AAAA Therapy: Using information from research on populations to help with clinical decisions about treatments Flashcards
4 stages of AAAA framework
- Assess
- Access
- Appraise
- Act
What is the aim of “assess”
Define the clinical question, study design and clinical components
What format are the clinical components in “assess” arranged in?
- PICO
> Population
> Intervention
> Comparator
> Outcomes
What is the aim of “access”
Finding studies (that are clinically relevant to the question)
2 stages of appraisal
- Quality of the conduct of the study (risk of bias)
- Interpretation of results
Basic questions of assessing quality of the conduct of the study in appraisal
- First
> Does the study address a research question relevant to the clinical problem?
> Did researchers use the study design most likely to provide a valid answer? - Then
> Was the study done well/is it trustworthy?
> If done well, what were the results?
Comparator vs control group
- Control = placebo or standard, pre-existing treatment
- Comparator = another treatment being assessed
Why have a control/comparator group?
- Both increase sample size
- Comparator gives patients a choice about which treatment
- Control allows evaluation of whether patients would have gotten better anyway with no treatment
Additional features of an RCT to mitigate against bias
- Allocation concealment
- Blinding
What is allocation concealment
The allocation sequence is concealed from participants and those recruiting them so it cannot influence who actually receives treatment
Which type of bias does allocation concealment prevent?
Selection bias
How is allocation concealment carried out
- Computer generated random sequence is performed distant to where trial is being conducted
- Treatment allocation should be concealed until administration (eg. sealed envelopes) (where possible)
Which types of bias does blinding remove?
- Measurement bias
- Performance bias
How does blinding remove measurement bias?
- Stops patients’ perception of pain, etc. being influenced by preconceived ideas about effectiveness of a treatment
- Stops patients giving socially desirable responses if they know they are in the intervention arm
- Stops preconceived ideas about how effective a treatment is influencing how observers measure subjective outcomes
How does blinding remove performance bias?
- Stops HCPs delivering treatment differently to different groups because of preconceived ideas about effectiveness of a treatment
- Stops participants behaving differently because they are disappointed in which group they were allocated
At which levels can blinding take place?
- Participants
- Clinicians delivering treatment
- Researchers measuring outcomes
- Statisticians undertaking analysis
Single vs double vs triple blind
- Single = blinding takes place on 1 level
- Double = blinding takes place on any 2 levels
- Triple = blinding takes place on any 3 levels
What is considered when interpreting results?
- Direction of effect
- Size of effect estimate
- Precision of effect estimate
- Statistical significance of effect estimate
- Clinical significance of effect
How is effectiveness measured?
- Risk of specific outcome in each group calculated then…
> Risk difference calculated
> Relative risk calculated
> Number needed to treat calculated
How is risk of specific outcome in each group calculated?
Number with specific outcome / Total number in group
How is risk difference calculated?
Risk of specific outcome in treatment group - Risk of specific outcome in comparator group
How is relative risk calculated?
Risk of specific outcome in treatment group / Risk of specific outcome in comparator group
How is number needed to treat calculated?
1 / Risk difference
What is number needed to treat?
Number of people needed to treat for 1 more person to have outcome
What different relative risk values mean
- RR = 1 –> no difference between groups
- RR > 1 –> more outcomes in treatment group vs comparator group
- RR < 1 –> less outcomes in treatment group vs comparator group
What different risk difference values mean
- RD = 0 –> no difference between groups
- RD > 0 –> more risk of outcome in treatment vs comparator group
- RD < 0 –> less risk of outcome in treatment vs comparator group
What is used to measure precision of results?
Confidence intervals
What is a confidence interval?
Estimation of the range within the true size of effect lies (eg. 95% CI for RD of +22 could be +18 to +26, meaning 95/100 repetitions of the trial would give RD value between +18 and +26)
What is the aim of “act”
Decide whether your practice should change due to the results of the evidence
