Paper 1 Flashcards
Describe evidence based dentistry
Use of best current available scientific evidence to make clinical decisions about the care of individual patients
6 uses of evidence based dentistry
Determining causes of dental disease
Diagnosis
Treatment
Prevention
Examine quality of care reported by patients
Economic tool
5 steps in the practice of evidence based dentistry
- Formulate clinical question
- Search for evidence
- Evaluate evidence through critical appraisal
- Apply results to clinical practice
- Implement and monitor the process
Give 5 sources of evidence
Colleagues
Textbooks
Dental journals
Internet
Conferences/courses
3 advantages of using a colleague as a source of evidence
Information is based on clinical experience
Easily accessible
Often more directly relevant
3 disadvantages of using a colleague as a source of evidence
May not be up to date
Conflicting opinions
May have limited or no experience
2 advantages of using a textbook as a source of evidence
Easily accessible
Written by experts in a field
2 disadvantages of using a textbook as a source of evidence
May not be up to date
Often not peer reviewed before publishing so rigour of evidence not assessed
4 advantages of using a dental journal as a source of evidence
Written by experts in a field
Most are peer reviewed
Up to date
Open access
3 disadvantages of using a dental journal as a source of evidence
Usually specialty based
Often niche topics
Expensive
2 advantages of using the internet as a source of evidence
Access to large datasbases
Free access to journals
3 disadvantages of using the internet as a source of evidence
Volume of material
Lack of quality control if not from peer-reviewed journals
Conflicting opinions
2 advantages of using conferences as a source of evidence
Evidence from experts
Up to date review of the current literature
2 disadvantages of using conferences as a source of evidence
Time consuming
Expensive to attend
4 important features of an information source
Valid
Relevant
Comprehensive
User friendly
What is strongest in the hierarchy of evidence
Meta-analysis and systematic reviews
What is weakest in the hierarchy of evidence
Case reports, opinion papers and letters
Why is research undertaken
To find out new knowledge
6 advantages of applying evidence based dentistry to clinical practice
Increase in confidence in diagnosis and treatment
Chance to offer more treatment options
Ensures effective and efficient treatment processes
Increases treatment acceptance
Improves satisfaction for patients
Increases referrals
5 disadvantages of applying evidence based dentistry to clinical practice
Information can overwhelm a clinician
Systematic reviews produce insufficient evidence
Clinician may not be trained in an appropriate procedure
Patients may not find intervention acceptable
Lack of resources can restrict choices available
3 examples of the use of EBD in contemporary clinical practice
NICE: National institute of health care and excellence
SIGN: Scottish intercollegiate guidelines network
Cochrane reviews
Define an observational study
Study that does not intervene in any way
Define an experimental study
Study in which the investigator deliberately intervenes to observe the effect of the intervention
Give 2 examples of observational studies
Cross sectional studies
Longitudinal studies
Define cross sectional studies
A descriptive observational study that provides a snapshot picture of a community at a point in time
4 advantages of cross sectional studies
Relatively simple
Data collected at around the same time
Little demanded of subjects
Does not have problems with prolonged follow-up
3 disadvantages of cross sectional studies
Lack time dimension therefore show association not causation
Subject to over-interpretation
Need a reasonable prevalence so not efficient for very rare conditions
Define longitudinal studies
An analytic observational study conducted over a length of time
2 examples of longitudinal studies
Case-control studies
Cohort studies
Define case-control studies
A retrospective longitudinal study where information is obtained from past records or interviews
2 advantages of case-controlled studies
Highly efficient, saves time
Lower costs
2 disadvantages of case-controlled studies
Medical records may be missing, incomplete or inaccurate
Interviews rely on accurate recall of information from many years ago
Define cohort studies
A prospective longitudinal study where data is collected forward from a given starting point
4 advantages of cohort studies
Multiple end-points can be assessed
Cause to effect time sequence clear
All measures of risk can be assessed
Exposure prior to outcome avoiding bias
3 disadvantages of cohort studies
Time consuming if the outcome does not occur for a long time after exposure may be costly, participants lost to follow-up
Large number subjects required to observe events if outcome not frequent
Subject to confounding
Define bias
Inaccuracy that is different in its size or direction in one of groups under study which can lead to systemic error in results
6 sources of bias
Selection
Allocation
Measurement
Recall
Observer
Publication
Define confounding
Confounding is distortion of the exposure - outcome association due to the association of another factor with both outcome and exposure
Give 4 examples of experimental studies
Randomised controlled trials
Quasi experiments
Parallel group studies
Matched design or cross-over studies
Define randomised controlled trials
Trials that randomly assigns participants to experimental or control group
Define Quasi experiments
Non-randomised trials
Define parallel group studies
Compare treatment between subjects
Define matched design or cross-over studies
Compare treatment within subjects
3 types of randomised controlled trials
Double blind RCT
Single blind RCT
Non blind RCT
Describe randomisation
Each participant equal chance of being placed in intervention or non-intervention group which can’t be predicted
What does randomisation achieve
Minimises some types of bias
Define fixed randomisation
Methods of allocation and randomisation are determined at the outset of the trial
4 methods of fixed randomisation
Random numbers
Simple randomisation
Block randomisation
Stratified randomisation
4 methods of random number randomisation
Computer generated
