INTRODUCTION – WHAT IS EVIDENCE BASED PRACTICE?
The practice of dentistry is becoming more complex and challenging.
Developments in dental materials and techniques, changing socio-demographic patterns, increasingly knowledgeable health care con-sumers, and the information ‘explosion’ all place greater demands on clinical decision making. As health care practitioners, it is important to offer the best possible care for patients. However, few decisions made in the health services are made as a result of good evidence.
Evidence based practice (EBP) aims to encourage the practitioner to look for, and make sense of, the available research evidence in order to apply it to everyday clinical problems. This presents a challenge to the practitioner as within dentistry alone there are around 500
journals publishing over 43 000 research articles a year. Given that a large proportion of these papers are of limited relevance to everyday practice and often of poor quality, how do you know which of these articles you should read to inform your practice and which you can disregard? You need to be able to identify articles that are both of a high quality and relevant to your clinical practice.
EBP has been defined as ‘the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients’1. It aims to inform, not replace clinical judgement and experience, by integrating the best evidence with clinical expertise and patient preferences. EBP can be broken down into five key stages, as illustrated in Fig. 8.1. The key stages require the development of skills that encourage the process of life-long learning, allowing the practitioner to identify and react appropriately to emerging information. This chapter will explore, in brief, some of the skills involved in the EBP process.
161
POOC08 02/18/2005 04:37PM Page 162
162 Chapter 8
5. Evaluation of practice
1. Identification of a clinically
4. Implementation of findings
relevant question
3. Appraisal of research
2. Searching for relevant
evidence
research evidence
Fig. 8.1 The key stages involved in evidence based practice.
IDENTIFYING AND DEFINING RELEVANT QUESTIONS
Everyday clinical practice can present a vast array of dilemmas regarding the prevention, diagnosis, prognosis or treatment of oral diseases and disorders. EBP encourages the recognition of uncertainty when making clinical decisions. The first step in the EBP
process requires the area of clinical uncertainty to be translated into a focused, clinically relevant question.
For example, an adult patient presents with a cavity in a posterior tooth. You explain the different restoration options available to them.
They stress that the appearance of the restoration is important to them and that they have heard that ceramic inlays can produce a good colour match with the surrounding tooth. However, they have concerns over the longevity of such restorations in comparison to amalgam and want to know exactly how long they might expect a ceramic inlay to last. You decide to examine the research literature to try and find an answer to their question.
Before looking for relevant information to answer your question, it is a good idea to define your question so that you know exactly what you’re looking for. One way to do this is to break it down into sections. Clinical research questions can usually be broken down into the following (PICO) components:
POOC08 02/18/2005 04:37PM Page 163
Evidence based practice 163
• Population – who are the participants within the study?
• Intervention/Exposure – what are the interventions, risk factors or exposures of interest?
• Comparison – what control comparisons do you want to make?
• Outcomes – what outcome measures are of interest to you and your patient?
Using the previous example, the question you want to answer is
‘how effective are ceramic inlays?’ Applying the PICO format allows us to define the question further. The patient group (P) of interest can be defined as adults requiring a restoration to any posterior tooth. The restoration/intervention (I) of interest to the patient is ceramic inlays.
The patient wants to know how long ceramic inlays are likely to last in comparison (C) to amalgam. The outcomes (O) of importance to both you and your patient include the longevity of the restoration and aesthetics.
Structuring the question in this way can help clarify the type of question you are asking and how to identify the relevant information to answer your question.
IDENTIFYING EVIDENCE
There are a number of sources of information a health professional can turn to in order to find the answer to a question. For example, they may draw upon their own clinical judgement or past experience, they may turn to colleagues for advice, read relevant textbooks or journal articles, search electronic bibliographic databases (e.g. MEDLINE or EMBASE) or secondary sources of information (e.g. The Cochrane Library, The Journal of Evidence Based Dental Practice, Evidence-based Dentistry). All of the available sources of information have advantages and disadvantages (Table 8.1).
