RESS Flashcards by Aimee Cunningham

What type of study generates new knowledge where there is limited research evidence available?

Research

How well did you know this?

Not at all

Perfectly

Why type of study measures existing practice against evidence based standards?

Audit: does… reach a certain standard?

How well did you know this?

Not at all

Perfectly

What type of study evaluates a proposed service or current practice with the intention of generating information to inform local decision making?

Service evaluation: What standard is being achieved?

How well did you know this?

Not at all

Perfectly

What is an audit-cum- service evaluation?

Audit to find what is achieving the standard

Service evaluation to find factors as to why

How well did you know this?

Not at all

Perfectly

What is PECOS?

Patient 
Exposure 
Comparison 
Outcome 
Study design

How well did you know this?

Not at all

Perfectly

What bias may occur in sampling and selection?

External validity: non-representative samples

- Confounding: Selection influences exposure and outcome

How well did you know this?

Not at all

Perfectly

What bias may occur in Measurement?

Information: participants know different amounts
Observer: influenced by prior knowledge/belief
Recall
Response: what the interviewer wants
Prestige: what appears to be favourable

How well did you know this?

Not at all

Perfectly

What bias may occur in analysis?

Failure to follow up
Omitted variable: imprecise confounding adjustment
Attributional Bias: interpretation of causality

How well did you know this?

Not at all

Perfectly

What bias may occur in dissemination?

Publication bias: eventful results are more likely to be published

How well did you know this?

Not at all

Perfectly

Hierarchy of studies designed to avoid bias

Meta-analysis: Evidence of reproducibility and generalisability 
Trial: finding cause 
Cohort: Direction of link
Case-control: Links between
Cross-sectional: links within

How well did you know this?

Not at all

Perfectly

What is an inductive approach to study?

Descriptive: observations: Case control and cross sectional

How well did you know this?

Not at all

Perfectly

What is a deductive approach to study?

Analytical:
Observational - cross, case & cohort
Experimental - trial by selecting exposure

How well did you know this?

Not at all

Perfectly

What is governance?

Permission

How well did you know this?

Not at all

Perfectly

What are Belmont’s 3 ethical concepts?

Respect
Justice & equality
Beneficence

How well did you know this?

Not at all

Perfectly

What are the six methods for ensuring voluntary participation?

Informed with sheet
Consent with form
Reward free
Freedom to decline
Freedom to withdraw
Rights - confidential and anonymous

How well did you know this?

Not at all

Perfectly

What are vulnerable groups?

Children
Diminished autonomy
Needs - poor/unwell
Unable to consent

How well did you know this?

Not at all

Perfectly

High risk ethical issues?

Vulnerable participants
Covert data collection
Sensitive
Admin of drugs/fluid/food
Additional stress
Data transfer outside of EU

How well did you know this?

Not at all

Perfectly

When is no approval required?

Data published
Non-humans
Existing data
New info but only on one existing service delivery

How well did you know this?

Not at all

Perfectly

What should be in a project protocol?

Why
What involve
Ethical, legal and governance issues addressed

How well did you know this?

Not at all

Perfectly

Unstratified and stratified sampling pros and cons

Unstratified
Pro: Easy. Con: Smaller groups under represented by chance.
Stratified
Pro: Representative of population & unequal sampling improved power for rare strata.
Con: Strata may not be known

How well did you know this?

Not at all

Perfectly

Estimates and hypothesis

Estimation allows you to estimate an effect - hypothesis testing tells you how likely you are to see that effect by chance if there is no effect.

How well did you know this?

Not at all

Perfectly

When are confidence intervals smallest?

When there is less variation and a larger sample size

How well did you know this?

Not at all

Perfectly

What is power?

The probability of rejecting the null hypothesis when the null is false: ie the probability of finding an association is there is one to be found.

How well did you know this?

Not at all

Perfectly

When is it easiest to detect power?

Greater mean effect
Smaller variation in effect
Sample size is larger: ONLY ONE WE CAN CHANGE

How well did you know this?

Not at all

Perfectly

Reasoning for calculating power?

- Ensure a study is well designed. - If nothing found to show there was power to detect an association. Estimate how small associated must be for it to be missed.

