Foraging Theory Applied to Medical Information Searching

Copyright © 2012 by Dr Mai Dwairy.

ISBN:          Softcover                                 978-1-4771-2803-9

                   Ebook                                      978-1-4771-2807-7

All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the copyright owner.

To order additional copies of this book, contact:

Xlibris Corporation

0800-891-366

www.xlibris.co.nz

[email protected]

700357

Contents

Abstract

Problem and research objectives

Method

Results

Conclusions

Acknowledgments

Glossary

1.    Delineation of the Problem

1.1    Information overload and Evidence-based practice

1.2    Information needs during consultations

1.2.1    Frequency of information needs

1.2.2    Types of information needs

1.3    Information seeking patterns

1.4    Constraints faced by GPs in accessing information

2.    Significance Of Problem

2.1    Impact of information on clinical decisions

2.2    GPs as first point of contact for most patients

3.    Conceptual Frameworks Available for Studying

the Problem

3.1    Conceptual diagram of GP’s information behaviour

3.2    Models of information behaviour

3.2.1    General Models

3.2.2    Models of information-seeking behaviour by health care practitioners

3.3    Optimal foraging theory

3.4    Information foraging

3.4.1    The prey concept in an information context

3.4.2    The patch concept in an information environment

3.4.3    Prey models in an information context

3.4.4    Information sent

3.5    The Evidence Based Medicine approach

3.5.1    General description and history

3.5.2    EBM and General Practice

3.5.3    Finding the answer—search skills

3.5.4    Barriers to EBM practice and implementations

3.5.5    Transferring research to practice

3.5.6    EBM training

3.5.7    EBM practice strategies

3.5.8    EBM and the Internet

3.5.9    The EBM debate—the adoption of best evidence into practice

3.5.10    Internet and Online sources

3.6    Potential strategies to improve the access to best available evidence

3.6.1    Introduction

3.6.2    Literature search service

3.6.3    Opinion leaders

3.6.4    Training for EBM skills

3.6.5    Computer-based clinical decision support systems

3.6.6    Evidence-based guidelines

4.    Research Design

4.1    Research objectives

4.1.1    Research questions

4.1.2    Conceptual model

4.1.3        Data requirements

5.    Methods

5.1    Procedure and sample

5.2    Data collection instruments

5.2.1    Background Questionnaire

Part 1: Obstacles to finding the evidence required for patient care (6 questions)

Part 2: Training in information searching (4 questions)

Part 3: Accessing the evidence (19 questions)

Part 4: Personal and practice details (8 questions).

Part 5: Workload, time pressure and satisfaction (5 questions)

5.2.2    Clinical Information Search Logbook

5.2.3    Background variables

5.3    Data analysis

5.3.1    General data treatment

5.3.2    Data analysis methods

Searching skills

Questionnaire components

Reliability and Factor analysis

5.3.3    Variables associated with the frequency of searching information sources in routine practice: High searchers (HS), Medium searchers (MS) and Low searchers (LS)

5.3.4    Variables associated with the use of electronic bibliographic databases: Electronic database searchers (EDS) and non-electronic database searchers (NEDS)

5.3.5    Information search time, success and efficiency

5.3.6    Influence of background and search variables on the types and order of information sources consulted

Variables associated with the type and order of information source consulted

HS versus LS and their different usage of information types in the logbooks

5.3.7    Influence of the complexity of an information need on search behaviour and outcome

6.    Results

6.1 Personal and practice backgrounds of GPs in the Wellington region

6.1.1    Personal details

6.1.2    Training

6.1.3    Practice settings

6.1.4    Workload and job satisfaction

6.2    Information searching background of GPs in the Wellington region

6.2.1    Training and perceived skills in information searching

6.2.2    Time allocated to information searching

6.2.3    Attributes of information sought

6.2.4    Use of and awareness of information sources

Overall pattern

Colleagues

Books

Journals

Electronic information sources

6.2.5    Accessing the evidence

6.2.6    Obstacles to information-searching

6.2.7    Profile of the last literature search that influenced clinical practice

6.2.8    Perceived best methods of facilitating information-searching

Methods currently used

GPs views on methods that should be used to seek the best available evidence

Perceived options for information-search facilitation

6.2.9    Attitude towards EBM

6.3    Background variables associated with information search frequency and use of information sources: High searchers (HS) versus low searchers (LS)