The ability to carry out quick and reliable searches to identify relevant research evidence is one of the key skills required when practising evidence based dentistry. Before starting to search for research evidence, it is necessary to consider what type of evidence would be most appropriate to answer your research question.
Different clinical research questions require evaluation through different study designs. The ceramic inlay example deals with the effectiveness of a treatment option and would be best answered by a randomised controlled trial (RCT) or, ideally, a systematic review of RCTs. However, it must be noted that although RCTs and systematic
POOC08 02/18/2005 04:37PM Page 164
164 Chapter 8
Table 8.1 Advantages and disadvantages of different sources of information.
Source
Advantage
Disadvantage
Colleague
Quick response
May be biased
May be out of date
Text book
Easy to access
Quickly out of date
Journal article
Full text of recent research
Difficult to access/identify
relevant information
Bibliographic
Available online
Can be difficult to search
databases
Access to large amounts
Databases are not
(e.g. MEDLINE)
of information
comprehensive in their
Sometimes links to full
coverage of published
text online
journals
Secondary sources
Provide summaries of
Can be difficult to search
(e.g. Systematic
relevant research evidence
Do not cover all topics
Reviews on The
Aim to be unbiased
Cochrane Library)
and up to date
reviews of RCTs may well be the ‘gold standard’ upon which to base decisions on the effectiveness of interventions, they are not necessarily appropriate, or ethical, to answer all questions. For example, you may be interested in the prognosis of endodontic treatment in patients with diabetes mellitus. An RCT would obviously not be helpful in answering this question. For questions regarding prognosis such as this, cohort studies would be more appropriate. Table 8.2 illustrates the types of studies designs most suitable for different types of research questions arising in endodontics. The most appropriate source of information will depend upon the type of study design being sought.
For questions regarding the effectiveness of healthcare interventions, the best place to start looking for information is The Cochrane Library: http://www.cochrane.org/reviews/clibintro.htm The Cochrane Library is recognised as the best single source of systematic reviews and controlled clinical trials, covering all aspects of healthcare. It is available through many university libraries and postgraduate centres. It is also freely available to all resid-ents in many countries including the British Isles (a full list of countries with free access is available at
http://www.update-software.com/cochrane/provisions.htm).
For questions not focusing on the effectiveness of interventions, other study designs will need to be sought, for which MEDLINE can
POOC08 02/18/2005 04:37PM Page 165
Evidence based practice 165
Table 8.2 Study designs and the types of questions they address.
Definition of study design
Used for
Experimental studies
Randomised controlled trial: parallel
Evaluating the effectiveness of an
group design – a group of participants (or intervention
other unit of analysis, e.g. teeth) is randomised Randomised controlled trial comparing the
into different treatment groups. These groups
longevity and aesthetics of two restorative are followed up for the outcomes of interest
materials
Randomised controlled trial: split-
Controlled trial comparing the longevity of mouth design – each patient is his/her
simply repairing or replacing restorations
own control. A pair of similar teeth, or groups of teeth (quadrants) may be selected and
randomly allocated to different treatment
groups
Non-randomised controlled trial –
allocation of participants under the control
of the investigator, but the method falls short of genuine randomisation
Observational studies
Cohort – a longitudinal study, identifying Measuring the incidence of a disease;
groups of participants according to their
looking at the causes of disease; determining
exposure/intervention status. Groups are
prognosis
followed forward in time to measure the
Cohort study looking at the progress of
development of different outcomes
carious lesions over time and relating this to external factors such as smoking or diet
Case–control – involves identifying two
Looking at the causes of disease;
groups; those who have the outcome of
identification of risk factors; suitable for
interest (cases) and those who have not
examining rare diseases
(controls). The investigator then looks back in Case–control study comparing patients with
time to see who had the exposure/intervention
endodontic infection (cases) with patients
of interest
without (controls), looking at associated risk factors
Cross-sectional survey – the observation Measuring the prevalence of a disease;
of a defined population at a single point in
examining the association
time or time interval. The status of an
Cross-sectional survey of caries in young
individual with respect to the presence or
adults in the North West of England
absence of both exposure/intervention and
outcome are determined at the same time
be a valuable source. MEDLINE is the US National Library of Medicine’s (NLM) premier bibliographic database. It can be accessed free of charge via http://www.ncbi.nlm.gov/PubMed but is also available through other database providers such as Ovid and SilverPlatter.