Dice
Random number table
Shuffled cards or envelopes
Describe simple randomisation
Each subject randomly allocated intervention or non-intervention as they enter the study
Describe block randomisation
Subjects are put in to blocks and are split equally and randomly between the treatment and the control group
Describe stratified randomisation
Subgroups created and confounders are equally distributed in the treatment and control groups and within each subgroup block randomisation takes place
Define adaptive randomisation
Used to balance characteristics and maintain similarity as a trial progresses
2 advantages of randomised controlled trials
Randomisation controls for confounding
Blinding controls for bias
5 disadvantages of randomised controlled trials
Complex with high requirements for organisation, time, cost and resources
Raises ethical questions
Large numbers needed to assess of small improvements
Long term follow up may be needed
Treatment may be obsolete by the time results available
Define single blinding
Only the investigator or the patient, knows whether they are in the intervention or non-intervention group
Define double blinding
Neither investigator or patients knows whether they are in the intervention or non-intervention group
Describe the aim of clinical trial CONSORT guidelines
Ensure that trials are correctly carried out and reported without bias
Describe a review of the literature
A systematic way of collecting and synthesising previous research
Define a standard review
An overview of selected published works written by a subject expert
2 disadvantages of a standard review
Can lack thoroughness
Risk of bias
Define a systematic review
A research method for identifying and critically appraising all relevant research and collecting and analysing the data
6 steps in conducting a systematic review
Defining the review question and eligibility criteria
Searching for studies
Selecting the studies
Data extraction
Data synthesis
Interpretation
3 advantages of systematic reviews
Thorough review of evidence base
Provide reliable and manageable summary data
Minimise bias
4 disadvantages of systematic reviews
Labour intensive
Need to be updated
Answer only a very focused question
Risk misinterpretation of evidence
3 features of a good systematic review
Quality
Thorough
Minimisation of bias
Define meta-analysis
Statistical combination of results from 2 or more homogenous studies to give a pooled effect estimate, presented on a forest plot
Why is meta-analysis carried out
Improves precision of results presented
Define a narrative review
Provides a brief synopsis of the appraised evidence in a literature review
Why is a narrative review carried out
Gives readers an understanding of how the included studies vary when there is a lack of homogeneity of available data for a full meta-analysis
Define GRADE
Overall judgement of all of the evidence and the quality of all the studies
6 features a GRADE outcome is based on
Risk of bias
Inconsistency across studies
Imprecision across studies
Indirectness across studies
Publication bias
Type of study
1 way of assessing risk of bias in randomised studies
Cochrane risk of bias tool
7 domains of Cochrane risk of bias tool for randomised trials
Random sequence generation
Allocation concealment
Blinding of participants and personnel
Blinding of outcome assessment
Incomplete outcome data
Selective reporting
Other bias
2 ways of assessing risk of bias in non-randomised studies
Newcastle-Ottawa scale * 0-9 scale
Down’s & Black
Describe Cochrane collaboration
Charitable organisation which produces high quality, relevant systematic reviews
3 goals of the Cochrane collaboration
Produce trusted evidence
Advocate for evidence
Inform health and care decisions
Define a population
The entire group of individuals
Define a sample
Selected group of individuals to represent the population in a research study
Define a variable
A characteristic or condition that can change or take on different values
Define independent variable
Characteristic/condition that is manipulated by the study
Define dependent variable
Characteristic/condition that is affected by the change in the value of the independent variable
Define 2 types of qualitative data
Nominal
Ordinal
Define nominal data
Data with no particular order. e.g, hair/eye colour
Define ordinal data
Data that can be catergorised with some order e.g. pain threshold (mild/moderate/severe)
Define 2 types of numerical (quantitative) data
Discrete
Continuous
Define discrete data
Data that consist of indivisible categories or fixed numbers. e.g., number of teeth, number of children in a family
Define continuous data
Data that is infinitely divisible into whatever units a researcher may choose e.g. weight in tonne, kilogram, gram, milligram
Describe a nominal scale
An unordered set of categories identified only by name, only determines whether two individuals are the same or different
Describe an ordinal scale
An ordered set of categories, can tell you the direction of difference between two individuals
Describe an interval scale
An ordered series of equal-sized categories which can identify the direction and magnitude of a difference, the 0 point is arbitrary
Describe a ratio scale
An ordered series of equal-sized categories which can identify the direction and magnitude of a difference, a value of 0 indicates none of the variable
Describe 3 measures of central tendency
Mean
Median
Mode
Define mean
Average or arithmetic mean of the data
Define median
Middle value when the scores are ranked in order
Define mode
Most frequently occurring value
Describe 3 measures of dispersion
Range
Interquartile range
Standard deviation
Define range
Highest to lowest values
Define interquartile range
Values that capture the middle 50% of the distribution, the upper and lower quartiles
Define standard deviation
Measure of spread of how individuals within the sample differ either side of the mean
Define standard error
Measure of how far the mean of the sample is likely to be from the actual target population mean
Define confidence interval
The probability that the mean for a population lies within a range of values, usually 95%
What does the width of a confidence interval suggest
The precision of the estimate and the wider the interval range, the less precise are the results