6.3.1    Personal and practice backgrounds

6.3.2    Training and perceived skills in information-searching

6.3.3    Time allocated to information-searching

6.3.4    Attributes of information sought

6.3.5    Use and awareness of information sources

Colleagues

Books

Electronic information sources

Use of electronic medical bibliographic databases

6.3.6    Accessing the evidence

6.3.7    Obstacles to information-searching

6.3.8    Perceived best methods to facilitate information-searching

6.3.9    Attitude towards EBM

6.4    Background variables associated with the use of electronic databases: electronic database searchers (EDS) versus non electronic database searchers (NEDS)

6.4.1    Personal and practice backgrounds

6.4.2    Training and perceived skills in information-searching

6.4.3    Time allocated to information-searching

6.4.4    Attributes of information sought

6.4.5    Use and awareness of information sources

Overall use of information source types

Colleagues

Books

Electronic information sources

Use of electronic medical bibliographic databases

6.4.6    Accessing the evidence

6.4.7    Obstacles to information-searching

Motivation to resolve the cost obstacle

6.4.8    Perceived methods currently used or recommended for general practice for seeking the best available evidence

6.4.9    Attitude towards EBM

6.5    Addressing an information need encountered during practice: Information foraging behavior and decision rules

6.5.1    GPs that completed a logbook compared to those that did not

6.5.2    Information needs that prompted information-searching

6.5.3    Constraints and expectations when starting the information-search

6.5.4    Number of information sources consulted

6.5.5    Types and order of information sources consulted

6.5.6    Access of information sources: location and time required

6.5.7    Criteria used for the selection of information sources

6.5.8    Previous search history and success rates within the information sources consulted

6.5.9    Source of awareness of the information sources consulted

6.5.10    Browsing methods used to select for relevant information items

6.5.11    Criteria used for selecting information items within sources

6.5.12    Criteria for stopping a search within an information source

6.5.13    Reasons for stopping an information-search altogether (end of last source consulted)

6.6    Addressing an information need encountered during practice: Information search time, search success and search efficiency

6.6.1    Search time, success and efficiency integrated over an entire search for information (from access of first source to exit from last source consulted)

6.6.2    Search time, success and efficiency in individual information sources and relationship with source access order

Overall search time, success and efficiency within an information source

Search time, success and efficiency in the first information source accessed

Search time, success and efficiency in the second information source accessed

Search time, success and efficiency in the third information source accessed

6.6.3    Time budget and organization of information searches according to search success in the sources consulted

6.6.4    Search time, success and efficiency associated with different types of information sources

Colleague

Book

Journal

Website

Database

Search engine

Differences in searching times in the 1st source by information source type

Search success versus information source type

6.6.5    Search success versus source search time

6.6.6    Search success versus previous success rate in the same source

6.6.7    Search time and search success versus attendance of training courses

6.6.8    GPs search levels (High, Medium, and Low Searchers) versus total search time

6.7    Background and search variables associated with the type and order of information sources consulted

6.7.1    Background and search variables associated with the type of information source selected as first source

Consulting a colleague as first information source

Consulting a book as first information source

Consulting a Website as first information source

Using a search engine as first information source

6.7.2    Background and search variables associated with the type of information source selected as second source

Consulting a colleague as second information source

Consulting a book as second information source

Consulting a Website as second information source

Using a search engine as second information source

6.7.3    Background and search variables associated with the type of information source selected as third source

Consulting a colleague as third information source

Consulting a Website as third information source

6.7.4    Relationship between the type of information source consulted and information searching frequency in routine practice: High Searchers versus Low Searchers

