HaDPop Flashcards
How do you calculate total period fertility rate?
Sum of all age specific fertility rates
How is standardised mortality ratio calculated?
Observed deaths/expected x 100
How is incidence rate calculated>
Cases / (persons x years)
How are error factors calculated?
e^ 2 x (sq root of 1/observed event)
How are confidence intervals calculated?
Lower = value/error factor Upper = value x error factor
What are two types of selection bias?
Systematic error in how two groups were collected.
Allocation bias
Healthy worker effect
How is odds ratio calculated?
(exposed cases x unexposed controls) / (unexposed cases x exposed controls)
What is the null hypothesis when talking about an odds ratio?
1
What type of studies use odds ratios?
Case control studies
What is meant by the cause effect relationship in epidemiology?
Disease results from interplay between host, environment and agent.
A cause is an exposure or factor that increases the probability of disease.
Exposures don’t have to be necessary or sufficient to be a cause
What are the bradford hill criteria?
Strength of association Specificity of association Biology plausability Dose response Consistency of association Temporal sequence Reversibility Coherence of theory Analogy
What are 2 assumptions of epidemiology?
Disease doesn’t happen at random
Disease has causal and preventable factors that can be identified through systematic investigation
What is a clinical trial?
Any form of planned experiment that involves patients and plans to identify the most appropriate method of treatment for future patients with a given condition
What is meant by efficacy?
The ability of a healthcare intervention to improve the health of a particular group under specific conditions
What are 3 characteristics of a good clinical trial?
Fair
Reproducible
Controlled (ie uses controls)
What are some reasons for differences in otucomes between different patient groups in clinical trials?
Chance
Patient knows type of treatment
Assessor knows type of treatment
Treating clinician knows type of treatment
One treatment is more effective than another
Why do pre-defined outcomes need to be set for clinical trials?
To agree criteria for measurement and assessment of outcomes
To prevent repeated analyses
To give protocol for data collection
How do we minimise losses to follow up?
Make follow up practical, convenient and simple
Be honest about commitment
Avoid coercion or inducements
Maintain contact with participant
What is meant by clinical equipoise?
There’s reasonable uncertainty or genuine ignorance of the better treatment of intervention
What is meant be a trial being scientifically robust?
That a trial can be judged scientifically either by looking at what’s actually being researched or by looking at the safety and ethics of the trial
What is classed as ethical recruitment?
That there’s no inappropriate inclusion of communities unlikely to benefit from results, participants with higher risk of harm than benefit, participants likely to be excluded from analysis
And no inappropriate exclusion of people who differ from a homogenous group, people who are difficult to gain valid consent from
What factors are need for consent to be valid?
Knowledgeable informant Appropriate information Informed participant Competent decision maker Legitimate authoriser Voluntariness
What is a systematic review?
Overview of primary studies that used explicit and reproducible methods. Involves a systematic literature search then an appraisal and synthesis. May involve a meta-analysis
What is a meta analysis?
Quantitative synthesis of two or more primary studies that addressed the same hypothesis in the same way
What is the purpose of a meta analysis?
To facilitate the synthesis of a large number of study results
To systematically collate study results
TO reduce problems of interpretation due to variations in sampling
TO quantify effect sizes and their uncertainty as a pooled estimate
How are meta-analyses carried out?
Odds ratios are carried out for each study and these are then combined to give a pooled estimate odds ratio. Studies are then weighted according to their size and uncertainty of result so a higher error factor will get a higher rating. Results are then compiled into a forest plot.
Why is heterogeneity a problem with meta-analyses?
Ideally all studies would be identical for analysis but it’s inevitable that some will be different
How are forest plots interpreted?
The solid line down the center is the null hypothesis
THe diamond is the overall odds ratio with the edges of the diamond showing the overall confidence intervals.
Boxes are individual odds ratios with size relating to weight given
Lines from boxes show individual confidence intervals
Dotted line shows overall odds ratio in relation to all studies
What is the difference between the fixed effect and the random effects models for meta-analysis?
Fixed effects - assumes studies are estimating exactly the same effect so only shows 1 true effect line
Random effects assumes studies are estimating similar effect sizes so there’s mean true effect lines but also trial specific true effect lines which widens confidence intervals and makes weighting more even
How is publication bias tested for?
Use statistical test
Check meta-analysis protocol for method of study selection
Create funnel plot with results plotted against their size. Should be balanced
