? | Combine the best search terms with AND. |
| Apply MOLES combined with AND. | |
| Pick the one search term most specific to the topic. | |
| Use meta-analysis search terms. | |
| Irrelevant References? | Select search terms that more accurately delineate the topic. |
| Consult the database’s thesaurus | |
| Not Satisfied? | Consult a reference librarian who understands MOLES and evidence-based databases. |
| Still Not Satisfied? | Admit to yourself, to your colleagues, and to your clients that empirical evidence is slim. |