6.8    Influence of the complexity of an information need on information searching behavior and outcome

6.8.1    Complexity of information need versus type of first information source consulted

6.8.2    Complexity of information need versus total information search time

6.8.3    Complexity of information need versus number of sources consulted

6.8.4    Complexity of information need versus search success

6.8.5    Complexity of information needs addressed versus search frequency in routine practice

7.    Discussion

7.1    Willingness of GPs to participate in this study

7.2    Limitations of the study

7.3    Personal, practice and information-searching backgrounds of GPs in the Wellington region

Summary of main results

Discussion

Previous studies of information searching by GPs in New Zealand

Perceived obstacles to information-searching

Perceived best solutions for facilitating access to the best available evidence

7.4    Background variables associated with information search frequency and use of information sources: high searchers (HS) versus low searchers (LS)

Summary of main results

Discussion

7.5    Background variables associated with the use of electronic databases: electronic database searchers (EDS) versus non electronic database searchers (NEDS)

Summary of main results

Discussion

7.6    Addressing an information need encountered during practice: information foraging behaviour and decision rules

Summary of main results

Discussion

7.7    Addressing an information need encountered during practice: time budget for information-searching, search success and search efficiency

Summary of main results

Discussion

7.8    Background and search variables associated with the types and order of information sources consulted

Summary of main results

Discussion

7.9    Influence of the complexity of an information need on information foraging behavior

Summary of main results

Discussion

8.    Conclusions

Are GPs optimal foragers?

Implications for the use of evidence in general practice

9.    Appendices

Questionnaire

Part 1: Obstacles to finding the evidence required for patient care

Part 2: Training in information searching

Part 3: Accessing the evidence

Part 4: Personal and practice details

Part 5: Workload, time pressure and satisfaction

Appendix 2

Appendix 3

Information searcher background variables (questionnaire)

Clinical information needs variables (logbook)

Information search variables (logbooks)

Appendix 4

Clinical Information Search Log.

References

Abstract

Problem and research objectives

Workload and other constraints may prevent General Practitioners from accessing medical evidence relevant for clinical decisions. Strategies to alleviate such constraints would be significant, as GPs are the first port of call for most patients.

Optimal Foraging Theory, developed in behavioural ecology and recently applied to information foraging, provides a framework and methodology to derive the mostly unknown parameters needed to explore such strategies. This approach is here applied to information foraging by GPs, modelled as sequential steps (associated with measurable costs and benefits) taken to initiate, organize and terminate a search.

The study objectives were: 1) to determine how the background and practice routine of GPs relate to their information searching behaviour and success; 2) to analyze the current searching behaviour of GPs, constraints on search options or success, and strategies used to optimise searching; 3) to examine insights that Optimal Foraging Theory may provide into this behaviour.

Method

A sample of GPs from the Wellington region were asked to document steps and outcomes of an actual information search in a pre-formatted logbook, and fill in a questionnaire pertaining to their personal, practice and information-searching backgrounds.

Results

The 114 respondents spent on average 2.1 hours per week searching for information that was preferentially reliable, up-to date, quickly accessible and understandable. Books and colleagues were their dominant sources; and workloads, time required, access to bibliographic databases and budget available their main obstacles. Most had no training in information searching, and many reported poor electronic searching skills. Most recommended a free literature search service, and would increase training and information searching effort if subsidised. Frequent searching and recent use of bibliographic databases were associated with more recent graduation, better training and skills in information searching, greater use and awareness of electronic sources, and fewer obstacles to searching.

Most documented information needs (logbooks) were therapy related, of moderate complexity and high importance, and required answers within the same day. GPs had 16 minutes available and anticipated 17 minutes searching and 72% success per search, close to the actual parameters (18 min, 89%). GPs consulted up to three sources per search, mostly colleagues (all sources) and websites (third source), and selected for their availability, ease of access and relevance. GPs were mostly familiar with first but not subsequent sources, and finding enough information was the main reason for terminating a search. GPs spent 12.1 minutes in a source whether successful or not, and had to search 9.5 min per answer (including unsuccessful searches) in books, 15.9 min with colleagues, and 34.3 minutes in databases. Complex needs were associated with consulting journals and increased search time, but not increased number of sources.

Conclusions

By consulting the most profitable sources (books, colleagues), rapidly switching sources when unsuccessful, and frequent double checking, GPs seem close to an optimal trade-off between maximizing search success and information reliability, and minimizing searching time. Fully subsidised training in information searching and provision of a literature search service are two inferred avenues to facilitate access to medical evidence.

