Research Concepts Flashcards
Retrospective Studies
- Looking back at data already collected.
- Tend to miss data.
Prospective Studies.
Collect data as study goes on
Experimental Study
You are manipulating something to get information.
Observational Study
- Simply looking at data, not actually manipulating anything to get information.
- Cannot make claims about causation.
What are Case Control studies, what do they look at, and what measure of statistics is used.
- Compare case (has disease) with control (doesn’t have disease).
- Cannot tell prevalence as you are selecting cases so prevalence artificially high.
- Tend to be experimental studies.
- Looking at “What is the Frequency of exposure? in those with disease and those without disease.
- Use Odds Ratio.
Cohort Study
- Take cohort as it comes, you aren’t selecting particular cases with disease.
- Reflects true prevalence.
- Tend to be observational studies.
- Looking at “Is the outcome occuring?” in those exposed and those not exposed.
- Use Relative Risk.
Intention to Treat
- Analysis includes all participants, no matter if they completed trial or not.
- Use if wanting to know if clinician should use treatment.
Per protocol Design
- Analysis only includes participants who complete trial.
- Use if wanting to know mechanism of drug (i.e. need people who have completed treatment).
Descriptive statistics
- Describe the sample data.
- Reduce many numbers down –> fewer numbers.
- E.g. mean, median, standard deviation.
- Sample statistics.
- Use english letters (e.g. M, SD)
Inferential statistics
- Use sample statistics to predict population parameters.
- Use greek letters (e.g. mu, delta).
Type 1 error
- Incorrectly reject the null hypothesis –> accept working hypothesis.
- The probability is the alpha-level (usually 0.05).
- False positive (e.g tell a male he is pregnant or e.g use a drug which is ineffective or has adverse effects).
- Due to P value.
Type 2 error
- Incorrectly retain the null hypothesis –> working hypothesis is rejected.
- The probability is the beta level (often 0.2).
- False negative (e.g tell a pregnant women she isn’t pregnant or e.g. don’t use a drug which is effective for a particular disease).
- Due to Statistical power i.e. sample size.
Confidence intervals
- We can be 95% confident that the true parameter sits between these two values.
- If we repeated the procedure (maths calculation to get confidence interval) many times over again, the confidence interval would capture the true value. 95% of the time.
- Never the probability of the parameter having a certain value.
Relative risk (risk ratio)
ratio of the probability of outcome in one group to the probability of the outcome in another group
RR = (TE / (TE + TN)) / (CE / (CE + CN))
TE = treatment with events.
TN = treatment, non-events.
CE = control with events.
CN = control, non-events
What are the Canadian Task force Level of evidence?
I: At least 1 RCT with proper randomisation.
II.1: Well designed cohort of case-control study.
II.2: Time series comparison or dramatic results from uncontrolled studies.
III: Expert opinion.
What are the levels of evidence from Sackett?
I: Large RCTs with clear cut results.
II: Small RCTs with unclear results.
III: Cohort and case-control studies.
IV: Historical cohort or case-control studies.
V: Case series, studies with no controls.
What are the levels of evidence for Prognostic studies?
I: High quality prospective cohort study with adequate power or systemic review of these studies.
II: Lesser quality prospective cohort, retrospective cohort study, untreated controls from RCT, or systematic review of these studies.
III: Case-control study or systematic review of these studies.
IV: Case series.
V: Expert opinion, case report or clinical example, or evidence based on physiology, bench research or “first principles”.
What are the levels of evidence for therapeutic studies?
1A: Systematic review (with homogeneity) of RCTs.
1B: Individual RCT (with narrow confidence intervals).
1C: All or none study.
2A: Systematic review (with homogeneity) of cohort studies.
2B: Individual cohort study (including low quality RCT e.g. < 80% follow-up).
2C: “Outcomes” research, ecological studies.
3A: Systematic review (with homogeneity) of case-control studies.
3B: Individual case-control study.
4: Case series (and poor quality cohort or case-control study).
5: Expert opinion without explicit critical appraisal or based on physiology bench research or “first principles”.
What are the Grade Practice Recommendations?
-Grade A
- Strong recommendation.
- Level I evidence or consistent findings from multiple studies of levels II, III or IV.
- Clinicians should follow a strong recommendation unless a clear and compelling rationale for an alternative approach is present.
What are the Grade Practice Recommendations?
-Grade B
- Recommendation.
- Levels II, III or IV evidence and findings are generally consistent.
- Generally clinicians should follow a recommendation but should remain alert to new information and sensitive to patient preferences.
What are the Grade Practice Recommendations?
-Grade C
- Option.
- Levels II, III, or IV evidence, but findings are inconsistent.
- Clinicians should be flexible in their decision-making regarding appropriate practice, although they may set bounds on alternatives; patient preference should have a substantial influencing role.
What are the Grade Practice Recommendations?
-Grade D
- Option.
- Level V evidence: little or no systematic empirical evidence.
- Clinicians should consider all options in their decision making and be alert to new published evidence that clarifies the balance of benefit vs harm; patient preference should have a substantial influencing role.