RESS 3 Lectures Flashcards
What is a ‘research’ type study in the context of clinically relevant studies?
Generates new knowledge where there is no or limited research evidence available and which has the potential to be generalisable or transferable.
What is a ‘audit’ type study in the context of clinically relevant studies?
a quality improvement process that seeks to improve patient care and outcomes through systematic review of care against explicit criteria and the implementation of change.
What is a ‘service evaluation’ type study in the context of clinically relevant studies?
evaluates a proposed service or current practice (e.g. acceptability and uptake) with the intention of generating information to inform local decision-making
What is a ‘audit-cum-service evaluation’ type study in the context of clinically relevant studies?
Combined audit and service evaluation. Establishes to entent to which clinical practice is achieveing a particular standard as well as a to estabilish what factors might be associated with those contexts , those patients or those practitioners who succeed in achieving this standard; and those who do not)
What is NICE?
The National Institute for Health & Care Excellence (NICE) is the independent organisation responsible for providing national guidance and advice to improve health and social care
What are the roles of NICE?
1) To reduce variation in the availability and quality of treatments and care
2) To help resolve uncertainty about which medicines and treatments work best and which are the best value for money for the NHS
3) To set national guidelines on how people with certain conditions should be treated
What is the mneumonic used to define the search terms used in research literature searches?
PECOS Patient/participant/people Exposure/event/experimental Intervention Comparison (if appropriate) Outcome Study design (if appropriate)
What are the 3 different options for locating studies with specific designs?
1) Free text searching using appropriate search terms
2) Using controlled vocab PECOS search terms
3) Using methodological/design PECOS search filters/limits
What does MeSH stand for?
Medical Subject Headings
What is Medlines controlled vocabulary thesaurus called?
MeSH
What are the 3 competing priorities that clinical studies require you to balance?
1) Design and conduct studies that provide the correct answers
2) Maximise the efficient use of resources
3) Comply with ethical, legal, institutional and professional regulations
Name the 4 potential sources of bias
Sampling/selection bias
Measurement bias
Analytical bias
Dissemination bias
What is sampling/selection bias?
Using non representative samples
This selection influences exposure and outcome
What are the 3 types of measurement bias?
Information bias - extent of information varies amongst participants
Observer bias - influence by prior knowledge/belief
Recall/response/prestige bias - influence by prior knowledge or belief
What are the 3 types of analytical bias?
Loss to follow up - specific participants excluded
Omitted variable bias - imprecise adjustment for confounding
Attribution bias - interpretation of causality
What is the 1 type of dissemination bias?
Publication bias: eventful results more likely to be published
What different types of studies have been developed to combat bias?
Cross sectional (lowest effectiveness) Case control Cohort Trial Meta analysis (highest effectiveness)
What are the 2 types of descriptive (inductive) studies?
Case study/series
Cross-sectional (descriptive)
What 2 categories are analytical studies further sub divided into?
Observational
Experimental
What are the 4 types of observational (selective sampling) studies?
Cross sectional (analytical)
Case/control
Cohort
What are the 2 types of experimental (selective exposure) studies?
Trial
Meta-analysis
What are the 3 key points outlined in the Belmont report in regards to ethics in research?
Respect for the person by gaining consent
Justice - equal burdens and benefits
Beneficence - risk is justified by knowledge gained by the study
What are the 4 types of potential sources of harm to participants?
Biological
Psychological
Social
Structural
What types of projects require formal ethical approval?
Research involving animals
Service evaluations involving vulnerable people
Projects involving a new intervention
Projects collecting new information on more than only existing service delivery
What is the definition of a target population?
The total finite population we wish to know about from which your sample is drawn
What is the definition of a study sample?
The participants drawn from the target population that constitutes our data set
What are the 3 different types of samples?
Complete samples
Unstratified random sampling
Stratified random sampling
What are complete samples?
The entire study population
What is unstratified random sampling?
Every member of the target population has the same chance of being sampled
What is stratified random sampling?
Randomly sample from the target population within each different variable e.g. age, sex etc
Every memeber within each group has the same chance of being sampled
At what figure does the p-value show that something is statistically significant?
<0.05
A 95% confidence interval without the null is the same as what?
a p-value of 0.05
What is the definition of ‘power’ in the context of research?
The probability of rejecting the null hypothesis when the null hypothesis is false
Making sure the sample size is big enough to prove an effect
How is the odds ratio calculated in categorical data?
Odds of the exposed/unexposed groups
What value that if it is included in the confidence interval shows that it is not significant in categorical data ?