Acknowledgments

Foraging Theory Applied to Medical Information was funded by a University of Otago research grant for collecting the data from the GPs who participated in this study. GPs were interviewed to fill the semi-structured questionnaire by the thesis researcher, and the second part to fill a logbook about their searching habits. I am grateful to the Wellington region GPs who gave the time and effort to participate in this study. Further financial help from an Otago stipend scholarship was arranged to enable me to work full time in order to complete this study.

I am particularly grateful to Dr. Charles Tustin Director Research Higher Degree & Scholarships, for his full support. My supervisor, Professor Tony Dowel provided important support in supervising this thesis, and input in classifying the GPs’ information needs. Dr. Jean Claude Stahl provided valuable input to the optimal foraging aspects, and time spent in editing; his contribution and support is gratefully acknowledged.

I would also like to acknowledge the support of Dr. Chris Parkin, Senior Teaching Fellow, and of Prof. Marwan Dwairy’s help and advice on the use of SPSS for analysing the data. I am grateful to Dr. Collin Walker for his time and help with statistical analysis.

Glossary

Bibliography

A reference list, a complete or selective list of works compiled upon some common principle, such as authorship, subject, place of publication, or printer.

Complete search

A search for information in all sources consulted to address an information need (from access of first source to exit from last source).

EBM

Evidence-based Medicine

EDS

Electronic database searchers

Electronic databases

Electronic database: structured collection of abstracts or full texts of published studies, that can be searched and retrieved from a computer to guide clinical decisions (e.g. Medline).

High Searcher (HS) and Low Searcher (LS)

HS = High frequency user of information sources

LS = Low frequency user of information sources

Information item

Smallest retrievable item containing all or part of the information required to answer an information need (e.g. verbal communication, article, guideline, book, webpage)

Information search

Sequence of steps undertaken from initiation through termination of a search for external evidence to answer a specific information need.

Information search facilitator

Professionals or search engines able to assemble a collection of relevant information items from more than one information source (e.g. librarian, web search engines such as Google, bibliographic retrieval engines such as Medline).

Information source

Browsable collection of information items (e.g. colleague, journal, personal reprint collection, webpage/website, temporary collection assembled by librarian, www search engine or bibliographic retrieval engine/database)

Medical bibliographic database

A comprehensive collection of medicine-related data organized for convenient access in a computer, e.g. Medline, PubMed etc.

NEDS

Non-electronic database searchers

Pursued information item

Item retrieved and read.

Relevant information item

Up-to-date information item which contains keywords relating to the search question.

Search efficiency

Average search time needed per successful answer or, in other words, the average information-searching time required to arrive at one answer to the information need pursued, either within a specific information source (source search efficiency) or during an entire search to address the information need (overall search efficiency). Search efficiency refers to the combined information search time (successful or unsuccessful) across a group of GPs, divided by the combined number of successful answers found by the same group.

Search engine

A computer program that retrieves documents or files or data from a database or from a computer network (especially from the internet)

Searching skills

Searching skills terminology includes the following skills (variables):

Formulating an answerable question; formulating an optimal strategy to search for information; awareness of information sources; knowing how to use Boolean operators (AND/OR and others); familiarity with information sources, including electronic bibliographic databases; assessment of information quality; knowing when to stop a search after finding all relevant evidence.

Search success or Success rate

Percentage of successful searches (i.e. percentage of searches that yield an answer to the information need) among all searches within a source (source search success, e.g. within first source accessed or when consulting a colleague), or among all searches conducted to address the information need (overall search success, with success defined after exit from the last source consulted).

Search time

Time spent searching for information, either within a specific information source, or up to the end of a search to address an information need.

Source search

A search for information within a specific information source.

Successful search relating to General Practitioners’ (GPs’) information needs

Successful searches were counted on the basis of GP’s responses to a question regarding the reason why they stopped searching a source at a particular point (Q11.8, Q13.8, Q15.8). Of the response options available to the GPs, option (a) (found the right answer/enough information), was interpreted as being based on the GPs’ perception that the answer they found was the right answer and that their search was successful.

Time budget

Ways in which GPs allocate time to the different activities involved in an information search (e.g. time spent accessing the first source, time spent searching the first source,