Searching electronic databases is a complex process, too detailed to cover in this chapter, particularly as databases and service providers
POOC08 02/18/2005 04:37PM Page 166
166 Chapter 8
all vary in terms of their specific searching ‘rules’2. However, there are certain elements that should be considered when developing a search strategy for an electronic database which we will cover briefly.
Controlled vocabulary and free text-terms
Many people when searching electronic databases tend to rely on free-text searching; simply typing a phrase, such as ‘ceramic inlays’, into the search box. The problem that arises when relying on free-text searching is that it depends on finding the exact match of text within the title, abstract (if present) or indexing field of the record. If the author of a study has described the restoration in any way other than
‘ceramic inlays’, for example ‘porcelain inlays’, the record will not be retrieved. To help overcome this, many databases use an indexing system for coding records entered onto the database. For example, MEDLINE and The Cochrane Library both use MeSH (Medical Subject Headings). All references added onto the database are assigned appropriate MeSH by indexers at the NLM. An example of some dental MeSH terms (section of MeSH Tree) are shown below. The broadest term ‘Dentistry, operative’ at the top and more specific headings indented below:
• Dentistry, operative
• Crown lengthening
• Dental cavity lining
• Dental restoration failure
• Dental restoration, permanent
—
Crowns
° Post and core technique
—
Inlays
—
Marginal adaptation (dentistry)
• Dental restoration, temporary
—
Crowns
° Post and core technique
It is recommended that a combination of appropriate MeSH and free-text terms be used when searching electronic databases to improve the identification of all relevant research articles.
Search operators and truncation symbols
Most databases allow the combining of search terms using the Boolean operators AND, OR and NOT. The operator AND is used
POOC08 02/18/2005 04:37PM Page 167
Evidence based practice 167
when the research article must contain both search terms, for example
‘ceramic’ AND ‘inlay’ (this will reduce the number of articles retrieved). The operator OR is used when either search term is acceptable, for example ‘porcelain’ OR ‘ceramic’ (this will broaden the number of articles retrieved). By using NOT between search terms will require the article to contain the first search term but not the second, for example ‘restoration’ NOT ‘crown’. In addition, some databases use proximity operators such as NEAR, NEXT or ADJACENT, which can also be helpful in focusing a search.
Truncation symbols can be used to broaden a search by allowing any ending to a given free-text word. For example, the truncation symbol in The Cochrane Library and PubMed is an asterisk. If placed at the end of ‘restorat*’ this will allow the retrieval of records containing the words restorative, restoration, restorations, restorating, etc. . . .
Truncation symbols vary according to databases. When choosing to search a particular database, ensure you are familiar with both the search operators and the truncation symbols available.
If we return to the ceramic inlays example, and acknowledge that the most appropriate database to search initially would be The Cochrane Library, we can explore how search results can vary depending upon the approach we take. Table 8.3 shows the results obtained when different free-text terms, MeSH terms, search operators and truncation symbols are used.
The results obtained through searching The Cochrane Library (Table 8.3) identified one relevant systematic review, entitled
‘Ceramic inlays for restoring posterior teeth’3. As systematic reviews provide a comprehensive, objective overview of research in a given area, this would be the obvious article to read first. If such a systematic review did not exist, individual studies of an appropriate study design (in this case RCTs) would have to be examined.
APPRAISAL OF RESEARCH LITERATURE
Let’s assume your searching of the literature has identified several relevant primary studies. All the studies use an appropriate study design for answering your research question. Consideration still needs to be given to the fact that within any particular design there is huge variability between studies with regard to how well they are conducted. An important issue to consider at this stage is the study’s validity, in particular internal validity.
