EBD

Question 1

Why is randomisation important

Answer 1

Eliminates bias in allocation of treatment - selection and confounding bias - operators shouldnt be able to choose who gets what treatment as they end up allocating patients who have better prognosis in the experimental group, confounding variables is one that influences the results other than the one we are investigating

Question 2

What is blinding and name 4 groups of people who are blinded

Answer 2

Withholding information about the type of intervention the patient is receiving - difficult in dentistry due to surgery looking very diff in diff procedures n materials r diff
Patient blinding - avoids placebo effect and patient dropping out if they dont receive intervention
Operator - reduces performance bias as the operator may treat patients from experimental group diff - see equipose revision card
Investigator/researcher - selection bias (allocation) and reporting bias
Outcome assessor - reporting bias, can be done by randomisation of patients as long as intervention doesn’t leave an identifiable mark

Question 3

What factors affect external validity

Answer 3

Setting - nhs/private/hospitals
Operators - were they specialists
Patients - were they similar to general population in terms of social status etc
Duration - isit long enough to deduce prognosi eg longevity of a filling
Outcomes - how were they achieved n were all reported - objective and validated measurements
Randomly selecting patients - so ppl from all backgrounds n characteristics are involved
Sample size - greater = more generalizable

Question 4

Wats the power calculation

Answer 4

Calculates number of people needed in the study to have a significant effect size and to account for drop outs

Question 5

Wats calibration

Answer 5

Used when more than one assessor is involved in measurements in a cross sectional study (large) involves ensuring uniform measurement methods, can be enhanced with training as assessors accuracy is compared to a standard

Question 6

The kappa statistic ?

Answer 6

Measuring level of agreement between diff tests/ ppl
0= never agree, by chance if they do
1= always agree 100%

Question 7

Wat is a metasnalysis

Answer 7

The studies in a systematic review can be combined in their results to give an average of them all.
CI will be narrower due to larger pop size.

Question 8

Wats the criteria for homogenous studies

Answer 8

Same interventions
Same population
Similar risk of bias of study

Homogenous studies can be combined in a meta analysis

Question 9

What is a systematic review

Answer 9

A type of quantitative, secondary research which collects primary research which address a similar PICO question to it, and if possible may combine the results in a meta analysis

Question 10

What is a sample frame

Answer 10

A list of all the people within a population who can be sampled, sample of ppl for cross sectional study is drawn from sample frame

Question 11

What is bias

Answer 11

Systematic error is a study design or data collection that leads to one outcome appearing superior to another.

Question 12

Factors affecting internal validity

Answer 12

Bias:
Selection 
Performance
Attrition
Reporting 

No blinding
Calibration
Validation of questionaire

Question 13

What are the three questions in the CASP approach

Answer 13

Is the study valid
What are the results
Will the results help locally

Question 14

What are guidelines

Answer 14

Systematically developed statements that help people make clinical, policy related and system related decisions in their treatment of patients
They critically appraise research papers and present us with evidence based guidelines

Question 15

Mnemonic for guidelines

Answer 15

Tim
Gave
Sam
Red
Cake
Red 
Balloons
Pink ribbon

Question 16

Actual guideline procedure of formation

Answer 16

Topic of guideline decided
Group recruit in developnent of it
Scope of guideline
Research
Critically appraise 
Reccommendations formed - experience + res
Barriers identified
Published and disseminated to stakeholders
Reviewed after 3 yrs and changes made

Question 17

Purpose of guidelines

Answer 17

Improve patient care
Make recommendations that can be afforded by NHS and are cost effective
They help busy clinicians, patients etc make evidence informed decisiobs

Question 18

What is equipose

Answer 18

All people involved in the clinical trial should be in equipose - uncertain as to which treatment is better so that they give both an equal chance

Question 19

Whole pop strategy adv n disadv

Answer 19

Targets whole population
Eg water fluoridation, tooth brushing scheme, smoking ban
Adv = may produce large effects in some groups of ppl, tackles the underlying reason for disease