Describe normal distribution
Symmetrical, bell curve where mean = median = mode
In a normal distribution, what proportion of the sample will have a value within ± 1 standard deviation of the mean
68%
In a normal distribution, what proportion of the sample will have a value within ± 2 standard deviations of the mean
95%
In a normal distribution, what proportion of the sample will have a value within ± 3 standard deviations of the mean
99.7%
Describe what a low standard deviation indicates
Most values are close to the average
2 ways of identifying if data is non-normally distributed
Mean is very different from the median
Two standard deviations below the mean gives an impossible answer
3 examples of normally distributed data
Height, weight, haemoglobin
3 examples of skewed data
Salaries, number of marriages, GCSE’s passed
Define internal validity
The degree to which the results of the study are likely to be true, believable and free from bias
Define external validity
The degree that the results obtained can be generalised to the wider population
Define intention to treat analysis
All the study participant are included in the analyses as part of the groups to which they are randomised regardless of whether they completed the study or not
4 methods of controlling confounding factors
Restriction
Matching
Randomisation using stratified sampling
Statistic methods using regression analysis
What information does hypothesis testing provide
The possibility that chance alone might be responsible for any observed differences
Describe non-statistical hypothesis testing
Hypothesis test where a decision is made whether the hypothesis is true purely on the basis of evidence presented
Describe statistical hypothesis testing
Hypothesis test where statistical decisions about populations or processes is based on empirical data
Define the null hypothesis
That the difference in observations between two groups is purely due to chance
Describe the p-value
The probability that an observed difference could have occurred by chance
Describe what a value p<0.05 indicates
There is enough evidence to reject the null hypothesis in favour of the alternative hypothesis
Describe what a value p≥0.05 indicates
There is not enough evidence to reject null hypothesis
Describe clinical significance
The practical importance of a treatment effect and whether it has a real genuine, palpable, noticeable effect on daily life
Describe critical appraisal
Assessing the validity of the research and the statistical techniques employed in studies and generating clinically useful information from them
6 steps in critically appraising research
Are aims, clinical question and hypothesis clearly defined?
Is the study design appropriate to answer the clinical question?
Is the sample size determined?
Is the statistical analysis appropriate?
Are the conclusions based on the results?
Can I apply the results of this study to my patients?
Describe Type I error
False-positive, an investigator rejects a null hypothesis that is actually true
Describe Type II error
False-negative, investigator fails to reject a null hypothesis that is actually false
Define selection bias and how it can be avoided
Differences between baseline characteristics of the groups being compared
Avoided through randomisation
Define performance bias and how it can be avoided
Difference between groups in the care provided
Avoided through blinding
Define detection bias and how it can be avoided
Differences between groups in how outcomes are determined
Avoided through blinding
Define attrition bias and how it can be avoided
Differences between groups in withdrawals from the study
Avoided through intention-to-treat analysis
Describe parametric tests
Statistical tests that allow us to inferences about normally distributed data
Describe non-parametric tests
Tests that make fewer assumptions about the data and can be used to assess a more diverse range of data
Describe a one-tailed test
Test used when the alternative hypothesis specifies a direction
Describe a two-tailed test
Test used when there isn’t a direction of difference between groups for the alternative hypothesis
Describe independent samples
Two groups with different members
Describe paired samples
Same population at different times/locations (natural coupling) or two individuals with the same characteristics (matched coupling)
2 descriptive ways of presenting categorical data
Mode
Frequency
2 descriptive ways of presenting continuous non-normally distributed data
Median
Range
2 descriptive ways of presenting continuous normally distributed data
Mean
Standard deviation
Which 2 statistical tests are most appropriate when 2 groups are being studied and the data is categorical
Chi-square test
Fisher’s Exact test
Which statistical test is most appropriate when more than 2 groups are being studied and the data is categorical
McNemars test
Which statistical test is most appropriate when 2 groups are being studied and the data is continuous and normally distributed
Student t-test
Which statistical test is most appropriate when more than 2 groups are being studied and the data is continuous and normally distributed
ANOVA Test
Which statistical test is most appropriate when 2 groups are being studied and the data is continuous and non-normally distributed
Mann-Whiteny U Test
Which statistical test is most appropriate when more than 2 groups are being studied and the data is continuous and non-normally distributed
Kruskal Wallis Test
4 considerations when determining if research findings are applicable to your patients
Is the study population similar to mine
Are the settings similar
Are all the outcomes accounted for
Are my patients values and beliefs similar
Describe the relationship between power and research study sample size
The power of research study increases the larger the sample size, a large power (minimum 80%)
Describe how to formulate the research question for experimental and observational studies
PI/ECO: population, intervention or exposure, comparator, outcome
Describe minimisation in adaptive randomisation and when it is used
Investigators decide which factors/traits they wish to be equally distributed between the arms of a study
Useful in small studies and where the distribution of multiple factors is important