POOC08 02/18/2005 04:37PM Page 168
168 Chapter 8
Table 8.3 Example of results obtained when different free-text terms, MeSH terms, search operators and truncation symbols are used when searching an electronic database ( The Cochrane Library, Issue 4, 2003).
Number of
Number of
systematic
controlled
Term
reviews
clinical trials
Ceramic inlays
1
25
Ceramic inlay
1
7
Ceramic inlay*
1
29
Porcelain inlays
0
4
Porcelain inlay
1
0
Porcelain inlay*
1
4
Ceramic* AND inlay*
1
46
CERAMIC (MeSH)
2
95
INLAYS (MeSH)
1
47
Combining free-text terms and MeSH can help increase the sensitivity and specificity of your search
(CERAMIC (MeSH)
1
48
OR ceramic* or porcelain)
AND (INLAYS (MeSH)
OR inlay*)
Internal validity refers to the degree to which the results of a study are likely to approximate to the ‘truth’ for the circumstances being studied4. Has the study been conducted in such a way that systematic error (bias) has been minimised? External validity refers to the degree to which the effects observed in the study are applicable to the outside world; how generalisable are the results to other circumstances5? If internal validity does not exist, there is little point in considering a study’s external validity. There are numerous ways in which bias can affect the internal validity of a study. Some of the main biases6 are described in Table 8.4.
There is a plethora of critical appraisal checklists and resources available. Some guidance on the types of questions to ask7,8 when reading research articles is presented below:
• What is the research question?
• Is the study design appropriate?
• How well was the study conducted? Consider issues of:
—
Internal validity
—
Sample size
—
Validity of outcome measures
—
Duration of follow-up
POOC08 02/18/2005 04:37PM Page 169
Evidence based practice 169
Table 8.4 Potential biases affecting the internal validity of a study.
Bias
Definition
Method of
overcoming bias
Selection bias
Systematic differences between comparison
Randomisation
groups in prognosis or responsiveness to
treatment
Performance bias
Systematic differences in care provided
Blinding of participants
to participants in a study, apart from the
and investigators to
intervention being evaluated
treatment allocation
Measurement bias
Systematic differences in how outcomes
Blind outcome assessment
are ascertained from the groups under
comparison
Attrition bias
Systematic differences between comparison
Intention to treat analysis
groups in withdrawals or exclusions of
participants from the results of a study
• What are the results?
—
Do the numbers add up?
—
Are all participants accounted for?
—
Was the statistical significance assessed?
—
Were the statistical techniques appropriate?
• What are the implications for practice?
It is acknowledged that this chapter does not cover the critical appraisal of all types of research articles published in the dental literature. For further guidance on how to appraise systematic reviews, clinical guidelines, qualitative research, economic evaluations in addition to the study designs already discussed, the following key resources are recommended:
• Critical Appraisal Skills Programme ( CASP) http://www.phru.nhs.uk/~casp/casp.htm and the CASP International Network (a collaboration of individuals and organisations across the world who are interested in helping people develop skills in making sense of scientific evidence – available in Chinese) http://www.caspinternational.org.uk/
• The How to read a paper series by Trisha Greenhalgh, available online via the British Medical Journal website (http://www.bmj.com). The series is also available as a book:
Greenhalgh T. How to Read a Paper. The Basics of Evidence Based Medicine. London, BMJ Publishing Group, 1997.
POOC08 02/18/2005 04:37PM Page 170
170 Chapter 8
• The Users’ guide to the medical literature initially published in the Journal of the American Medical Association and available online at http://www.cche.net/usersguides/main.asp. This series has also been updated and published as a book:
Guyatt G, Rennie D. Users’ Guide to the Medical Literature. A Manual for Evidence-based Clinical Practice. Chicago, American Medical Association, 2002.
• Crombie I. The Pocket Guide to Critical Appraisal: A Handbook for Health Care Professionals. London, BMJ Publishing Group, 1996.
• The Appraisal of Guidelines Research and Evaluation ( AGREE) instrument provides a framework for evaluating the quality of clinical practice guidelines http://www.agreecollaboration.org/
Often a first attempt at critical appraisal can feel slow and cumber-some, but with practice it becomes rapid and almost automatic.