What is an odds ratio?

Probability event occurs /probability it does not. | Odds of an event for exposed individuals/ odds of an event for unexposed. (ad/bc)

How do odds ratios form confidence intervals?

Exponential of long OR -/+ 1.96 root of 1/a+1/b + 1/c +1/d)

How much of the observation will be within 3 standard deviations of the mean? And what is the approximate difference between the highest and lowest values?

99.72% 6 x standard deviation

What covariate causes the outcome AND exposure?

Confounders

What covariate causes the outcome NOT the exposure?

Competing exposures

What covariate is caused by the exposure that causes the outcome?

Mediators

What is DAG?

Directed ACYCLIC Graph to summarise relationship between variable

Why should confounders be adjusted for?

- Generate relationship between exposure and outcome even though there isn't one

Why should competing exposures be adjusted for?

Adjustment can make association easier to detect

Why are mediators left alone?

- Part of causal pathway between exposure and outcome

What types of causal relationships are there?

- Functional - Empirical - based on previous analysis - Theoretical - Speculative

What is a retrospective study?

Case control, cross sectional: looking backwards to measure exposure. Less time & resources Allows over sampling for rare outcomes More susceptible to bias: recall & recruitment

What is a proforma?

Method of research collection: questionnaire

What is the purpose of general linear modelling?

Indicates and association between two variables (using regress on stata) If categorical variables: xi regress i categorical variable (logistic regression)

What is R2 and what does it mean if it = 1?

Proportion of variation explained by linear model. | 1 = perfect fit

What is research that assumes objectivity and includes critical examination for bias?

Positivist medical research

What checklist may be used for qualitative research on patient experience?

COREQ

What is purposive sampling?

- People who provide relevant rich and diverse data. | - Increased explanatory power through inclusion of deviant cases.

What can increase trustworthiness in obtaining qualitative data?

- Prolonged engagement - Multiple sources - Not foreclosing prematurely

What can be a threat to trustworthiness in qualitative data?

- Stereotyping - Access - Misreading - Translation

What are the three key aspects of health economics?

1. Opportunity cost 2. Efficiency 3. Marginal analysis

What is described as the benefit foregone from choosing something else?

Opportunity cost

What are Quality Adjusted Life Years?

- Measures output of healthcare and all types of effect: value

What is used to ensure value of outputs is greater than value of resources?

- Economic evaluation

What type of economic evaluation measured benefits as clinical outcomes, considering life years gained?

Cost effectiveness analysis

What type of economic evaluation considers both quantity and quality of life?

Cost utility analysis

What type of economic evaluation considers resource use and health benefits in monetary terms?

Cost benefit analysis

What is is technical efficiency?

Meeting an objective at the least cost

What is allocative efficiency

Producing supply that meets demand.

What is marginal analysis?

Compares benefit of next step (marginal benefit) with cost of next step (marginal cost)

What is ICER?

Incremental cost effectiveness ratio: estimated cost per additional unit of health produced by a new service

How is threshold of opportunity cost raised?

- Opportunity cost is at level of current worst treatment in QUALYs - becomes displaced by new treatment within budget = new threshold

What is the cost effectiveness threshold?

- Maximum opportunity cost consistent with improving population health by introducing a new intervention - NICE £20-30K per QALY gained

What is optional appraisal?

Evaluation alternative methods to achieve an agreed objective. Involves weighing up non-monetary and monetary costs of each option

What are non-monetary costs and benefits?

- QoL - Access - Health outcomes - Quality of care indicators - Patient safety - Patient preference - Adherence and compliance - Staff experience

What is the declaration of Helsinki?

Ethical research guidelines ensuring safety, effectiveness, efficiency and quality

What is the Social Value Requirement?

- Importance of objective outweighs risks and burdens to research subject. (improve health or important new knowledge/treatment )

How is consenting patients who are unable to consent for research approached?

- Consent gained ASAP from subject or legal representative

How should opt out/tacit consent be approached?

- Inform in advance and give easy option to opt out - Must know if they do not opt out then they have consented - Must be disseminated

RESS Flashcards

(64 cards)