1
What 3 values are calculated in continuous data to calculate confidence interval?
Mean
Standard deviation
Estimated sample size
What value if not included in the confidence interval in continuous data shows it is significant?
0
How are the odds ratio for 2 groups calculated?
Odds of the event for exposed individuals/odds of an event for unexposed individuals
What is a covariante?
variables that affect the exposure/outcome
What 3 questions are used to assess which covarieantes may be important to include in an analytical study?
1) What are the (likely) causes of the exposure?
2) What are the (likely) causes of the outcome?
3) What are (likely to be) caused by the exposures?
The confounders are formed from what 2 covariates?
Causes the exposure
Causes the outcome
The competing exposure is formed from what 2 covariates?
Causes the outcome
Likely to be caused by the exposure
The mediators are also known as what covariate?
what is likely to cause the outcome
What needs to be adjusted for?
Confounders
What does DAG stand for?
Directed Acyclic Graph
What is the function of a DAG?
1) Summaries functional, theoretical and speculate relationships between variables
2) Identifies covariants acting as competing exposures so this can be adjusted for
3) Identify covarients as mediators so they are not adjusted for
Why do we need to adjust for confounders?
They create a psuedo-casual path between the outcome and the exposure which will generate a statistical relationship between the two even when non exists
Why do we not adjust for mediators?
They are part of the casual path between the outcome and the exposure and if we adjust for these we are adjusting out this potentially important pathway
Why might we adjust for competing exposures?
IF they cause a substantial amount of variation in the exposure, ‘adjusting this out’ can make any association between the remainder and the exposure easier to detect
What are the 2 types of study that your variables can be sourced from?
Prospective study - records variables over the study period, with the outcome measured subsequently e.g. cohort
Retrospective study - measures the outcome and then look backwards to measure the exposure and other variables e.g. case-control/cross sectional
What are the pro’s of prospective data collection?
Fewer sources of bias e.g. recall or ascertainment
Less chance of confounding
Some data can only be measured prospectively
What are the cons of prospective data collection?
Time and resource intensive
Tendency to collect data on more variables that you need or can use in your analyses
Subjects can drop out (bias)
Usually unfeasible for rare outcomes
What are the pro’s of retrospective data collection?
Data collection is less time and resource intensive
Allows oversampling of rare outcomes
What are the con’s of retrospective data collection?
More susceptible to bias (recruitment and data recall)
Some variables cannot be measured directly e.g. BMI
If data are from records, little control over these
Name an example of a prospective data collection study
Biobank
Name an example of a retrospective data collection study
Genome project
Name 2 ways to ensure a standardised data collection
A research diary - ensures you keep a timely record of what you do and any observations on data collection processes influencing data quality and completeness
A data collection proforma - ensures that variables are measured/recorded consistently
What is the requirement of a questionnaire styled data collection proforma for collecting participants responses?
Clearly written questions Unambiguous Definitive Instructive Closed/pre-coded answers
What is a proxy?
Not the variable you require but something close to it e.g. to know what people eat you could ask about how much veg they eat
What are 3 sources of measurement errors?
Inherent variation/Instability in variables: can be further exacerbated when using proxies
Imperfect measurement equipment/technique
The same variable arising from the same questions may be biased in different ways depending on the context
Name 4 types of biases that occur from the same variable arising from the same questions may be biased in different ways depending on the context
1) response bias - what the interviewer wants/gets
2) prestige bias - what appears favourable
3) recall bias - prospective/retrospective
4) multiple transcription from sources to proforma
What is a latent key variable?
A key variable that is missing
What should be done if a latent variable occurs?
If it is an exposure/outcome then you must find an alternative
If it is a potential cofounder, try using a proxy
What are the 3 different options if there is missing data in some key variables?
1) compare participants with complete and missing data - are they comparable? if not missing may cause bias
2) create a missing category for missing measurements
3) interpolate/statistically estimate what the missing measurement might/are likely to have been
What is the function of general linear modelling?
Tests the relationship between 2 variables
Works for many different types of variables
What is a nuisance variable?
aka confounders and competing exposures 0 need to adjust for these
What is the equation of the line used in the simplest linear model?
y=mx+b+e
e = residual variation (noise) - want this value to be as low as possible
If the exposure is categorical, what is written before regress in the stata command?
xi:
When is logistic regression used?
if the outcome is binary
What does R2(squared) represent?
The proportion of the variation explained by the linear model
Value is between 0-1
The higher the value, the more variable that is explained by your model - this is good
What is the adjusted R2(squared) used for?