By developing skills associated with critical appraisal healthcare workers will become more adept at managing the ever increasing information overload. It is imperative that any healthcare worker wishing to practise in an evidence based manner develops the ability to quickly identify and interpret high quality, valid and clinically relevant research articles to inform their practice.
IMPLEMENTATION OF RESEARCH EVIDENCE AND EVALUATION
OF ITS APPLICATION
Not all valid research articles will be relevant to your clinical practice.
Even if a study focuses on a particular intervention or outcome of interest, consideration needs to be given to the external validity of the piece of research. Are the participants included in the research article similar enough to the patients you deal with, or are they so different that the results don’t apply? Even if you feel the study is relevant to your clinical practice, implementing the findings is not always easy.
There are many barriers that may prevent the integration of research findings into your everyday practice. Time, financial resources and the organisation of the practice in which you work may all prevent the adoption of research evidence9. Similarly, opinion leaders within the practice, current beliefs and attitudes regarding treatment options and, increasingly, patients’ perceptions about what they consider to be appropriate care all can play their part in preventing best evidence from changing clinical practice.
POOC08 02/18/2005 04:37PM Page 171
Evidence based practice 171
A variety of implementation strategies have been used to try and improve the uptake of research findings, including educational outreach visits, manual or computerised reminders, interactive educational sessions and audit and feedback. The effectiveness of an implementation strategy appears to be dependent upon individual circumstances, with multifaceted interventions targeting barriers to change being more likely to be effective than a single intervention5.
There appears to be no single implementation model that can be applied across all settings. Before attempting to change clinical practice, local barriers need to be identified and strategies to overcome them explored.
CONCLUSION
In order for healthcare practitioners to offer the best possible care for their patients they need to practise in an evidence based manner. This involves the development of skills that will enable the practitioner to find and assess emerging information relevant to his particular clinical practice. These skills include forming appropriate questions, searching the literature, appraisal of research papers and the implementation of the findings into everyday practice. This is an ongoing process which will ensure that patients receive the most up-to-date effective care.
REFERENCES
1.
Sackett D., Strau S., Richardson W., Rosenberg W. and Haynes R.
Evidence-based Medicine, 2nd edn. Toronto, Churchill Livingstone, 2000.
2.
Bickley S. and Harrison J. How to . . . find the evidence. J Orthod, 2003; 30: 72–78.
3.
Hayashi M. and Yeung C. Ceramic inlays for restoring posterior teeth. In: The Cochrane Library, Issue 3, 2004.
4.
Juni P., Altman D. and Egger M. Assessing the quality of randomised controlled trials . In: Egger, M., Davey Smith, G. and Altman, D. (eds) Systematic Reviews in Health Care. Meta-analysis in Context. London, BMJ
Publishing Group, 2001, pp. 87–108.
5.
NHS Centre for Reviews and Dissemination. Undertaking Systematic Reviews of Research on Effectiveness: CRD’s Guidance for Carrying out or Commissioning Reviews, 2nd edn. NHS Centre for Reviews and Dissemination, University of York, 2001.
POOC08 02/18/2005 04:37PM Page 172
172 Chapter 8
6.
Clarke M. and Oxman A. Cochrane Reviewers’ Handbook 4.2.1 [updated March 2004]. In: The Cochrane Library, Issue 4, 2004, pp. 49–60.
7.
Crombie I. The Pocket Guide to Critical Appraisal: A Handbook for Health Care Professionals. London, BMJ Publishing Group, 1996.
8.
Greenhalgh T. How to read a paper: getting your bearings (deciding what a paper is about). BMJ, 1997; 315: 243–6.
9.
Bero L., Grilli R., Grimshaw J., Harvey, E., Oxman, AD and Thomson MA . Closing the gap between research and practice: an overview of systematic reviews of interventions to promote the implementation of research findings. The Cochrane Effective Practice and Organization of Care Review Group. BMJ, 1998; 317: 465–8.
POOZ01-Index 02/18/2005 04:37PM Page 173