the WISDOM centre  

educational resources for the medical professions    

Home

Library

Clinical Governance

Caldicott

Informatics Course

 Clinical Governance Course

Basic Informatics Course

This is an extract from one of the units on the basic informatics course.

Medical Databases

Planning and executing a search of the Medline bibliographic database

by Debbie Hornby

Introduction

Database Coverage

Before starting a search of any bibliographic database it helps to know a little about the coverage of the database with respect to the quantity and type of records it contains, as this will affect how complex your search strategy will need to be.In general, the smaller the number of records in the database, the simpler your search.

Medline

Medline (MEDlars onLINE) produced by the National Library of Medicine (NLM) in the US is the largest of the bibliographic databases providing access to healthcare literature. It covers the fields of medicine, nursing, dentistry, veterinary medicine, the health care system, and the preclinical sciences and contains around 11 million bibliographic citations and abstracts from approximately 3,900 current biomedical journals dating back to 1966. Coverage is intended to be world-wide, with publications from 70 foreign countries as well as those from the United States but there remains a significant US bias.

The National Library of Medicine offers one of several ways of accessing Medline via the internet, through the PubMed website: http://www4.ncbi.nlm.nih.gov/PubMed/

Amongst the alternative ways of accessing Medline, is the service provided by OVID. Available to registered users of the University of Sheffield campus network, Medline is one of several biomedical resources included in the Biomed suite of databases.


Learning Aim and Objectives

Learning Aim: devise and run a search on Medline for a clinical problem to retrieve a reasonable number of papers

Learning Objectives - By the end of this keynote you should be able to


Devise a Search Plan

Note: The search plan comprises the following steps, the details of which should be defined in advance, if the search is to be effective.

    Accurately define your search question

    Identify the key concepts from the search question

    Devise a search strategy incorporating use of the MeSH thesaurus, Boolean operators and any other means of focusing your search, such as truncation, limits or filters as appropriate.


1. Define the search question

1.1 Be clear about the meaning of terms and phrases within your search topic

1.2 Summarise the question into a single sentence or statement.

At the risk of stating the obvious, the quality of the information you eventually retrieve from a search will depend significantly upon quality of the question it sets out to answer. Generally at the beginning of a search for information, your question probably exists in the form of a clinical scenario. To form an adequate search question however, it will almost certainly need to be more concisely defined; focusing the information in the scenario into a more succinct statement.


Scenario: (with thanks to Chris Roberts and Alan O’Rourke of the Wisdom Centre.)

You have a patient with rheumatoid arthritis, treated with anti-inflammatory drugs, who now needs anticoagulant therapy (you would prefer to use Warfarin). You need to know more about any increased risk of bleeding due to the interaction between the two therapies, so that you can adequately monitor the patient and prescribe the optimal drug regime.

Search question (statement): How can the increased risk of bleeding associated with the combined use of anticoagulant therapy, (preferably Warfarin), and anti-inflammatory drugs be kept to a minimum?

Back to Learning Aims and Objectives


2. Identify key concepts from the search question

To help structure your search, the structure, or anatomy of your statement needs to be identified, which should reveal its constituent concepts. The following model (Richardson 1995) is often used as a framework for the separation of clinical questions into their component concepts. Often referred to by the acronym P I O C, (or if more appropriate PICO) it comprises the following elements:

Problem/Condition/Population:

Who is the question about?

Who or what is the problem you are dealing with? Factors to consider may include age, sex, socio-economic status and ethnicity of the patient, and any specific technical terms or jargon associated with the problem.

Intervention/Exposure:

What is being done/is happening to the patient/population?

What interventions are you going to consider introducing into the ‘Problem’ situation? What are your options for action? You may even be attempting to compare the “if in doubt, do nowt” option with a more active alternative.

Outcomes:

How does the intervention affect the patient/population?

What do you hope to achieve from the intervention? For which you will need to know how the outcomes you are interested in might be measured, or evaluated, which is often more difficult and typically requires more thought than is necessary for identification of the problem or potential interventions.

Comparison:

What other outcomes could be considered – for example, following an alternative intervention?

This is an optional category used in situations where there may be two or more interventions, whose outcomes may be considered in terms of cost and/or effectivenesss.

Example:

Concept 1

(Patient/Problem)

Concept 2

(Intervention/Exposure)

Concept 3

(Outcome)

Concept 4

(Comparison-optional)

Patient (with rheumatoid arthritis) requiring anti-inflammatory drugs

Anti-inflammatory drugs & anticoagulant therapy/Warfarin

Increased risk of bleeding

Increased risk of bleeding with alternative drug/monitoring regime

Back to Learning Aims and Objectives