When there are more variables, the R2(squared) will always improve so this accounts and adjusts for this for each variable
If the confidence interval includes 0, what does this mean?
There is a significant association
What are the advantages of qualitative studies?
Assume objectivity - researchers seek to avoid their own presence, behaviour or attitude affecting the results
Positivist researchers critically examine their methods and conclusions for possible bias therefore studies are able to be replicated
Can interrogate taken for granted concepts to give new insight, define problems/solutions or connect wider political and social views
Complex issues can be researched in depth
Give reasons WHY peoples behaviour is a certain way
What are the 4 steps of a qualitative study design?
1) theoretical framework
2) participant selection
3) setting
4) data collection
What is meant by ‘theoretical framework?’
Assumption on how you think the world works
What is ‘induction’ in the context of theoretical framework?
Generate universal statement from individual cases
What is ‘deduction’ in the context of theoretical framework?
Use theories to explain individual cases
How many people would you want to choose to sample in a qualitative study?
The amount that will reach saturation which is around 15-25 people per variable
What are the different groups of people that can be sampled in a qualitative study?
1) purposive: participants who have the potential to provide rich, relevant ans diverse data relevant to the research question
2) Maximum variability: people with a wide range of experiences and no preconceptions
3) Deviant cases: People who will increase explanation as they do not fit the norm
4) Convenience: e.g. using students at research based uni’s - may lead to bias
5) Theoretical: allowing new theories to be found
What data sources are used in qualitative studies?
Interviews
Focus groups
Be aware of when to use each one e.g. sensitive topics, confidentiality, stigmatised topics etc.
What is the definition of ethnography?
The scientific description of peoples and cultures with their customs, habits, and mutual differences
What does IPA stand for?
Interpretive Phenomological Analysis
What is IPA (Interpretive Phenomological Analysis )?
A method to focus on how interviewees make sense of things
What is the grounded theory?
Focuses on identifying shared meanings in the data without reference to existing theories
What different parts have to be assessed to deem a qualitative study trustworthy?
Demonstration of knowledge of the researcher
Demonstration of skill and systematic approach
Volume and richness of data supporting the claims made
Systematic formulation of tentative claims, testing and refinement
Plausibility of the claims in light of earlier theory and evidence
What is opportunity cost?
the value of something when a particular course of action is chosen e.g. going on holiday vs replacing a car
What does it mean if the opportunity cost is greater than the value of what you have chosen?
You have made the wrong choice
What does QALY stand for?
Quality adjusted life years
What is the definition of quality adjusted life years?
A measure of the output of healthcare that can measure all types of healthcare
What is the definition of economic evaluation?
the comparison of two or more alternative courses of action in terms of their costs and their outcomes
What does economic evaluation do?
Ensures the value of the outputs from an activity is greater than the value of the resources consumed by the activity
Provides information about how much it costs per unit of health gain to choose one course of action over another
What is the definition of efficiency?
maximising the benefit for the resources used - helps the most people at the least cost
What are the 2 types of efficiency and their definition?
Technical efficiency: meeting an given objective at the least cost
Allocative efficiency: producing the output the matches what the consumer wants (supply vs demand)
What are the 2 aspects of marginal analysis?
1) Marginal Benefit = the benefit from the next step
2) Marginal cost = the cost of taking the next step
What are the 3 types of economic evaluation that we use when we want to compare across conditions?
1) cost effectiveness analysis - life years gained
2) cost utility analysis - quality and quantity life
3) cost benefit analysis
What is the definition of cost effectiveness analysis?
benefits measured in terms of standard clinical outcomes
What is the incremental cost effectiveness ratio? (ICER)
A calculation of the (expected) cost per additional unit of health produced by a new intervention compared to the current practice
What is the cost effectiveness threshold?
The maximum amount the health service will pay per unit of health gained
What are monetary costs?
the physical money cost
What are non monetary costs?
E.g. Quality of life/access/health outcomes etc.
What is the deceleration that codes for medical ethics?
Deceleration of Helsinki
What does the deceleration of Helsinki say in regards to consent?
Free informed consent is required by participants, though there can be exceptions
If unable to gain consent then:
Their condition must be characteristic for the study
They must get consent as soon as possible
Gain approval from an ethics committee
What is the definition of tacit?
Opt out consent
What is the social value requirement?
Allows you to ask people to participate in research if there is a substantial gain for others - as being part of clinical trials is normally not in the best interest of the patient
What are the 3 requirements that qualify a research project to be ethical?
1) consent
2) well designed project
3) social value requirement
